The Safer Affordable Fuel-Efficient (SAFE) Vehicles Rule for Model Years 2021-2026 Passenger Cars and Light Trucks

Published date30 April 2020
Citation85 FR 24174
Record Number2020-06967
SectionRules and Regulations
CourtNational Highway Traffic Safety Administration
Federal Register, Volume 85 Issue 84 (Thursday, April 30, 2020)
[Federal Register Volume 85, Number 84 (Thursday, April 30, 2020)]
                [Rules and Regulations]
                [Pages 24174-25278]
                From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
                [FR Doc No: 2020-06967]
                [[Page 24173]]
                Vol. 85
                Thursday,
                No. 84
                April 30, 2020
                Part IV
                Book 2 of 3 Books
                Pages 24173-25278
                Environmental Protection Agency
                -----------------------------------------------------------------------
                40 CFR Parts 86 and 600
                Department of Transportation
                -----------------------------------------------------------------------
                National Highway Traffic Safety Administration
                49 CFR Parts 523, 531, 533, et al.
                -----------------------------------------------------------------------
                The Safer Affordable Fuel-Efficient (SAFE) Vehicles Rule for Model
                Years 2021-2026 Passenger Cars and Light Trucks; Final Rule
                Federal Register / Vol. 85 , No. 84 / Thursday, April 30, 2020 /
                Rules and Regulations
                [[Page 24174]]
                -----------------------------------------------------------------------
                ENVIRONMENTAL PROTECTION AGENCY
                40 CFR Parts 86 and 600
                DEPARTMENT OF TRANSPORTATION
                National Highway Traffic Safety Administration
                49 CFR Parts 523, 531, 533, 536, and 537
                [NHTSA-2018-0067; EPA-HQ-OAR-2018-0283; FRL 10000-45-OAR]
                RIN 2127-AL76; RIN 2060-AU09
                The Safer Affordable Fuel-Efficient (SAFE) Vehicles Rule for
                Model Years 2021-2026 Passenger Cars and Light Trucks
                AGENCY: Environmental Protection Agency and National Highway Traffic
                Safety Administration.
                ACTION: Final rule.
                -----------------------------------------------------------------------
                SUMMARY: EPA and NHTSA, on behalf of the Department of Transportation,
                are issuing final rules to amend and establish carbon dioxide and fuel
                economy standards. Specifically, EPA is amending carbon dioxide
                standards for model years 2021 and later, and NHTSA is amending fuel
                economy standards for model year 2021 and setting new fuel economy
                standards for model years 2022-2026. The standards set by this action
                apply to passenger cars and light trucks, and will continue our
                nation's progress toward energy independence and carbon dioxide
                reduction, while recognizing the realities of the marketplace and
                consumers' interest in purchasing vehicles that meet all of their
                diverse needs. These final rules represent the second part of the
                Administration's action related to the August 24, 2018 proposed Safer
                Affordable Fuel-Efficient (SAFE) Vehicles Rule. These final rules
                follow the agencies' actions, taken September 19, 2019, to ensure One
                National Program for automobile fuel economy and carbon dioxide
                emissions standards, by finalizing regulatory text related to
                preemption under the Energy Policy and Conservation Act and withdrawing
                a waiver previously provided to California under the Clean Air Act.
                DATES: This final rule is effective on June 29, 2020.
                 Judicial Review: NHTSA and EPA undertake this joint action under
                their respective authorities pursuant to the Energy Policy and
                Conservation Act and the Clean Air Act. Pursuant to CAA section 307(b),
                42 U.S.C. 7607(b), any petitions for judicial review of this action
                must be filed in the United States Court of Appeals for the D.C.
                Circuit. Given the inherent relationship between the agencies' action,
                any challenges to NHTSA's regulation under 49 U.S.C. 32909 should also
                be filed in the United States Court of Appeals for the D.C. Circuit.
                ADDRESSES: EPA and NHTSA have established dockets for this action under
                Docket ID Nos. EPA-HQ-OAR-2018-0283 and NHTSA-2018-0067, respectively.
                All documents in the docket are listed in the http://www.regulations.gov index. Although listed in the index, some
                information is not publicly available, e.g., confidential business
                information (CBI) or other information whose disclosure is restricted
                by statute. Certain other material, such as copyrighted material, will
                be publicly available in hard copy in EPA's docket, and electronically
                in NHTSA's online docket. Publicly available docket materials can be
                found either electronically in www.regulations.gov by searching for the
                dockets using the Docket ID numbers above, or in hard copy at the
                following locations:
                 EPA: EPA Docket Center, EPA/DC, EPA West, Room 3334, 1301
                Constitution Ave. NW, Washington, DC. The Public Reading Room is open
                from 8:30 a.m. to 4:30 p.m., Monday through Friday, excluding legal
                holidays. The telephone number for the Public Reading Room is (202)
                566-1744.
                 NHTSA: Docket Management Facility, M-30, U.S. Department of
                Transportation (DOT), West Building, Ground Floor, Rm. W12-140, 1200
                New Jersey Ave. SE, Washington, DC 20590. The DOT Docket Management
                Facility is open between 9 a.m. and 5 p.m. Eastern Time, Monday through
                Friday, except Federal holidays.
                FOR FURTHER INFORMATION CONTACT: EPA: Christopher Lieske, Office of
                Transportation and Air Quality, Assessment and Standards Division,
                Environmental Protection Agency, 2000 Traverwood Drive, Ann Arbor, MI
                48105; telephone number: (734) 214-4584; fax number: (734) 214-4816;
                email address: [email protected], or contact the Assessment
                and Standards Division, email address: [email protected]. NHTSA: James Tamm,
                Office of Rulemaking, Fuel Economy Division, National Highway Traffic
                Safety Administration, 1200 New Jersey Avenue SE, Washington, DC 20590;
                telephone number: (202) 493-0515.
                SUPPLEMENTARY INFORMATION:
                Does this action apply to me?
                 This action affects companies that manufacture or sell new light-
                duty vehicles, light-duty trucks, and medium-duty passenger vehicles,
                as defined under EPA's CAA regulations,\1\ and passenger automobiles
                (passenger cars) and non-passenger automobiles (light trucks) as
                defined under NHTSA's CAFE regulations.\2\ Regulated categories and
                entities include:
                ---------------------------------------------------------------------------
                 \1\ ``Light-duty vehicle,'' ``light-duty truck,'' and ``medium-
                duty passenger vehicle'' are defined in 40 CFR 86.1803-01. Generally
                speaking, a ``light-duty vehicle'' is a passenger car, a ``light-
                duty truck'' is a pick-up truck, sport-utility vehicle, or minivan
                up to 8,500 lbs. gross vehicle weight rating, and a ``medium-duty
                passenger vehicle'' is a sport-utility vehicle or passenger van from
                8,500 to 10,000 lbs. gross vehicle weight rating.
                 \2\ ``Passenger car'' and ``light truck'' are defined in 49 CFR
                part 523.
                ---------------------------------------------------------------------------
                [[Page 24175]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.000
                 This list is not intended to be exhaustive, but rather provides a
                guide regarding entities likely to be regulated by this action. To
                determine whether particular activities may be regulated by this
                action, you should carefully examine the regulations. You may direct
                questions regarding the applicability of this action to the person
                listed in FOR FURTHER INFORMATION CONTACT.
                I. Executive Summary
                II. Overview of Final Rule
                III. Purpose of the Rule
                IV. Purpose of Analytical Approach Considered as Part of Decision-
                Making
                V. Regulatory Alternatives Considered
                VI. Analytical Approach as Applied to Regulatory Alternatives
                VII. What does the analysis show, and what does it mean?
                VIII. How do the final standards fulfill the agencies' statutory
                obligations?
                IX. Compliance and Enforcement
                X. Regulatory Notices and Analyses
                I. Executive Summary
                 NHTSA (on behalf of the Department of Transportation) and EPA are
                issuing final rules to adopt and modify standards regulating corporate
                average fuel economy and tailpipe carbon dioxide (CO2)
                emissions and use/leakage of other air conditioning refrigerants for
                passenger cars and light trucks for MYs 2021-2026.\3\ These final rules
                follow the proposal issued in August 2018 and respond to each agency's
                legal obligation to set standards based on the factors Congress
                directed them to consider, as well as the direction of the United
                States Supreme Court in Massachusetts v. EPA, which stated that ``there
                is no reason to think the two agencies cannot both administer their
                obligations and yet avoid inconsistency.'' \4\ These standards are the
                product of significant and ongoing work by both agencies to craft
                regulatory requirements for the same group of vehicles and vehicle
                manufacturers. This work aims to facilitate, to the extent possible
                within the statutory directives issued to each agency, the ability of
                automobile manufacturers to meet all requirements under both programs
                with a single national fleet under one national program of fuel economy
                and tailpipe CO2 emission regulation.
                ---------------------------------------------------------------------------
                 \3\ Throughout this document and the accompanying FRIA, the
                agencies will often use the term ``CO2'' or ``tailpipe
                CO2'' to refer broadly to EPA's suite of light duty
                vehicle GHG standards.
                 \4\ 549 U.S. 497, 532 (2007).
                ---------------------------------------------------------------------------
                 The CAFE and CO2 emissions standards established by
                these final rules will increase in stringency at 1.5 percent per year
                from MY 2020 levels over MYs 2021-2026. The ``1.5 percent'' regulatory
                alternative is new for the final rule and was not expressly analyzed in
                the NPRM, but it is a logical outgrowth of the NPRM analysis, being
                well within the range of alternatives then considered and consistent
                with discussions by both the agencies and commenters that there are
                benefits to having standards that increase at the same rate for all
                fleets. These standards apply to light-duty vehicles, which NHTSA
                divides for purposes of regulation into passenger cars and light
                trucks, and EPA divides into passenger cars, light-duty trucks, and
                medium-duty passenger vehicles (i.e., sport utility vehicles, cross-
                over utility vehicles, and light trucks). Both the CAFE and
                CO2 standards are vehicle-footprint-based, as are the
                standards currently in effect. These standards will become more
                stringent for each model year from 2021 to 2026, relative to the MY
                2020 standards. Generally, the larger the vehicle footprint, the less
                numerically stringent the corresponding vehicle CO2 and
                miles-per-gallon (mpg) targets. As a result of the footprint-based
                standards, the burden of compliance is distributed across all vehicle
                footprints and across all manufacturers. Each manufacturer is subject
                to individualized standards for passenger cars and light trucks, in
                each model year, based on the vehicles it produces. When standards are
                carefully crafted, both in terms of the footprint curves and the rate
                of increase in stringency of those curves, manufacturers are not
                [[Page 24176]]
                compelled to build vehicles of any particular size or type.
                 Knowing that many readers are accustomed to considering CAFE and
                CO2 emissions standards in terms of the mpg and grams-per-
                mile (g/mi) values that the standards are projected to eventually
                require, the agencies include those projections here. EPA's standards
                are projected to require, on an average industry fleet-wide basis, 201
                grams per mile (g/mi) of CO2 in model year 2030, while
                NHTSA's standards are projected to require, on an average industry
                fleet-wide basis, 40.5 miles per gallon (mpg) in model year 2030. The
                agencies note that real-world CO2 is typically 25 percent
                higher and real-world fuel economy is typically 20 percent lower than
                the CO2 and CAFE compliance values discussed here, and also
                note that a portion of EPA's expected ``CO2'' improvements
                will in fact be made through improvements in minimizing air
                conditioning leakage and through use of alternative refrigerants, which
                will not contribute to fuel economy but will contribute toward
                reductions of climate-related emissions.
                 In these final rules, NHTSA and EPA are reaching similar
                conclusions on similar grounds: even though each agency has its own
                distinct statutory authority and factors, the relevant considerations
                overlap in many ways. Both agencies recognize that they are balancing
                the relevant considerations in somewhat different ways from how they
                may have been balanced previously, as in the 2012 final rule and in
                EPA's Initial Determination, but the current balancing is called for in
                light of the facts before the agencies. The balancing in these final
                rules is also somewhat different from how the agencies balanced their
                respective considerations in the proposal, in part because of updates
                to analytical inputs and methodologies, previewed in the NPRM and made
                in response to public comments, that collectively resulted in changes
                to the analytical outputs. For example, between the notice and final
                rule, the agencies updated fuel price projections to somewhat greater
                values, updated the analysis fleet to MY 2017, updated estimates of the
                efficacy and cost of fuel-saving technologies, revised procedures for
                calculating impacts on vehicle sales and scrappage, updated models for
                estimating highway safety impacts, updated estimates of highway
                congestion costs, and updated estimates of annual mileage accumulation,
                holding VMT (before applying the rebound effect) constant between
                regulatory alternative. Moreover, the cost-benefit analysis conducted
                for these final rules has even been overtaken by events in many ways
                over recent weeks. Based upon current events, and for additional
                reasons discussed in Section VI.D.1 the benefits of saving additional
                fuel through more stringent standards are potentially even smaller than
                estimated in this rulemaking analysis.
                 The standards finalized today fit the pattern of gradual, tough,
                but feasible stringency increases that take into account real world
                performance, shifts in fuel prices, and changes in consumer behavior
                toward crossovers and SUVs and away from more efficient sedans. This
                approach ensures that manufacturers are provided with sufficient lead
                time to achieve standards, considering the cost of compliance. The
                costs to both industry and automotive consumers would have been too
                high under the standards set forth in 2012, and by lowering the auto
                industry's costs to comply with the program, with a commensurate
                reduction in per-vehicle costs to consumers, the standards enhance the
                ability of the fleet to turn over to newer, cleaner and safer vehicles.
                 More stringent standards also have the potential for overly
                aggressive penetration rates for advanced technologies relative to the
                penetration rates seen in the final standards, especially in the face
                of an unknown degree of consumer acceptance of both the increased costs
                and of the technologies themselves--particularly given current
                projections of relatively low fuel prices during that timeframe. As a
                kind of insurance policy against future fuel price volatility,
                standards that increase at 1.5 percent per year for cars and trucks
                will help to keep fleet fuel economy higher than they would be
                otherwise when fuel prices are low, which is not improbable over the
                next several years.\5\ At the same time, the standards help to address
                these issues by maintaining incentives to promote broader deployment of
                advanced technologies, and so provides a means of encouraging their
                further penetration while leaving manufacturers alternative technology
                choices. Steady, gradual increases in stringency ensure that the
                benefits of reduced GHG emissions and fuel consumption are achieved
                without the potential for disruption to automakers or consumers.
                ---------------------------------------------------------------------------
                 \5\ For example, EIA currently expects U.S. retail gasoline
                prices to average $2.14/gallon in 2020, compared to $2.69/gallon in
                2019 (see https://www.eia.gov/outlooks/steo/archives/mar20.pdf), and
                $3.68/gallon in 2012 (see https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=PET&s=EMM_EPM0_PTE_NUS_DPG&f=A). While gasoline
                prices may foreseeably rise over the rulemaking time frame, it is
                also very foreseeable that they will not rise to the $4-5/gallon
                that many Americans saw over the 2008-2009 time frame, that caused
                the largest shift seen toward smaller and higher-fuel-economy
                vehicles. See, e.g., Figure VIII-2 below.
                ---------------------------------------------------------------------------
                 Standards that increase at 1.5 percent per year represent a
                reasonable balance of additional technology and required per-vehicle
                costs, consumer demand for fuel economy, fuel savings and emissions
                avoided given the foreseeable state of the global oil market and the
                minimal effect on climate between finalizing 1.5 percent standards
                versus more stringent standards. The final standards will also result
                in year-over-year improvements in fleetwide fuel economy, resulting in
                energy conservation that helps address environmental concerns,
                including criteria pollutant, air toxic pollutant, and carbon
                emissions.
                 The agencies project that under these final standards, required
                technology costs would be reduced by $86 to $126 billion over the
                lifetimes of vehicles through MY 2029. Equally important, purchase
                prices costs to U.S. consumers for new vehicles would be $977 to $1,083
                lower, on average, than they would have been if the agencies had
                retained the standards set forth in the 2012 final rule and originally
                upheld by EPA in January 2017. While these final standards are
                estimated to result in 1.9 to 2.0 additional billion barrels of fuel
                consumed and from 867 to 923 additional million metric tons of
                CO2 as compared to current estimates of what the standards
                set forth in 2012 would require, the agencies explain at length below
                why the overall benefits of the final standards outweigh these
                additional costs.\6\
                ---------------------------------------------------------------------------
                 \6\ 1.9 to 2.0 barrels of fuel is approximately 78 to 84 gallons
                of fuel.
                ---------------------------------------------------------------------------
                 For the CAFE program, overall (fleetwide) net benefits vary from
                $16.1 billion at a 7 percent discount rate to -$13.1 billion at a 3
                percent discount rate. For the CO2 program, overall
                (fleetwide) societal net benefits vary from $6.4 billion at a 7 percent
                discount rate to -$22.0 billion at a 3 percent discount rate. The net
                benefits straddle zero, and are very small relative to the scale of
                reduced required technology costs, which range from $86.3 billion to
                $126.0 billion for the CAFE and CO2 programs across 7
                percent and 3 percent discount rates. Likewise, net benefits are very
                small relative to the scale of reduced retail fuel savings over the
                full life of all vehicles manufactured during the 2021 through 2029
                model years, which range from $108.6 billion to $185.1 billion for the
                CAFE and CO2 programs across 7 percent and 3 percent
                discount rates. Similarly, all of the alternatives have small net
                benefits, ranging from $18.4 billion to -$31.1
                [[Page 24177]]
                billion for the CAFE and CO2 programs across 7 percent and 3
                percent discount rates.\7\
                ---------------------------------------------------------------------------
                 \7\ See Table II-12 to Table II-15 for costs, benefits and net
                benefits.
                ---------------------------------------------------------------------------
                 NHTSA and EPA believe their analysis of the final rule represents
                the best available science, evidence, and methodologies for assessing
                the impacts of changes in CAFE and CO2 emission standards.
                In fact, the agencies note that today's analysis represents a marked
                improvement over prior rulemakings. Previously, the agencies were
                unable to model the impact of the standards on new vehicle sales or the
                retirement of older vehicles in the fleet, and, instead, were forced to
                assume, contrary to economic theory and empirical evidence, that the
                number of new vehicles sold and older vehicles scrapped remained static
                across regulatory alternatives. Today's analysis--as commenters to
                previous rulemakings and EPA's Science Advisory Board have argued is
                necessary \8\--quantifies the sales and scrappage impacts of the
                standards, including the associated safety benefits, and represents a
                significant step forward in agencies' ability to comprehensively
                analyze the impacts of CAFE and CO2 emission standards.
                ---------------------------------------------------------------------------
                 \8\ Science Advisory Board, U.S. EPA. Review of EPA's Proposed
                SAFE rule at 4 (Feb. 27, 2020), available at https://
                yosemite.epa.gov/sab/sabproduct.nsf/LookupWebProjectsCurrentBOARD/
                1FACEE5C03725F268525851F006319BB/$File/EPA-SAB-20-003+.pdf
                [hereinafter ``SAB Report''].
                ---------------------------------------------------------------------------
                 However, the agencies also believe it is important to be
                transparent about analytical limitations. For example, EPA's Science
                Advisory Board stressed that the agencies account for ``evolving
                consumer preferences for performance and other vehicle attributes,''
                \9\ yet due to limitations on the agencies' current ability to model
                buyers' choices among combinations of various attributes and their
                costs, the primary analysis does not account for the consumer benefits
                of other vehicle features that may be sacrificed for costly
                technologies that improve fuel economy. The agencies' analysis assumes
                that under these final standards, attributes of new cars and light
                trucks other than fuel economy would remain identical to those under
                the baseline standards, so that changes in sales prices and fuel
                economy would be the only sources of benefits or costs to new car and
                light truck buyers. In other words, the agencies' primary analysis does
                not consider that producers will likely respond to buyers' demands by
                reallocating some their savings in production costs due to lower
                technology costs to add or improve other attributes that consumers
                value more highly than the increases in fuel economy the augural
                standards would have required. The agencies have long debated whether
                and how best to model the consumer benefits of other vehicle
                attributes, and note that they have made considerable progress.\10\
                However, despite these potential analytical shortcomings, the agencies
                reaffirm that today's analysis represents the most complete and
                rigorous examination of CAFE and CO2 emission standards to
                date, and provide decision-makers a powerful analytical tool--
                especially since the limitations are known, do not bias the central
                analysis' results, and are afforded due consideration.
                ---------------------------------------------------------------------------
                 \9\ SAB at 10.
                 \10\ In their evaluations of previous CAFE and CO2
                rules, the agencies attempted to account for this possibility by
                conducting sensitivity analyses that reduced the fuel savings and
                other benefits to vehicle buyers by a significant fraction. For
                example, NHTSA's analysis supporting the Final Rule establishing
                CAFE standards for model year 2012-16 cars and light trucks tested
                the sensitivity of their central estimates of social costs and
                benefits to the assumptions that 25 percent and 50 percent of
                benefits to buyers were offset by opportunity costs of foregone
                improvements in attributes other than fuel economy; see NHTSA, Final
                Regulatory Impact Analysis: Corporate Average Fuel Economy for Model
                year 2012-16 Passenger Cars and Light Trucks, March 2010, at 563-565
                and Table X-9, at 566-56; see also, NHTSA, Final Regulatory Impact
                Analysis: Corporate Average Fuel Economy for Model year 2017-25
                Passenger Cars and Light Trucks, August 2012, at 1087 and Tables X-
                18a, X-18b, and X-18c, at 1099-1104. The agencies acknowledged that
                this was not a completely satisfactory way to represent the
                sacrifices in vehicles' other attributes that car and light truck
                manufacturers might find it necessary to make in order to comply
                with the increasingly stringent standards those previous rules
                established. At the time, however, the agencies were unable to
                identify specific attributes that manufacturers were most likely to
                sacrifice, measure the tradeoffs between increased fuel economy and
                improvements in those attributes, or assess the potential losses in
                utility to car and light truck buyers. In an effort to improve on
                their previous treatment of this issue, the agencies' evaluation of
                this final rule includes a sensitivity case that assumes
                manufacturers redirect their technology cost savings from complying
                with less stringent standards to instead improve a combination of
                cars' and light trucks' other attributes that offers benefits to
                their buyers significantly exceeding those costs. The magnitude of
                these (net) benefits is interpreted as the opportunity cost of the
                improvements in vehicles' other attributes that would have been
                sacrificed if the augural standards had been enacted. The method the
                agencies use to approximate these benefits, together with its effect
                on the rule's overall benefits and costs, is discussed in detail in
                Section VI.D.1.b)(8). Briefly, the results of this sensitivity
                analysis suggest the Final Rule would generate net benefits for the
                CAFE and CO2 programs ranging from $34.9 to $55.4 billion
                at 3% and 7% discount rates.
                ---------------------------------------------------------------------------
                 In terms of the agencies' respective statutory authorities, EPA is
                setting national tailpipe CO2 emissions standards for
                passenger cars and light trucks under section 202(a) of the Clean Air
                Act (CAA),\11\ and taking other actions under its authority to
                establish metrics and measure passenger car and light truck fleet fuel
                economy pursuant to the Energy Policy and Conservation Act (EPCA),\12\
                while NHTSA is setting national corporate average fuel economy (CAFE)
                standards under EPCA, as amended by the Energy Independence and
                Security Act (EISA) of 2007.\13\ As summarized above and as discussed
                in much greater detail below, the agencies believe that these represent
                appropriate levels of CO2 emissions standards and maximum
                feasible CAFE standards for MYs 2021-2026, pursuant to their respective
                statutory authorities. Sections III and VIII below contain detailed
                discussions of both agencies' statutory obligations and authorities.
                ---------------------------------------------------------------------------
                 \11\ 42 U.S.C. 7521(a).
                 \12\ 49 U.S.C. 32904(c).
                 \13\ 49 U.S.C. 32902.
                ---------------------------------------------------------------------------
                 Section 202(a) of the CAA requires EPA to establish standards for
                emissions of pollutants from new motor vehicles that cause or
                contribute to air pollution that may reasonably be anticipated to
                endanger public health or welfare. Standards under section 202(a) thus
                take effect only ``after providing such period as the Administrator
                finds necessary to permit the development and application of the
                requisite technology, giving appropriate consideration to the cost of
                compliance within such period.'' \14\ In establishing such standards,
                EPA must consider issues of technical feasibility, cost, and available
                lead time, among other things.
                ---------------------------------------------------------------------------
                 \14\ CAA Sec. 202(a); 42 U.S.C. 7512(a)(2).
                ---------------------------------------------------------------------------
                 EPCA, as amended by EISA, contains a number of provisions governing
                how NHTSA must set CAFE standards. EPCA requires that the Department of
                Transportation establish separate passenger car and light truck
                standards \15\ at ``the maximum feasible average fuel economy level
                that the Secretary decides the manufacturers can achieve in that model
                year,'' \16\ based on the agency's consideration of four statutory
                factors: technological feasibility, economic practicability, the effect
                of other standards of the Government on fuel economy, and the need of
                the United States to conserve energy.\17\ EPCA does not define these
                terms or specify what weight to give each concern in balancing them--
                such considerations are left within the discretion of the Secretary of
                Transportation (delegated to NHTSA) based upon current information.
                Accordingly, NHTSA interprets these factors and determines the
                appropriate weighting that leads to the maximum
                [[Page 24178]]
                feasible standards given the circumstances present at the time of
                promulgating each CAFE standard rulemaking. While EISA, for MYs 2011-
                2020, additionally required that standards increase ``ratably'' and be
                set at levels to ensure that the CAFE of the industry-wide combined
                fleet of new passenger cars and light trucks reach at least 35 mpg by
                MY 2020,\18\ EISA requires that standards for MYs 2021-2030 simply be
                set at the maximum feasible level as determined by the Secretary (and
                by delegation, NHTSA).\19\
                ---------------------------------------------------------------------------
                 \15\ 49 U.S.C. 32902(b)(1).
                 \16\ 49 U.S.C. 32902(a).
                 \17\ 49 U.S.C. 32902(f).
                 \18\ 49 U.S.C. 32902(b)(2)(A) and (C).
                 \19\ 49 U.S.C. 32902(b)(2)(B).
                ---------------------------------------------------------------------------
                 In the NPRM, the agencies sought comment on a variety of possible
                changes to existing compliance flexibilities that have been created
                over the past several years. The vast majority of the existing
                compliance flexibilities are not being changed, but a small number of
                flexibilities related to real-world fuel efficiency improvements are
                being finalized. In addition, EPA will continue to allow manufacturers
                to make improvements relating to air conditioning refrigerants and
                leakage and will credit those improvements toward CO2
                compliance, and EPA is making no changes in the amounts of credits
                available. EPA is also not making any changes to the existing
                CH4 and N2O standards. EPA is also extending the
                ``0 g/mi upstream'' incentive for electric vehicles beyond its current
                sunset of MY 2021, through MY 2026. EPA is also establishing a credit
                multiplier for natural gas vehicles through the 2026 model year.
                Otherwise, compliance flexibilities in the two programs do not change
                significantly for the final rule. These changes should help to
                streamline manufacturer use of those flexibilities in certain respects.
                While manufacturers and suppliers sought a number of other additional
                compliance flexibilities, the agencies have concluded that the
                aforementioned existing flexibilities are reasonable and appropriate,
                and that additional flexibilities are not justified.
                 Table I-1 and Table I-2 present the total costs, benefits, and net
                benefits for the 2021-2026 preferred alternative CAFE and
                CO2 levels, relative to the MY 2022-2025 existing/augural
                standards (with the MY 2025 standards repeated for MY 2026) and current
                MY 2021 standard. The preferred alternative exhibits a stringency rate
                increase of 1.5 percent per year for both passenger cars and light
                trucks. The values in Table I-1 and Table I-2 display (in total and
                annualized forms) costs for all MYs 1978-2029 vehicles, and the
                benefits and net benefits represent the impacts of the standards over
                the full lifetimes of the vehicles sold or projected to be sold during
                model years 1978-2029.
                 For this analysis, negative signs are used for changes in costs or
                benefits that decrease from those that would have resulted from the
                existing/augural standards. Any changes that would increase either
                costs or benefits are shown as positive changes. Thus, an alternative
                that decreases both costs and benefits, will show declines (i.e., a
                negative sign) in both categories. From Table I-1 and Table I-2, the
                preferred alternative (Alternative 3) is estimated to decrease costs
                relative to the baseline by $182 to $280 billion over the lifetime of
                MYs 1978-2029 passenger vehicles (range determined by discount rate
                across both CAFE and CO2 programs). It will also decrease
                benefits from $175 to $294 billion over the life of these MY fleets.
                The net impact will be a decrease from $22 billion to an increase of
                $16 billion in total net benefits to society over this roughly 52-year
                timeframe. Annualized, this amounts to roughly -$0.8 to 1.2 billion in
                net benefits per year.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.001
                [GRAPHIC] [TIFF OMITTED] TR30AP20.002
                [[Page 24179]]
                 Table I-3 and Table I-4 lists costs, benefits, and net benefits for
                all seven alternatives that were examined.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.003
                [GRAPHIC] [TIFF OMITTED] TR30AP20.004
                 Table I-5 and Table I-6 show a summary of various impacts of the
                preferred alternative for CAFE and CO2 standards. Impacts
                are presented in monetized and non-monetized values, as well as from
                the perspective of society and the consumer.
                [[Page 24180]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.005
                [[Page 24181]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.006
                BILLING CODE 4910-59-C
                 The agencies note that the NPRM drew more public comments (and,
                particularly, more pages of substantive comments) than any rulemaking
                in the history of the CAFE or CO2 tailpipe emissions
                programs--exceeding 750,000 comments. The agencies recognized in the
                NPRM that the proposal was significantly different from the final rules
                set forth in 2012, and explained at length the reasons for those
                differences--namely, that new information and considerations, along
                with an expanded and updated analysis, had led to different tentative
                conclusions. Today's final rules represent a further evolution of the
                work that supported the proposal, based on improved quantitative
                methodology and in careful consideration of the hundreds of thousands
                of public comments and deep reflection on the serious issues before the
                agencies. Simply put, the agencies have heard the comments, and today's
                analysis and decision reflect the agencies' grappling with the issues
                commenters raised, as well as all of the other information before the
                agencies. These programs and issues are weighty, and the agencies
                believe that a reasonable balance has been struck in these final rules
                between the many competing national needs that these regulatory
                programs collectively address.
                II. Overview of Final Rule
                A. Summary of Proposal
                 In the NPRM, the National Highway Traffic Safety Administration
                (NHTSA) and the Environmental Protection Agency (EPA) (collectively,
                ``the
                [[Page 24182]]
                agencies'') proposed the ``Safer Affordable Fuel-Efficient (SAFE)
                Vehicles Rule for Model Years 2021-2026 Passenger Cars and Light
                Trucks'' (SAFE Vehicles Rule). The proposed SAFE Vehicles Rule would
                set Corporate Average Fuel Economy (CAFE) and carbon dioxide
                (CO2) emissions standards, respectively, for passenger cars
                and light trucks manufactured for sale in the United States in model
                years (MYs) 2021 through 2026.\20\
                ---------------------------------------------------------------------------
                 \20\ NHTSA sets CAFE standards under the Energy Policy and
                Conservation Act of 1975 (EPCA), as amended by the Energy
                Independence and Security Act of 2007 (EISA). EPA sets
                CO2 standards under the Clean Air Act (CAA).
                ---------------------------------------------------------------------------
                 The agencies explained that they must act to propose and finalize
                these standards and do not have discretion to decline to regulate.
                Congress requires NHTSA to set CAFE standards for each model year.\21\
                Congress also requires EPA to set emissions standards for light-duty
                vehicles if EPA has made an ``endangerment finding'' that the pollutant
                in question--in this case, CO2--``cause[s] or contribute[s]
                to air pollution which may reasonably be anticipated to endanger public
                health or welfare.'' \22\ NHTSA and EPA proposed the standards
                concurrently because tailpipe CO2 emissions standards are
                directly and inherently related to fuel economy standards,\23\ and, if
                finalized, the rules would apply concurrently to the same fleet of
                vehicles. By working together to develop the proposals, the agencies
                aimed to reduce regulatory burden on industry and improve
                administrative efficiency.
                ---------------------------------------------------------------------------
                 \21\ 49 U.S.C. 32902.
                 \22\ 42 U.S.C. 7521; see also 74 FR 66495 (Dec. 15, 2009)
                (``Endangerment and Cause or Contribute Findings for Greenhouse
                Gases under Section 202(a) of the Clean Air Act'').
                 \23\ See, e.g., 75 FR 25324, at 25327 (May 7, 2010) (``The
                National Program is both needed and possible because the
                relationship between improving fuel economy and reducing tailpipe
                CO2 emissions is a very direct and close one. The amount
                of those CO2 emissions is essentially constant per gallon
                combusted of a given type of fuel. Thus, the more fuel efficient a
                vehicle is, the less fuel it burns to travel a given distance. The
                less fuel it burns, the less CO2 it emits in traveling
                that distance. [citation omitted] While there are emission control
                technologies that reduce the pollutants (e.g., carbon monoxide)
                produced by imperfect combustion of fuel by capturing or converting
                them to other compounds, there is no such technology for
                CO2. Further, while some of those pollutants can also be
                reduced by achieving a more complete combustion of fuel, doing so
                only increases the tailpipe emissions of CO2. Thus, there
                is a single pool of technologies for addressing these twin problems,
                i.e., those that reduce fuel consumption and thereby reduce
                CO2 emissions as well.'').
                ---------------------------------------------------------------------------
                 The agencies discussed some of the history leading to the proposal,
                including the 2012 final rule, the expectations regarding a mid-term
                evaluation as required by EPA regulation, and the rapid process over
                2016 and early 2017 by which EPA issued its first Final Determination
                that the CO2 standards set in 2012 for MYs 2022-2025
                remained appropriate based on the information then before the EPA
                Administrator.\24\ The agencies also discussed President Trump's
                direction in March 2017 to restore the original mid-term evaluation
                timeline, and EPA's subsequent information-gathering process and
                announcement that it would reconsider the January 2017
                Determination.\25\ EPA ultimately concluded that the standards set in
                2012 for MYs 2022-2025 were no longer appropriate.\26\ For NHTSA, in
                turn, the ``augural'' CAFE standards for MYs 2022-2025 were never
                final, and as explained in the 2012 final rule, NHTSA was obligated
                from the beginning to undertake a new rulemaking to set CAFE standards
                for MYs 2022-2025.
                ---------------------------------------------------------------------------
                 \24\ See 83 FR at 42987 (Aug.24, 2018).
                 \25\ Id.
                 \26\ 83 FR 16077 (Apr. 2, 2018).
                ---------------------------------------------------------------------------
                 The NPRM thus began the rulemaking process for both agencies to
                establish new standards for MYs 2022-2025 passenger cars and light
                trucks. Standards were concurrently proposed for MY 2026 in order to
                provide regulatory stability for as many years as is legally
                permissible for both agencies together. The NPRM also included revised
                standards for MY 2021 passenger cars and light trucks, because the
                agencies tentatively concluded, based on the information and analysis
                then before them, that the CAFE standards previously set for MY 2021
                were no longer maximum feasible, and the CO2 standards
                previously set for MY 2021 were no longer appropriate. Agencies always
                have authority under the Administrative Procedure Act to revisit
                previous decisions in light of new facts, as long as they provide
                notice and an opportunity for comment, and the agencies stated that it
                is plainly the best practice to do so when changed circumstances so
                warrant.\27\
                ---------------------------------------------------------------------------
                 \27\ See FCC v. Fox Television, 556 U.S. 502 (2009).
                ---------------------------------------------------------------------------
                 The NPRM proposed to maintain the CAFE and CO2 standards
                applicable in MY 2020 for MYs 2021-2026, and took comment on a wide
                range of alternatives, including different stringencies and retaining
                existing CO2 standards and the augural CAFE standards.\28\
                Table II-1, Table II-2, and Table II-3 show the estimates, under the
                NPRM analysis, of what the MY 2020 CAFE and CO2 curves would
                translate to, in terms of miles per gallon (mpg) and grams per mile (g/
                mi), in MYs 2021-2026, as well as the regulatory alternatives
                considered in the NPRM. In addition to retaining the MY 2020
                CO2 standards through MY 2026, EPA proposed and sought
                comment on excluding air conditioning refrigerants and leakage, and
                nitrous oxide and methane emissions for compliance with CO2
                standards after model year 2020, in order to improve harmonization with
                the CAFE program. EPA also sought comment on whether to change existing
                methane and nitrous oxide standards that were finalized in the 2012
                rule. The proposal was accompanied by a 1,600 page Preliminary
                Regulatory Impact Analysis (PRIA) and, for NHTSA, a 500 page Draft
                Environmental Impact Statement (DEIS), with more than 800 pages of
                appendices and the entire CAFE model, including the software source
                code and documentation, all of which were also subject to comment in
                their entirety and all of which received significant comments.
                ---------------------------------------------------------------------------
                 \28\ The agencies noted that this did not mean that the miles
                per gallon and grams per mile levels that were estimated for the MY
                2020 fleet in 2012 would be the ``standards'' going forward into MYs
                2021-2026. Both NHTSA and EPA set CAFE and CO2 standards,
                respectively, as mathematical functions based on vehicle footprint.
                These mathematical functions that are the actual standards are
                defined as ``curves'' that are separate for passenger cars and light
                trucks, under which each vehicle manufacturer's compliance
                obligation varies depending on the footprints of the cars and trucks
                that it ultimately produces for sale in a given model year. It was
                the MY 2020 CAFE and CO2 curves that the agencies
                proposed would continue to apply to the passenger car and light
                truck fleets for MYs 2021-2026. The mpg and g/mi values which those
                curves would eventually require of the fleets in those model years
                would be known for certain only at the ends of each of those model
                years. While it is convenient to discuss CAFE and CO2
                standards as a set ``mpg,'' ``g/mi,'' or ``mpg-e'' number,
                attempting to define those values based on the information then
                before the agency would necessarily end up being inaccurate.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24183]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.007
                [[Page 24184]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.008
                
                ---------------------------------------------------------------------------
                 \29\ The carbon dioxide equivalents of air conditioning
                refrigerant leakage, nitrous oxide emissions, and methane emissions
                were included for compliance with the EPA standards for all MYs
                under the baseline/no action alternative in the NPRM. Carbon dioxide
                equivalent is calculated using the Global Warming Potential (GWP) of
                each of the emissions.
                 \30\ Beginning in MY 2021, the proposal provided that the GWP
                equivalents of air conditioning refrigerant leakage, nitrous oxide
                emissions, and methane emissions would no longer be able to be
                included with the tailpipe CO2 for compliance with
                tailpipe CO2 standards.
                ---------------------------------------------------------------------------
                [[Page 24185]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.009
                BILLING CODE 4910-59-C
                 The agencies explained in the NPRM that new information had been
                gathered and new analysis performed since publication of the 2012 final
                rule establishing CAFE and CO2 standards for MYs 2017 and
                beyond and since issuance of the 2016 Draft TAR and EPA's 2016 and
                early 2017 ``mid-term evaluation'' process. This new information and
                analysis helped lead the agencies to the tentative conclusion that
                holding standards constant at MY 2020 levels through MY 2026 was
                maximum feasible, for CAFE purposes, and appropriate, for
                CO2 purposes.
                 The agencies further explained that technologies had played out
                differently in the fleet from what the agencies previously assumed:
                That while there remain a wide variety of technologies available to
                improve fuel economy and reduce CO2 emissions, it had become
                clear that there were reasons to temper previous optimism about the
                costs, effectiveness, and consumer acceptance of a number of
                technologies. In addition, over the years between the previous analyses
                and the NPRM, automakers had added considerable amounts of technologies
                to their new vehicle fleets, meaning that the agencies were no longer
                free to make certain assumptions about how some of those technologies
                could be used going forward. For example, some technologies that could
                be used to improve fuel economy and reduce emissions had not been used
                entirely for that purpose, and some of the benefit of these
                technologies had gone instead toward improving other vehicle
                attributes. Other technologies had been tried, and had been met with
                significant customer acceptance issues. The agencies underscored the
                importance of reflecting the fleet as it stands today, with the
                technology it has and as that technology has been used, and considering
                what technology remains on the table at this point, whether and when it
                can realistically be available for widespread use in production, and
                how much it would cost to implement.
                 The agencies also acknowledged the math of diminishing returns: As
                CAFE and CO2 emissions standards increase in stringency, the
                benefit of continuing to increase in stringency decreases. In mpg
                terms, a vehicle owner who drives a light vehicle 15,000 miles per year
                (a typical assumption for analytical purposes) \31\ and trades in a
                vehicle with fuel economy of 15 mpg for one with fuel economy of 20
                mpg, will reduce their annual fuel consumption from 1,000 gallons to
                750 gallons--saving 250 gallons annually. If, however, that owner were
                to trade in a vehicle with fuel economy of 30 mpg for one with fuel
                economy of 40 mpg, the owner's annual gasoline consumption would drop
                from 500 gallons/year to 375 gallons/year--only 125 gallons even though
                the mpg improvement is twice as large. Going from 40 to 50 mpg would
                save only 75 gallons/year. Yet each additional fuel economy improvement
                becomes much more expensive as the easiest to achieve low-cost
                technological improvement options are chosen. In CO2 terms,
                if a vehicle emits 300 g/mi CO2,
                [[Page 24186]]
                a 20 percent improvement is 60 g/mi, so the vehicle would emit 240 g/
                mi; but if the vehicle emits 180 g/mi, a 20 percent improvement is only
                36 g/mi, so the vehicle would get 144 g/mi. In order to continue
                achieving similarly large (on an absolute basis) emissions reductions,
                the percentage reduction must also continue to increase.
                ---------------------------------------------------------------------------
                 \31\ A different vehicle-miles-traveled (VMT) assumption would
                change the absolute numbers in the example, but would not change the
                mathematical principles.
                ---------------------------------------------------------------------------
                 Related, average real-world fuel economy is lower than average fuel
                economy required under CAFE and CO2 standards. The 2012
                Federal Register notice announcing augural CAFE and CO2
                standards extending through MY 2025 indicated that, if met entirely
                through the application of fuel-saving technology, the MY 2025
                CO2 standards would result in an average requirement
                equivalent to 54.5 mpg. However, because the CO2 standards
                provide credit for reducing leakage of AC refrigerants and/or switching
                to lower-GWP refrigerants, and these actions do not affect fuel
                economy, the notice explained that the corresponding fuel economy
                requirement (under the CAFE program) would be 49.7 mpg. These estimates
                were based on a market forecast grounded in the MY 2008 fleet. The
                notice also presented analysis using a market forecast grounded in the
                MY 2010 fleet, showing a 48.7 mpg average CAFE requirement.
                 In the real world, fuel economy is, on average, about 20% lower
                than as measured under regulatory test procedures. In the real world,
                then, these new standards were estimated to require 39.0-39.8 mpg.
                 Today's analysis indicates that the requirements under the
                baseline/augural CAFE standards would average 46.6 mpg in MY 2029. The
                lower value results from changes in the fleet forecast which reflects
                consumer preference for larger vehicles than was forecast for the 2012
                rulemaking. In the real world, the requirements average about 37.1 mpg.
                Under the final standards issued today, the regulatory test procedure
                requirements average 40.5 mpg, corresponding to 33.2 mpg in the real
                world. Buyers of new vehicles experience real-world fuel economy, with
                levels varying among drivers (due to a wide range of factors). Vehicle
                fuel economy labels provide average real-world fuel economy information
                to buyers.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.010
                 Vehicle owners also face fuel prices at the pump. The agencies
                noted in the NPRM that when fuel prices are high, the value of fuel
                saved may be enough to offset the cost of further fuel economy/
                emissions reduction improvements, but the agencies recognized that
                then-current projections of fuel prices by the Energy Information
                Administration did not indicate particularly high fuel prices in the
                foreseeable future. The agencies explained that fundamental structural
                shifts had occurred in global oil markets since the 2012 final rule,
                largely due to the rise of U.S. production and export of shale oil. The
                consequence over time of diminishing returns from more stringent fuel
                economy/emissions reduction standards, especially when combined with
                relatively low fuel prices, is greater difficulty for automakers to
                find a market of consumers willing to buy vehicles that meet the
                increasingly stringent standards. American consumers have long
                demonstrated that in times of relatively low fuel prices, fuel economy
                is not a top priority for the majority of them, even when highly fuel
                efficient vehicle models are available.
                 The NPRM analysis sought to improve how the agencies captured the
                effects of higher new vehicle prices on fleet composition as a whole by
                including an improved model for vehicle scrappage rates. As new vehicle
                prices increase, consumers tend to continue using older vehicles for
                longer, slowing fleet turnover and thus slowing improvements in fleet-
                wide fuel economy, reductions in CO2 emissions, reductions
                in criteria pollutant emissions, and advances in safety. That aspect of
                the analysis was also driven by the agencies' updated estimates of
                average per-vehicle cost increases due to
                [[Page 24187]]
                higher standards, which were several hundred dollars higher than
                previously estimated. The agencies cited growing concerns about
                affordability and negative equity for many consumers under these
                circumstances, as loan amounts grow and loan terms extend.
                 For all of the above reasons, the agencies proposed to maintain the
                MY 2020 fuel economy and CO2 emissions standards for MYs
                2021-2026. The agencies explained that they estimated, relative to the
                standards for MYs 2021-2026 put forth in 2012, that an additional 0.5
                million barrels of oil would be consumed per day (about 2 to 3 percent
                of projected U.S. consumption) if that proposal were finalized, but
                that they also expected the additional fuel costs to be outweighed by
                the cost savings from new vehicle purchases; that more than 12,700 on-
                road fatalities and significantly more injuries would be prevented over
                the lifetimes of vehicles through MY 2029 as compared to the standards
                set forth in the 2012 final rule over the lifetimes of vehicles as more
                new and safer vehicles are purchased than the current (and augural)
                standards; and that environmental impacts, on net, would be relatively
                minor, with criteria and toxic air pollutants not changing noticeably,
                and with estimated atmospheric CO2 concentrations increasing
                by 0.65 ppm (a 0.08 percent increase), which the agencies estimated
                would translate to 0.003 degrees Celsius of additional temperature
                increase relative to the standards finalized in 2012.
                 Under the NPRM analysis, the agencies tentatively concluded that
                maintaining the MY 2020 curves for MYs 2021-2026 would save American
                auto consumers, the auto industry, and the public a considerable amount
                of money as compared to EPA retaining the previously-set CO2
                standards and NHTSA finalizing the augural standards. The agencies
                explained that this had been identified as the preferred alternative,
                in part, because it appeared to maximize net benefits compared to the
                other alternatives analyzed, and recognizing the statutory
                considerations for both agencies. Relative to the standards issued in
                2012, under CAFE standards, the NPRM analysis estimated that costs
                would decrease by $502 billion overall at a three-percent discount rate
                ($335 billion at a seven-percent discount rate) and benefits were
                estimated to decrease by $326 billion at a three-percent discount rate
                ($204 billion at a seven-percent discount rate). Thus, net benefits
                were estimated to increase by $176 billion at a three-percent discount
                rate and $132 billion at a seven-percent discount rate. The estimated
                impacts under CO2 standards were estimated to be similar,
                with net benefits estimated to increase by $201 billion at a three-
                percent discount rate and $141 billion at a seven-percent discount
                rate.
                 The NPRM also sought comment on a variety of potential changes to
                NHTSA's and EPA's compliance programs for CAFE and CO2 as
                well as related programs, including questions about automaker requests
                for additional flexibilities and agency interest in reducing market-
                distorting incentives and improving transparency; and on a proposal to
                withdraw California's CAA preemption waiver for its ``Advanced Clean
                Car'' regulations, with an accompanying discussion of preemption of
                State standards under EPCA.\32\ The agencies sought comment broadly on
                all aspects of the proposal.
                ---------------------------------------------------------------------------
                 \32\ Agency actions relating to California's CAA waiver and EPCA
                preemption have since been finalized, see 84 FR 51310 (Sept. 27,
                2019), and will not be discussed in great detail as part of this
                final rule.
                ---------------------------------------------------------------------------
                B. Public Participation Opportunities and Summary of Comments
                 The NPRM was published on NHTSA's and EPA's websites on August 2,
                2018, and published in the Federal Register on August 24, 2018,
                beginning a 60-day comment period. The agencies subsequently extended
                the official comment period for an additional three days, and left the
                dockets open for more than a year after the start of the comment
                period, considering late comments to the extent practicable. A separate
                Federal Register notice also published on August 24, 2018, which
                announced the locations, dates, and times of three public hearings to
                be held on the proposal: One in Fresno, California, on September 24,
                2018; one in Dearborn, Michigan, on September 25, 2018; and one in
                Pittsburgh, Pennsylvania, on September 26, 2018. Each hearing started
                at 10 a.m. local time; the Fresno hearing ended at 5:10 p.m. and
                resulted in a 235 page transcript; the Dearborn hearing ran until 5:26
                p.m. and resulted in a 330 page transcript; and the Pittsburgh hearing
                ran until 5:06 p.m. and also resulted in a 330 page transcript. Each
                hearing also collected several hundred pages of comments from
                participants, in addition to the hearing transcripts.
                 Besides the comments submitted as part of the public hearings,
                NHTSA's docket received a total of 173,359 public comments in response
                to the proposal as of September 18, 2019, and EPA's docket a total of
                618,647 public comments, for an overall total of 792,006. NHTSA also
                received several hundred comments on its DEIS to the separate DEIS
                docket. While the majority of individual comments were form letters,
                the agencies received over 6,000 pages of substantive comments on the
                proposal.
                 Many commenters generally supported the proposal and many
                commenters opposed it. Commenters supporting the proposal tended to
                cite concerns about the cost of new vehicles, while commenters opposing
                the proposal tended to cite concerns about additional fuel expenditures
                and the impact on climate change. Many comments addressed the modeling
                used for the analysis, and specifically the inclusion, operation, and
                results of the sales and scrappage modules that were part of the NPRM's
                analysis, while many addressed the NPRM's safety findings and the role
                that those findings played in the proposal's justification. Many other
                comments addressed California's standards and role in Federal decision-
                making; as discussed above, those comments are further summarized and
                responded to in the separate Federal Register notice published in
                September 2019. Nearly every aspect of the NPRM's analysis and
                discussion received some level of comment by at least one commenter.
                The comments received, as a whole, were both broad and deep, and the
                agencies appreciate the level of engagement of commenters in the public
                comment process and the information and opinions provided.
                C. Changes in Light of Public Comments and New Information
                 The agencies made a number of changes to the analysis between the
                NPRM and the final rule in response to public comments and new
                information that was received in those comments or otherwise became
                available to the agencies. While these changes, their rationales, and
                their effects are discussed in detail in the sections below, the
                following represents a high-level list of some of the most significant
                changes:
                 Some regulatory alternatives were dropped from
                consideration, and one was added;
                 updated analysis fleet, and changes to technologies on
                ``baseline'' vehicles within the fleet to reflect better their current
                properties and improve modeling precision;
                 no civil penalties assumed to be paid after MY 2020 under
                CAFE program;
                 updates and expansions in accounting for certain over-
                compliance
                [[Page 24188]]
                credits, including early credits earned in EPA's program;
                 updates and expansions to CAFE Model's technology paths;
                 updates to inputs defining the range of manufacturer-,
                technology-, and product-specific constraints;
                 updates to allow the model to adopt a more advanced
                technology if it is more cost-effective than an earlier technology on
                the path;
                 precision improvements to the modeling of A/C efficiency
                and off-cycle credits;
                 updates to model's ``effective cost'' metric;
                 extended explicit simulation of technology application
                through MY 2050;
                 expanded presentation of the results to include ``calendar
                year'' analysis;
                 quantifying different types of health impacts from changes
                in air pollution, rather than only accounting for such impacts in
                aggregate estimates of the social costs of air pollution;
                 updated costs to 2018 dollars;
                 updated fuel costs based on the AEO 2019 version of NEMS;
                 a variety of technology updates in response to comments
                and new information;
                 updated accounting of rebound VMT between regulatory
                alternatives;
                 updated estimates of the macroeconomic cost of petroleum
                dependence;
                 updated response of total new vehicle sales to increases
                in fuel efficiency and price; and
                 updated response of vehicle retirement rates to changes in
                new vehicle fuel efficiency and transaction price.
                 Sections IV and VI below discuss these updates in significant
                detail.
                D. Final Standards--Stringency
                 As explained above, the agencies have chosen to set CAFE and
                CO2 standards that increase in stringency by 1.5 percent
                year over year for MYs 2021-2026. Separately, EPA has decided to retain
                the A/C refrigerant and leakage and CH4 and N2O
                standards set forth in 2012 for MYs 2021 and beyond, and the stringency
                of the CO2 standards in this final rule reflect the
                ``offset'' also established in 2012 based on assumptions made at that
                time about anticipated HFC emissions reductions.
                 When the agencies state that stringency will increase at 1.5
                percent per year, that means that the footprint curves which actually
                define the standards for CAFE and CO2 emissions will become
                more stringent at 1.5 percent per year. Consistent with Congress's
                direction in EISA to set CAFE standards based on a mathematical
                formula, which EPA harmonized with for the CO2 emissions
                standards, the standard curves are equations, which are slightly
                different for CAFE and CO2, and within each program,
                slightly different for passenger cars and light trucks. Each program
                has a basic equation for a fleet standard, and then values that change
                to cause the stringency changes are the coefficients within the
                equations. For passenger cars, consistent with prior rulemakings, NHTSA
                is defining fuel economy targets as follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.011
                where:
                TARGETFE is the fuel economy target (in mpg) applicable to a
                specific vehicle model type with a unique footprint combination,
                a is a minimum fuel economy target (in mpg),
                b is a maximum fuel economy target (in mpg),
                c is the slope (in gallons per mile per square foot, or gpm, per
                square foot) of a line relating fuel consumption (the inverse of
                fuel economy) to footprint, and
                d is an intercept (in gpm) of the same line.
                 Here, MIN and MAX are functions that take the minimum and maximum
                values, respectively, of the set of included values. For example,
                MIN[40,35] = 35 and MAX(40, 25) = 40, such that MIN[MAX(40, 25), 35] =
                35.
                 For light trucks, also consistent with prior rulemakings, NHTSA is
                defining fuel economy targets as follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.012
                where:
                TARGETFE is the fuel economy target (in mpg) applicable to a
                specific vehicle model type with a unique footprint combination,
                a, b, c, and d are as for passenger cars, but taking values specific
                to light trucks,
                e is a second minimum fuel economy target (in mpg),
                f is a second maximum fuel economy target (in mpg),
                g is the slope (in gpm per square foot) of a second line relating
                fuel consumption (the inverse of fuel economy) to footprint, and
                h is an intercept (in gpm) of the same second line.
                 The final CAFE standards (described in terms of their footprint-
                based curves) are as follows, with the values for the coefficients
                changing over time:
                [[Page 24189]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.013
                 These equations are presented graphically below, where the x-axis
                represents vehicle footprint and the y-axis represents fuel economy,
                showing that in the CAFE context, targets are higher (fuel economy) for
                smaller footprint vehicles and lower for larger footprint vehicles:
                BILLING CODE 4910-59-C
                [[Page 24190]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.014
                BILLING CODE 4910-59-P
                [[Page 24191]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.015
                BILLING CODE 4910-59-C
                 EPCA, as amended by EISA, requires that any manufacturer's
                domestically-manufactured passenger car fleet must meet the greater of
                either 27.5 mpg on average, or 92 percent of the average fuel economy
                projected by the Secretary for the combined domestic and non-domestic
                passenger automobile fleets manufactured for sale in the U.S. by all
                manufacturers in the model year, which projection shall be published in
                the Federal Register when the standard for that model year is
                promulgated in accordance with 49 U.S.C. 32902(b).\33\ Any time NHTSA
                establishes or changes a passenger car standard for a model year, the
                MDPCS for that model year must also be evaluated or re-evaluated and
                established accordingly. Thus, this final rule establishes the
                applicable MDPCS for MYs 2021-2026. Table II-8 lists the minimum
                domestic passenger car standards.
                ---------------------------------------------------------------------------
                 \33\ 49 U.S.C. 32902(b)(4).
                 [GRAPHIC] [TIFF OMITTED] TR30AP20.016
                
                 EPA CO2 standards are as follows. Rather than expressing
                these standards as linear functions with accompanying minima and
                maxima, similar to the approach NHTSA has followed since 2005 in
                specifying attribute-based standards, the following tables specify flat
                standards that apply below and above specified footprints, and a linear
                function that applies between those footprints. The two approaches are
                mathematically identical. For passenger cars with a footprint of less
                than or equal to 41 square feet, the gram/mile CO2 target
                value is selected for the appropriate model year from Table II-9:
                [[Page 24192]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.017
                 For passenger cars with a footprint of greater than 56 square feet,
                the gram/mile CO2 target value is selected for the
                appropriate model year from Table II-10:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.018
                 For passenger cars with a footprint that is greater than 41 square
                feet and less than or equal to 56 square feet, the gram/mile
                CO2 target value is calculated using the following equation
                and rounded to the nearest 0.1 grams/mile.
                [[Page 24193]]
                Target CO2 = [a x f] + b
                Where f is the vehicle footprint and a and b are selected from Table
                II-11 for the appropriate model year:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.019
                 For light trucks with a footprint of less than or equal to 41
                square feet, the gram/mile CO2 target value is selected for
                the appropriate model year from Table II-12:
                [[Page 24194]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.020
                 For light trucks with a footprint greater than the minimum value
                specified in the table below for each model year, the gram/mile
                CO2 target value is selected for the appropriate model year
                from Table II-13:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.021
                [[Page 24195]]
                 For light trucks with a footprint that is greater than 41 square
                feet and less than or equal to the maximum footprint value specified in
                Table II-14 below for each model year, the gram/mile CO2
                target value is calculated using the following equation and rounded to
                the nearest 0.1 grams/mile.
                Target CO2 = (a x f) + b
                Where f is the footprint and a and b are selected from Table II-14
                below for the appropriate model year:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.022
                 These equations are presented graphically below, where the x-axis
                represents vehicle footprint and the y-axis represents the
                CO2 target. The targets are lower for smaller footprint
                vehicles and higher for larger footprint vehicles:
                BILLING CODE 4910-59-P
                [[Page 24196]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.023
                [[Page 24197]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.024
                BILLING CODE 4910-59-C
                 Except that EPA elected to apply a slightly different slope when
                defining passenger car targets, CO2 targets may be expressed
                as direct conversion of fuel economy targets, as follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.025
                where 8887 g/gal relates grams of CO2 emitted to gallons
                of fuel consumed, and OFFSET reflects the fact that that HFC
                emissions from lower-GWP A/C refrigerants and less leak-prone A/C
                systems are counted toward average CO2 emissions, but
                EPCA provides no basis to count reduced HFC emissions toward CAFE
                levels.
                 For the reader's benefit, Table II-15, Table II-16, and Table II-17
                show the estimates, under the final rule analysis, of what the MYs
                2021-2026 CAFE and CO2 curves would translate to, in terms
                of miles per gallon (mpg) and grams per mile (g/mi).
                BILLING CODE 4910-59-P
                [[Page 24198]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.026
                BILLING CODE 4910-59-C
                [[Page 24199]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.027
                 As the following tables demonstrate, averages of manufacturers'
                estimated requirements are more stringent (i.e., for CAFE, higher, and
                for CO2, lower) under the final standards than under the
                proposed standards:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.028
                [[Page 24200]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.029
                E. Final Standards--Impacts
                 This section summarizes the estimated costs and benefits of the MYs
                2021-2026 CAFE and CO2 emissions standards for passenger
                cars and light trucks, as compared to the regulatory alternatives
                considered. These estimates helped inform the agencies' choices among
                the regulatory alternatives considered and provide further confirmation
                that the final standards are maximum feasible, for NHTSA, and
                appropriate, for EPA. The costs and benefits estimated to result from
                the CAFE standards are presented first, followed by those estimated to
                result from the CO2 standards. For several reasons, the
                estimates for costs and benefits presented for the different programs,
                while consistent, are not identical. NHTSA's and EPA's standards are
                projected to result in slightly different fuel efficiency improvements.
                EPA's CO2 standard is nominally more stringent in part due
                to its assumptions about manufacturers' use of air conditioning
                leakage/refrigerant replacement credits, which are expected to result
                in reduced emissions of HFCs. NHTSA's final standards are based solely
                on assumptions about fuel economy improvements, and do not account for
                emissions reductions that do not relate to fuel economy. In addition,
                the CAFE and CO2 programs offer somewhat different program
                flexibilities and provisions, primarily because NHTSA is statutorily
                prohibited from considering some flexibilities when establishing CAFE
                standards, while EPA is not.\34\ The analysis underlying this final
                rule reflects many of those additional EPA flexibilities, which
                contributes to differences in how the agencies estimate manufacturers
                could comply with the respective sets of standards, which in turn
                contributes to differences in estimated impacts of the standards. These
                differences in compliance flexibilities are discussed in more detail in
                Section IX below.
                ---------------------------------------------------------------------------
                 \34\ See 49 U.S.C. 32902(h); CAA Sec. 202(a).
                ---------------------------------------------------------------------------
                 Table II-20 to Table II-23 present all subcategories of costs and
                benefits of this final rule for all seven alternatives proposed. Costs
                include application of fuel economy technology to new vehicles,
                consumer surplus, crash costs due to changes in VMT, as well as, noise
                and congestion. Benefits include fuel savings, consumer surplus,
                refueling time, and clean air.
                BILLING CODE 4910-59-P
                [[Page 24201]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.030
                [[Page 24202]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.031
                [[Page 24203]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.032
                [[Page 24204]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.033
                [[Page 24205]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.034
                [[Page 24206]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.035
                [[Page 24207]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.036
                [[Page 24208]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.037
                [[Page 24209]]
                BILLING CODE 4910-59-C
                F. Other Programmatic Elements
                1. Compliance and Flexibilities
                 Automakers seeking to comply with the CAFE and CO2
                standards are generally expected to add fuel economy-improving
                technologies to their new vehicles to boost their overall fleet fuel
                economy levels. Readers will remember that improving fuel economy
                directly reduces CO2 emissions, because CO2 is a
                natural and inevitable byproduct of fossil fuel combustion to power
                vehicles. The CAFE and CO2 programs contain a variety of
                compliance provisions and flexibilities to accommodate better
                automakers' production cycles, to reward real-world fuel economy
                improvements that cannot be reflected in the 1975-developed test
                procedures, and to incentivize the production of certain types of
                vehicles. While the agencies sought comment on a broad variety of
                changes and potential expansions of the programs' compliance
                flexibilities in the NPRM, the agencies determined, after considering
                the comments, to make a few changes to the flexibilities proposed in
                the NPRM in this final rule. The most noteworthy change is the
                retention, in the CO2 program, of the flexibilities that
                allow automakers to continue to use HFC reductions toward their
                CO2 compliance, and that extend the ``0 grams/mile''
                assumption for electric vehicles through MY 2026 (i.e., recognizing
                only the tailpipe emissions of full battery-electric vehicles and not
                recognizing the upstream emissions caused by the electricity usage of
                those vehicles). In the NPRM, EPA had proposed to remove and sought
                comment on removing those flexibilities from the CO2
                program, but determined not to remove them in this final rule. EPA and
                NHTSA are also removing from the programs, starting in MY 2022, the
                credit/FCIV for full-size pickup trucks that are either hybrids or
                over-performing by a certain amount relative to their targets, and
                allowing technology suppliers to begin the petition process for off-
                cycle credits/adjustments.
                 Table II-24, Table II-25, Table II-26, and Table II-27 provide a
                summary of the various compliance provisions in the two programs; their
                authorities; and any changes included as part of this final rule:
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.038
                [[Page 24210]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.039
                BILLING CODE 4910-59-P
                [[Page 24211]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.040
                
                ---------------------------------------------------------------------------
                 \35\ The CAFE program uses an energy efficiency metric and
                standards that are expressed in miles per gallon. For PHEVs and
                BEVs, to determine gasoline the equivalent fuel economy for
                operation on electricity, a Petroleum Equivalency Factor (PEF) is
                applied to the measured electrical consumption. The PEF for
                electricity was established by the Department of Energy, as required
                by statute, and includes an accounting for upstream energy
                associated with the production and distribution for electricity
                relative to gasoline. Therefore, the CAFE program includes upstream
                accounting based on the metric that is consistent with the fuel
                economy metric. The PEF for electricity also includes an incentive
                that effectively counts only 15 percent of the electrical energy
                consumed.
                ---------------------------------------------------------------------------
                 Providing a technology neutral basis by which manufacturers meet
                fuel economy and CO2 emissions standards encourages an
                efficient and level playing field. The agencies continue to have a
                desire to minimize incentives that disproportionately favor one
                technology over another. Some of this may involve regulations
                established by other Federal agencies. In the near future, NHTSA and
                EPA intend to work with other relevant Federal agencies to pursue
                regulatory means by which we can further ensure technology neutrality
                in this field.
                2. Preemption/Waiver
                 As discussed above, the issues of Clean Air Act waivers of
                preemption under Section 209 and EPCA/EISA preemption under 49 U.S.C.
                32919 are not addressed in today's final rule, as
                [[Page 24212]]
                they were the subject of a separate final rulemaking action by the
                agencies in September 2019. While many comments were received in
                response to the NPRM discussion of those issues, those comments have
                been addressed and responded to as part of that separate rulemaking
                action.
                III. Purpose of the Rule
                 The Administrative Procedure Act (APA) requires agencies to
                incorporate in their final rules a ``concise general statement of their
                basis and purpose.'' \36\ While the entire preamble document represents
                the agencies' overall explanation of the basis and purpose for this
                regulatory action, this section within the preamble is intended as a
                direct response to that APA (and related CAA) requirements. Executive
                Order 12866 further states that ``Federal agencies should promulgate
                only such regulations as are required by law, are necessary to
                interpret the law, or are made necessary by compelling public need,
                such as material failures of private markets to protect or improve the
                health and safety of the public, the environment, or the well-being of
                the American people.'' \37\ Section III.C of the FRIA accompanying this
                rulemaking discusses at greater length the question of whether a market
                failure exists that these final rules may address.
                ---------------------------------------------------------------------------
                 \36\ 5 U.S.C. 553(c); see also Clean Air Act section
                307(d)(6)(A), 42 U.S.C. 7607(d)(6)(A).
                 \37\ E.O. 12866, Section 1(a).
                ---------------------------------------------------------------------------
                 NHTSA and EPA are legally obligated to set CAFE and GHG standards,
                respectively, and do not have the authority to decline to regulate.\38\
                The agencies are issuing these final rules to fulfill their respective
                statutory obligations to provide maximum feasible fuel economy
                standards and limit emissions of pollutants from new motor vehicles
                which have been found to endanger public health and welfare (in this
                case, specifically carbon dioxide (CO2); EPA has already set
                standards for methane (CH4), nitrous oxide (N2O),
                and hydrofluorocarbons (HFCs) and is not revising them in this rule).
                Continued progress in meeting these statutory obligations is both
                legally necessary and good for America--greater energy security and
                reduced emissions protect the American public. The final standards
                continue that progress, albeit at a slower rate than the standards
                finalized in 2012.
                ---------------------------------------------------------------------------
                 \38\ For CAFE, see 49 U.S.C. 32902; for CO2, see 42
                U.S.C. 7521(a).
                ---------------------------------------------------------------------------
                 National annual gasoline consumption and CO2 emissions
                currently total about 140 billion gallons and 5,300 million metric
                tons, respectively. The majority of this gasoline (about 130 billion
                gallons) is used to fuel passenger cars and light trucks, such as will
                be covered by the CAFE and CO2 standards issued today.
                Accounting for both tailpipe emissions and emissions from ``upstream''
                processes (e.g., domestic refining) involved in producing and
                delivering fuel, passenger cars and light trucks account for about
                1,500 million metric tons (mmt) of current annual CO2
                emissions. The agencies estimate that under the standards issued in
                2012, passenger car and light truck annual gasoline consumption would
                steadily decline, reaching about 80 billion gallons by 2050. The
                agencies further estimate that, because of this decrease in gasoline
                consumption under the standards issued in 2012, passenger car and light
                truck annual CO2 emissions would also steadily decline,
                reaching about 1,000 mmt by 2050. Under the standards issued today, the
                agencies estimate that, instead of declining from about 140 billion
                gallons annually today to about 80 billion gallons annually in 2050,
                passenger car and light truck gasoline consumption would decline to
                about 95 billion gallons. The agencies correspondingly estimate that
                instead of declining from about 1,500 mmt annually today to about 1,000
                mmt annually in 2050, passenger car and light truck CO2
                emissions would decline to about 1,100 mmt. In short, the agencies
                estimate that under the standards issued today, annual passenger car
                and light truck gasoline consumption and CO2 emissions will
                continue to steadily decline over the next three decades, even if not
                quite as rapidly as under the previously-issued standards.
                 The agencies also estimate that these impacts on passenger car and
                light truck gasoline consumption and CO2 emissions will be
                accompanied by a range of other energy- and climate-related impacts,
                such as reduced electricity consumption (because today's standards
                reduce the estimated rate at which the market might shift toward
                electric vehicles) and increased CH4 and N2O
                emissions. These estimated impacts, discussed below and in the FEIS
                accompanying today's notice, are dwarfed by estimated impacts on
                gasoline consumption and CO2 emissions.
                 As explained above, these final rules set or amend fuel economy and
                carbon dioxide standards for model years 2021-2026. Many commenters
                argued that it was not appropriate to amend previously-established
                CO2 and CAFE standards, generally because those commenters
                believed that the administrative record established for the 2012 final
                rule and EPA's January 2017 Final Determination was superior to the
                record that informed the NPRM, and that that prior record led
                necessarily to the policy conclusion that the previously-established
                standards should remain in place.\39\ Some commenters similarly argued
                that EPA's Revised Final Determination--which, for EPA, preceded this
                regulatory action--was invalid because, they allege, it did not follow
                the procedures established for the mid-term evaluation that EPA
                codified into regulation,\40\ and also because the Revised Final
                Determination was not based on the prior record.\41\
                ---------------------------------------------------------------------------
                 \39\ Comments arguing that the prior record was superior to the
                current record, and thus a better basis for decision-making, will be
                addressed throughout the balance of this preamble.
                 \40\ 40 CFR 86.1818-12(h).
                 \41\ See, e.g., comments from the States and Cities, Attachment
                1, Docket No. NHTSA-2018-0067-11735, at 40-42; CARB, Detailed
                Comments, Docket No. NHTSA-2018-0067-11873, at 71-72; CBD et. al,
                Appendix A, Docket No. NHTSA-2018-0067-12000, at 214-228.
                ---------------------------------------------------------------------------
                 The agencies considered a range of alternatives in the proposal,
                including the baseline/no action alternative of retaining the existing
                EPA carbon dioxide standards. As the agencies explained in the
                proposal, the proposal was entirely de novo, based on an entirely new
                analysis reflecting the best and most up-to-date information available
                to the agencies.\42\ This rulemaking action is separate and distinct
                from EPA's Revised Final Determination, which itself was neither a
                proposed nor a final decision that the standards ``must'' be revised.
                EPA retained full discretion in this rulemaking to revise the standards
                or not revise them. In any event, the case law is clear that agencies
                are free to reconsider their prior decisions.\43\ With that legal
                principle in mind, the agencies agree with commenters that the amended
                (and new) CO2 and CAFE standards must be consistent with the
                [[Page 24213]]
                CAA and EPCA/EISA, respectively, and this preamble and the accompanying
                FRIA explain in detail why the agencies believe they are consistent.
                The section below discusses briefly the authority given to the agencies
                by their respective governing statutes, and the factors that Congress
                directed the agencies to consider as they exercise that authority in
                pursuit of fulfilling their statutory obligations.
                ---------------------------------------------------------------------------
                 \42\ 83 FR 42968, 42987 (Aug. 24, 2018).
                 \43\ See, e.g., Encino Motorcars, LLC v. Navarro, 136 S. Ct.
                2117, 2125 (2016) (``Agencies are free to change their existing
                policies as long as they provide a reasoned explanation for the
                change.''); FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515
                (2009) (When an agency changes its existing position, it ``need not
                always provide a more detailed justification than what would suffice
                for a new policy created on a blank slate. Sometimes it must--when,
                for example, its new policy rests on factual findings that
                contradict those which underlay its prior policy; or when its prior
                policy has engendered serious reliance interests that must be taken
                into account . . . . In such cases it is not that further
                justification is demanded by the mere fact of policy change, but
                that a reasoned explanation is needed for disregarding facts and
                circumstances that underlay or were engendered by the prior
                policy.'')
                ---------------------------------------------------------------------------
                A. EPA's Statutory Requirements
                 EPA is setting national CO2 standards for passenger cars
                and light trucks under Section 202(a) of the Clean Air Act (CAA).\44\
                Section 202(a) of the CAA requires EPA to establish standards for
                emissions of pollutants from new motor vehicles which cause or
                contribute to air pollution which may reasonably be anticipated to
                endanger public health or welfare.\45\ In establishing such standards,
                EPA considers issues of technical feasibility, cost, available lead
                time, and other factors. Standards under section 202(a) thus take
                effect only ``after providing such period as the Administrator finds
                necessary to permit the development and application of the requisite
                technology, giving appropriate consideration to the cost of compliance
                within such period.'' \46\ EPA's statutory requirements are further
                discussed in Section VIII.A.
                ---------------------------------------------------------------------------
                 \44\ 42 U.S.C. 7521(a).
                 \45\ See Coalition for Responsible Regulation v. EPA, 684 F.3d
                102, 114-115 (D.C. Cir. 2012) (`` `If EPA makes a finding of
                endangerment, the Clean Air Act requires the [a]gency to regulate
                emissions of the deleterious pollutant from new motor vehicles . . .
                . Given the non-discretionary duty in Section 202(a)(1) and the
                limited flexibility available under Section 202(a)(2), which this
                court has held related only to the motor vehicle industry, . . . EPA
                had no statutory basis on which it could ground [any] reasons for
                further inaction' '') (quoting Massachusetts v. EPA, 549 U.S. 497,
                533-35 (2007).
                 \46\ 42 U.S.C. 7521(a)(2).
                ---------------------------------------------------------------------------
                B. NHTSA's Statutory Requirements
                 NHTSA is setting national Corporate Average Fuel Economy (CAFE)
                standards for passenger cars and light trucks for each model year as
                required under EPCA, as amended by EISA.\47\ EPCA mandates a motor
                vehicle fuel economy regulatory program that balances statutory factors
                in setting minimum fuel economy standards to facilitate energy
                conservation. EPCA allocates the responsibility for implementing the
                program between NHTSA and EPA as follows: NHTSA sets CAFE standards for
                passenger cars and light trucks; EPA establishes the procedures for
                testing, tests vehicles, collects and analyzes manufacturers' data, and
                calculates the individual and average fuel economy of each
                manufacturer's passenger cars and light trucks; and NHTSA enforces the
                standards based on EPA's calculations.
                ---------------------------------------------------------------------------
                 \47\ EPCA and EISA direct the Secretary of Transportation to
                develop, implement, and enforce fuel economy standards (see 49
                U.S.C. 32901 et. seq.), which authority the Secretary has delegated
                to NHTSA at 49 CFR 1.94(c).
                ---------------------------------------------------------------------------
                 The following sections enumerate specific statutory requirements
                for NHTSA in setting CAFE standards and NHTSA's interpretations of
                them, where applicable. Many comments were received on these
                requirements and interpretations. Because this is intended as an
                overview section, those comments will be addressed below in Section
                VIII rather than here, and the agencies refer readers to that part of
                the document for more information.
                 For each future model year, EPCA (as amended by EISA) requires that
                DOT (by delegation, NHTSA) establish separate passenger car and light
                truck standards at ``the maximum feasible average fuel economy level
                that the Secretary decides the manufacturers can achieve in that model
                year,'' \48\ based on the agency's consideration of four statutory
                factors: ``technological feasibility, economic practicability, the
                effect of other motor vehicle standards of the Government on fuel
                economy, and the need of the United States to conserve energy.'' \49\
                The law also allows NHTSA to amend standards that are already in place,
                as long as doing so meets these requirements.\50\ EPCA does not define
                these terms or specify what weight to give each concern in balancing
                them; thus, NHTSA defines them and determines the appropriate weighting
                that leads to the maximum feasible standards given the circumstances in
                each CAFE standard rulemaking.\51\
                ---------------------------------------------------------------------------
                 \48\ 49 U.S.C. 32902(a) and (b).
                 \49\ 49 U.S.C. 32902(f).
                 \50\ 49 U.S.C. 32902(g).
                 \51\ See Center for Biological Diversity v. NHTSA, 538 F.3d
                1172, 1195 (9th Cir. 2008) (hereafter ``CBD v. NHTSA'') (``The EPCA
                clearly requires the agency to consider these four factors, but it
                gives NHTSA discretion to decide how to balance the statutory
                factors--as long as NHTSA's balancing does not undermine the
                fundamental purpose of the EPCA: Energy conservation.'')
                ---------------------------------------------------------------------------
                 EISA added several other requirements to the setting of separate
                passenger car and light truck standards. Standards must be ``based on 1
                or more vehicle attributes related to fuel economy and express[ed] . .
                . in the form of a mathematical function.'' \52\ New standards must
                also be set at least 18 months before the model year in question, as
                would amendments to increase standards previously set.\53\ NHTSA must
                regulations prescribing average fuel economy standards for at least 1,
                but not more than 5, model years at a time.\54\ A number of comments
                addressed these requirements; for the reader's reference, those
                comments will be summarized and responded to in Section VIII. EISA also
                added the requirement that NHTSA set a minimum standard for
                domestically-manufactured passenger cars,\55\ which will also be
                discussed further in Section VIII below.
                ---------------------------------------------------------------------------
                 \52\ 49 U.S.C. 32902(b)(3)(A).
                 \53\ 49 U.S.C. 32902(a), (g)(2).
                 \54\ 49 U.S.C. 39202(b)(3)(B).
                 \55\ 49 U.S.C. 32902(b)(4).
                ---------------------------------------------------------------------------
                 For MYs 2011-2020, EISA further required that the separate
                standards for passenger cars and for light trucks be set at levels high
                enough to ensure that the achieved average fuel economy for the entire
                industry-wide combined fleet of new passenger cars and light trucks
                reach at least 35 mpg not later than MY 2020, and standards for those
                years were also required to ``increase ratably.'' \56\ For model years
                after 2020, standards must be set at the maximum feasible level.\57\
                ---------------------------------------------------------------------------
                 \56\ 49 U.S.C. 32902(b)(2)(A) and (C). NHTSA has CAFE standards
                in place that are projected to result in industry-achieved fuel
                economy levels over 35 mpg in MY 2020. EPA typically provides
                verified final CAFE data from manufacturers to NHTSA several months
                or longer after the close of the MY in question, so the actual MY
                2020 fuel economy will not be known until well after MY 2020 has
                ended. The standards for all MYs up to and including 2020 are known
                and not at issue in this regulatory action, so these provisions are
                noted for completeness rather than immediate relevance to this final
                rule. Because neither of these requirements apply after MY 2020,
                they are not relevant to this rulemaking and will not be discussed
                further.
                 \57\ 49 U.S.C. 32902(b)(2)(B).
                ---------------------------------------------------------------------------
                1. Factors That Must Be Considered in Deciding What Levels of CAFE
                Standards are ``Maximum Feasible''
                (a) Technological Feasibility
                 ``Technological feasibility'' refers to whether a particular method
                of improving fuel economy can be available for commercial application
                in the model year for which a standard is being established. Thus, in
                determining the level of new standards, the agency is not limited to
                technology that is already being commercially applied at the time of
                the rulemaking. For this rulemaking, NHTSA has evaluated and considered
                all types of technologies that improve real-world fuel economy,
                although not every possible technology was expressly included in the
                analysis, as discussed in Section VI and also in Section VIII.
                (b) Economic Practicability
                 ``Economic practicability'' refers to whether a standard is one
                ``within the
                [[Page 24214]]
                financial capability of the industry, but not so stringent as to'' lead
                to ``adverse economic consequences, such as a significant loss of jobs
                or the unreasonable elimination of consumer choice.'' \58\ The agency
                has explained in the past that this factor can be especially important
                during rulemakings in which the automobile industry is facing
                significantly adverse economic conditions (with corresponding risks to
                jobs). Economic practicability is a broad factor that includes
                considerations of the uncertainty surrounding future market conditions
                and consumer demand for fuel economy in addition to other vehicle
                attributes.\59\ In an attempt to evaluate the economic practicability
                of different future levels of CAFE standards (i.e., the regulatory
                alternatives considered in this rulemaking), NHTSA considers a variety
                of factors, including the annual rate at which manufacturers can
                increase the percentage of their fleet(s) that employ a particular type
                of fuel-saving technology, the specific fleet mixes of different
                manufacturers, assumptions about the cost of the standards to
                consumers, and consumers' valuation of fuel economy, among other
                things, including, in part, safety.
                ---------------------------------------------------------------------------
                 \58\ 67 FR 77015, 77021 (Dec. 16, 2002).
                 \59\ See, e.g., Center for Auto Safety v. NHTSA (``CAS''), 793
                F.2d 1322 (D.C. Cir. 1986) (Administrator's consideration of market
                demand as component of economic practicability found to be
                reasonable); Public Citizen v. NHTSA, 848 F.2d 256 (D.C. Cir. 1988)
                (Congress established broad guidelines in the fuel economy statute;
                agency's decision to set lower standard was a reasonable
                accommodation of conflicting policies).
                ---------------------------------------------------------------------------
                 It is important to note, however, that the law does not preclude a
                CAFE standard that poses considerable challenges to any individual
                manufacturer. The Conference Report for EPCA, as enacted in 1975, makes
                clear, and the case law affirms, ``a determination of maximum feasible
                average fuel economy should not be keyed to the single manufacturer
                which might have the most difficulty achieving a given level of average
                fuel economy.'' \60\ Instead, NHTSA is compelled ``to weigh the
                benefits to the nation of a higher fuel economy standard against the
                difficulties of individual automobile manufacturers.'' \61\
                Accordingly, while the law permits NHTSA to set CAFE standards that
                exceed the projected capability of a particular manufacturer as long as
                the standard is economically practicable for the industry as a whole,
                the agency cannot simply disregard that impact on individual
                manufacturers.\62\ That said, in setting fuel economy standards, NHTSA
                does not seek to maintain competitive positions among the industry
                players, and notes that while a particular CAFE standard may pose
                difficulties for one manufacturer as being too high or too low, it may
                also present opportunities for another. NHTSA has long held that the
                CAFE program is not necessarily intended to maintain the competitive
                positioning of each particular company. Rather, it is intended to
                enhance the fuel economy of the vehicle fleet on American roads, while
                protecting motor vehicle safety and paying close attention to the
                economic risks.
                ---------------------------------------------------------------------------
                 \60\ Center for Auto Safety v. NHTSA (``CAS''), 793 F.2d 1322,
                1352 (D.C. Cir. 1986).
                 \61\ Id.
                 \62\ Id. (``. . . the Secretary must weigh the benefits to the
                nation of a higher average fuel economy standard against the
                difficulties of individual automobile manufacturers.'')
                ---------------------------------------------------------------------------
                (c) The Effect of Other Motor Vehicle Standards of the Government on
                Fuel Economy
                 ``The effect of other motor vehicle standards of the Government on
                fuel economy'' involves an analysis of the effects of compliance with
                emission, safety, noise, or damageability standards on fuel economy
                capability and thus on average fuel economy. In many past CAFE
                rulemakings, NHTSA has said that it considers the adverse effects of
                other motor vehicle standards on fuel economy. It said so because, from
                the CAFE program's earliest years,\63\ the effects of such compliance
                on fuel economy capability over the history of the program have been
                negative ones. For example, safety standards that have the effect of
                increasing vehicle weight lower vehicle fuel economy capability and
                thus decrease the level of average fuel economy that the agency can
                determine to be feasible. NHTSA has considered the additional weight
                that it estimates would be added in response to new safety standards
                during the rulemaking timeframe. NHTSA has also accounted for EPA's
                ``Tier 3'' standards for criteria pollutants in its estimates of
                technology effectiveness.\64\
                ---------------------------------------------------------------------------
                 \63\ 42 FR 63184, 63188 (Dec. 15, 1977). See also 42 FR 33534,
                33537 (Jun. 30, 1977).
                 \64\ See Section VI, below.
                ---------------------------------------------------------------------------
                 The NPRM also discussed how EPA's CO2 standards for
                light-duty vehicles and California's Advanced Clean Cars program fit
                into NHTSA's consideration of ``the effect of other motor vehicle
                standards of the Government on fuel economy.'' The agencies note that
                on September 19, 2019, to ensure One National Program for automobile
                fuel economy and carbon dioxide emissions standards, the agencies
                finalized regulatory text related to preemption of State tailpipe
                CO2 standards and Zero Emission Vehicle (ZEV) mandates under
                EPCA and partial withdrawal of a waiver previously provided to
                California under the Clean Air Act.\65\ This final rule's impact on
                State programs--including California's--will therefore be somewhat
                different from the NPRM's consideration. In the interest of brevity,
                this preamble will hold further discussion of that point, along with
                responses to comments received, until Section VIII.
                ---------------------------------------------------------------------------
                 \65\ 84 FR 51310 (Sept. 27, 2019).
                ---------------------------------------------------------------------------
                (d) The Need of the United States To Conserve Energy
                 ``The need of the United States to conserve energy'' means ``the
                consumer cost, national balance of payments, environmental, and foreign
                policy implications of our need for large quantities of petroleum,
                especially imported petroleum.'' \66\ Environmental implications
                principally include changes in emissions of carbon dioxide and criteria
                pollutants and air toxics. Prime examples of foreign policy
                implications are energy independence and security concerns.
                ---------------------------------------------------------------------------
                 \66\ 42 FR 63184, 63188 (1977).
                ---------------------------------------------------------------------------
                (1) Consumer Costs and Fuel Prices
                 Fuel for vehicles costs money for vehicle owners and operators. All
                else equal (and this is an important qualification), consumers benefit
                from vehicles that need less fuel to perform the same amount of work.
                Future fuel prices are a critical input into the economic analysis of
                potential CAFE standards because they determine the value of fuel
                savings both to new vehicle buyers and to society, the amount of fuel
                economy that the new vehicle market is likely to demand in the absence
                of new standards, and they inform NHTSA about the consumer cost of the
                nation's need for large quantities of petroleum. In this final rule,
                NHTSA's analysis relies on fuel price projections estimated using the
                version of NEMS used for the U.S. Energy Information Administration's
                (EIA) Annual Energy Outlook for 2019.\67\ Federal government agencies
                generally use EIA's price projections in their assessment of future
                energy-related policies.
                ---------------------------------------------------------------------------
                 \67\ The analysis for the proposal relied on fuel price
                projections from AEO 2017; the difference in the projections is
                discussed in Section VI.
                ---------------------------------------------------------------------------
                (2) National Balance of Payments
                 Historically, the need of the United States to conserve energy has
                included consideration of the ``national balance of payments'' because
                of concerns that importing large amounts of oil created a
                [[Page 24215]]
                significant wealth transfer to oil-exporting countries and left the
                U.S. economically vulnerable.\68\ As recently as 2009, nearly half of
                the U.S. trade deficit was driven by petroleum,\69\ yet this concern
                has largely lain fallow in more recent CAFE actions, in part because
                other factors besides petroleum consumption have since played a bigger
                role in the U.S. trade deficit.\70\ Given significant recent increases
                in U.S. oil production and corresponding decreases in oil imports, this
                concern seems likely to remain fallow for the foreseeable future.\71\
                Increasingly, changes in the price of fuel have come to represent
                transfers between domestic consumers of fuel and domestic producers of
                petroleum rather than gains or losses to foreign entities.
                ---------------------------------------------------------------------------
                 \68\ See, e.g., 42 FR 63184, 63192 (Dec. 15, 1977) (``A major
                reason for this need [to reduce petroleum consumption] is that the
                importation of large quantities of petroleum creates serious balance
                of payments and foreign policy problems. The United States currently
                spends approximately $45 billion annually for imported petroleum.
                But for this large expenditure, the current large U.S. trade deficit
                would be a surplus.'')
                 \69\ See ``Today in Energy: Recent improvements in petroleum
                trade balance mitigate U.S. trade deficit,'' U.S. Energy Information
                Administration (Jul. 21, 2014), available at https://www.eia.gov/todayinenergy/detail.php?id=17191.
                 \70\ See, e.g., Nida [Ccedil]akir Melek and Jun Nie, ``What
                Could Resurging U.S. Energy Production Mean for the U.S. Trade
                Deficit,'' Mar. 7, 2018, Federal Reserve Bank of Kansas City.
                Available at https://www.kansascityfed.org/publications/research/mb/articles/2018/what-could-resurging-energy-production-mean. The
                authors state that ``The decline in U.S. net energy imports has
                prevented the total U.S. trade deficit from widening further. . . .
                In 2006, petroleum accounted for about 16 percent of U.S. goods
                imports and about 3 percent of U.S. goods exports. By the end of
                2017, the share of petroleum in total goods imports declined to 8
                percent, while the share in total goods exports almost tripled,
                shrinking the U.S. petroleum trade deficit. Had the petroleum trade
                deficit not improved, all else unchanged, the total U.S. trade
                deficit would likely have been more than 35 percent wider by the end
                of 2017.''
                 \71\ For an illustration of recent increases in U.S. production,
                see, e.g., `U.S. crude oil and liquid fuels production,'' Short-Term
                Energy Outlook, U.S. Energy Information Administration (Aug. 2019),
                available at http://www.eia.gov/outlooks/steo/images/Fig16.png. EIA
                noted in April 2019 that ``Annual U.S. crude oil production reached
                a record level of 10.96 million barrels per day (b/d) in 2018, 1.6
                million b/d (17%) higher than 2017 levels. In December 2018, monthly
                U.S. crude oil production reached 11.96 million b/d, the highest
                monthly level of crude oil production in U.S. history. U.S crude oil
                production has increased significantly over the past 10 years,
                driven mainly by production from tight rock formations using
                horizontal drilling and hydraulic fracturing. EIA projects that U.S.
                crude oil production will continue to grow in 2019 and 2020,
                averaging 12.3 million b/d and 13.0 million b/d, respectively.''
                ``Today in Energy: U.S. crude oil production grew 17% in 2018,
                surpassing the previous record in 1970,'' EIA, Apr. 9, 2019.
                Available at http://www.eia.gov/todayinenergy/detail.php?id=38992.
                ---------------------------------------------------------------------------
                 As flagged in the NPRM, some commenters raised concerns about
                potential economic consequences for automaker and supplier operations
                in the U.S. due to disparities between CAFE standards at home and their
                counterpart fuel economy/efficiency and CO2 standards
                abroad. NHTSA finds these concerns more relevant to technological
                feasibility and economic practicability considerations than to the
                national balance of payments. The discussion in Section VIII below
                addresses this topic in more detail.
                (3) Environmental Implications
                 Higher fleet fuel economy can reduce U.S. emissions of various
                pollutants by reducing the amount of oil that is produced and refined
                for the U.S. vehicle fleet, but can also increase emissions by reducing
                the cost of driving, which can result in more vehicle miles traveled
                (i.e., the rebound effect). Thus, the net effect of more stringent CAFE
                standards on emissions of each pollutant depends on the relative
                magnitude of both its reduced emissions in fuel refining and
                distribution and increases in its emissions from vehicle use. Fuel
                savings from CAFE standards also necessarily results in lower emissions
                of CO2, the main greenhouse gas emitted as a result of
                refining, distributing, and using transportation fuels. Reducing fuel
                consumption directly reduces CO2 emissions because the
                primary source of transportation-related CO2 emissions is
                fuel combustion in internal combustion engines.
                 NHTSA has considered environmental issues, both within the context
                of EPCA and the context of the National Environmental Policy Act
                (NEPA), in making decisions about the setting of standards since the
                earliest days of the CAFE program. As courts of appeal have noted in
                three decisions stretching over the last 20 years,\72\ NHTSA defined
                ``the need of the United States to conserve energy'' in the late 1970s
                as including, among other things, environmental implications. In 1988,
                NHTSA included climate change concepts in its CAFE notices and prepared
                its first environmental assessment addressing that subject.\73\ It
                cited concerns about climate change as one of its reasons for limiting
                the extent of its reduction of the CAFE standard for MY 1989 passenger
                cars.\74\ Since then, NHTSA has considered the effects of reducing
                tailpipe emissions of CO2 in its fuel economy rulemakings
                pursuant to the need of the United States to conserve energy by
                reducing petroleum consumption.
                ---------------------------------------------------------------------------
                 \72\ CAS, 793 F.2d 1322, 1325 n. 12 (D.C. Cir. 1986); Public
                Citizen, 848 F.2d 256, 262-63 n. 27 (D.C. Cir 1988) (noting that
                ``NHTSA itself has interpreted the factors it must consider in
                setting CAFE standards as including environmental effects''); CBD,
                538 F.3d 1172 (9th Cir. 2007).
                 \73\ 53 FR 33080, 33096 (Aug. 29, 1988).
                 \74\ 53 FR 39275, 39302 (Oct. 6, 1988).
                ---------------------------------------------------------------------------
                (4) Foreign Policy Implications
                 U.S. consumption and imports of petroleum products can impose
                additional costs (i.e., externalities) on the domestic economy that are
                not reflected in the market price for crude petroleum or in the prices
                paid by consumers for petroleum products such as gasoline. NHTSA has
                said previously that these costs can include (1) higher prices for
                petroleum products resulting from the effect of U.S. oil demand on
                world oil prices, (2) the risk of disruptions to the U.S. economy
                caused by sudden increases in the global price of oil and its resulting
                impact on fuel prices faced by U.S. consumers, and (3) expenses for
                maintaining the strategic petroleum reserve (SPR) to provide a response
                option should a disruption in commercial oil supplies threaten the U.S.
                economy, to allow the U.S. to meet part of its International Energy
                Agency obligation to maintain emergency oil stocks, and to provide a
                national defense fuel reserve.\75\ Higher U.S. consumption of crude oil
                or refined petroleum products increases the magnitude of these external
                economic costs, thus increasing the true economic cost of supplying
                transportation fuels above the resource costs of producing them.
                Conversely, reducing U.S. consumption of crude oil or refined petroleum
                products (by reducing motor fuel use) can reduce these external costs.
                ---------------------------------------------------------------------------
                 \75\ While the U.S. maintains a military presence in certain
                parts of the world to help secure global access to petroleum
                supplies, that is neither the primary nor the sole mission of U.S.
                forces overseas. Additionally, the scale of oil consumption
                reductions associated with CAFE standards would be insufficient to
                alter any existing military missions focused on ensuring the safe
                and expedient production and transportation of oil around the globe.
                See the FRIA's discussion on energy security for more information on
                this topic.
                ---------------------------------------------------------------------------
                 While these costs are considerations, the United States has
                significantly increased oil production capabilities in recent years, to
                the extent that the U.S. is currently producing enough oil to satisfy
                nearly all of its energy needs and is projected to continue to do so
                (or even become a net energy exporter in the near future).\76\ This has
                added stable new supply to the global oil market, which ameliorates the
                U.S.' need to
                [[Page 24216]]
                conserve energy from a security perspective even given that oil is a
                global commodity. The agencies discuss this issue in more detail in
                Section VIII below.
                ---------------------------------------------------------------------------
                 \76\ See AEO 2019, at 14 (``In the Reference case, the United
                States becomes a net exporter of petroleum liquids after 2020 as
                U.S. crude oil production increases and domestic consumption of
                petroleum products decreases.''). Available at https://www.eia.gov/outlooks/aeo/pdf/aeo2019.pdf.
                ---------------------------------------------------------------------------
                (2) Factors That NHTSA Is Prohibited From Considering
                 EPCA states that in determining the level at which it should set
                CAFE standards for a particular model year, NHTSA may not consider the
                ability of manufacturers to take advantage of several EPCA provisions
                that facilitate compliance with CAFE standards and thereby can reduce
                their costs of compliance.\77\ As discussed further below, NHTSA cannot
                consider compliance credits that manufacturers earn by exceeding the
                CAFE standards and then use to achieve compliance in years in which
                their measured average fuel economy falls below the standards. NHTSA
                also cannot consider the use of alternative fuels by dual-fueled
                vehicles (such as plug-in hybrid electric vehicles) nor the
                availability of dedicated alternative fuel vehicles (such as battery
                electric or hydrogen fuel cell vehicles) in any model year. EPCA
                encourages the production of alternative fuel vehicles by specifying
                that their fuel economy is to be determined using a special calculation
                procedure that results in those vehicles being assigned a higher fuel
                economy level than they actually achieve. For non-statutory incentives
                that NHTSA developed by regulation, NHTSA does not consider these
                incentives subject to the EPCA prohibition on considering
                flexibilities. These topics will be addressed further in Section VIII
                below.
                ---------------------------------------------------------------------------
                 \77\ 49 U.S.C. 32902(h).
                ---------------------------------------------------------------------------
                (3) Other Considerations in Determining Maximum Feasible CAFE Standards
                 NHTSA historically has interpreted EPCA's statutory factors as
                including consideration for potential adverse safety consequences in
                setting CAFE standards. Courts have consistently recognized that this
                interpretation is reasonable. As courts have recognized, ``NHTSA has
                always examined the safety consequences of the CAFE standards in its
                overall consideration of relevant factors since its earliest rulemaking
                under the CAFE program.'' \78\ The courts have consistently upheld
                NHTSA's implementation of EPCA in this manner.\79\ Thus, in evaluating
                what levels of stringency would result in maximum feasible standards,
                NHTSA assesses the potential safety impacts and considers them in
                balancing the statutory considerations and to determine the maximum
                feasible level of the standards.\80\ Many commenters addressed the
                NPRM's analysis of safety impacts; those comments will be summarized
                and responded to in Section VI.D.2 and also in each agency's discussion
                in Section VIII.
                ---------------------------------------------------------------------------
                 \78\ Competitive Enterprise Institute v. NHTSA, 901 F.2d 107,
                120 n. 11 (D.C. Cir. 1990) (``CEI-I'') (citing 42 FR 33534, 33551
                (Jun. 30, 1977).
                 \79\ See, e.g., Competitive Enterprise Institute v. NHTSA, 956
                F.2d 321, 322 (D.C. Cir. 1992) (``CEI-II'') (in determining the
                maximum feasible fuel economy standard, ``NHTSA has always taken
                passenger safety into account,'' citing CEI-I, 901 F.2d at 120 n.
                11); Competitive Enterprise Institute v. NHTSA, 49 F.3d 481, 483-83
                (D.C. Cir. 1995) (same); Center for Biological Diversity v. NHTSA,
                538 F.3d 1172, 1203-04 (9th Cir. 2008) (upholding NHTSA's analysis
                of vehicle safety issues with weight in connection with the MYs
                2008-2011 light truck CAFE rulemaking).
                 \80\ NHTSA stated in the NPRM that ``While we discuss safety as
                a separate consideration, NHTSA also considers safety as closely
                related to, and in some circumstances a subcomponent of, economic
                practicability. On a broad level, manufacturers have finite
                resources to invest in research and development. Investment into the
                development and implementation of fuel saving technology necessarily
                comes at the expense of investing in other areas such as safety
                technology. On a more direct level, when making decisions on how to
                equip vehicles, manufacturers must balance cost considerations to
                avoid pricing further consumers out of the market. As manufacturers
                add technology to increase fuel efficiency, they may decide against
                installing new safety equipment to reduce cost increases. And as the
                price of vehicles increase beyond the reach of more consumers, such
                consumers continue to drive or purchase older, less safe vehicles.
                In assessing practicability, NHTSA also considers the harm to the
                nation's economy caused by highway fatalities and injuries.'' 83 FR
                at 43209 (Aug. 24, 2018). Many comments were received on this issue,
                which will be discussed further in Section VIII below.
                ---------------------------------------------------------------------------
                 The above sections explain what Congress thought was important
                enough to codify when it directed each agency to regulate, and begin to
                explain how the agencies have interpreted those directions over time
                and in this final rule. The next section looks more closely at the
                interplay between Congress's direction to the agencies and the aspects
                of the market that these regulations affect, as follows.
                IV. Purpose of Analytical Approach Considered as Part of Decision-
                Making
                A. Relationship of Analytical Approach to Governing Law
                 Like the NPRM, today's final rule is supported by extensive
                analysis of potential impacts of the regulatory alternatives under
                consideration. Below, Section VI reviews the analytical approach,
                Section VII summarizes the results of the analysis, and Section VIII
                explains how the final standards--informed by this analysis--fulfill
                the agencies' statutory obligations. Accompanying today's notice, a
                final Regulatory Impact Analysis (FRIA) and, for NHTSA's consideration,
                a final Environmental Impact Analysis (FEIS), together provide a more
                extensive and detailed enumeration of related methods, estimates,
                assumptions, and results. The agencies' analysis has been constructed
                specifically to reflect various aspects of governing law applicable to
                CAFE and CO2 standards, and has been expanded and improved
                in response to comments received to the NPRM and based on additional
                work by the agencies. The analysis aided the agencies in implementing
                their statutory obligations, including the weighing of competing
                considerations, by reasonably informing the agencies about the
                estimated effects of choosing different regulatory alternatives.
                 The agencies' analysis makes use of a range of data (i.e.,
                observations of things that have occurred), estimates (i.e., things
                that may occur in the future), and models (i.e., methods for making
                estimates). Two examples of data include (1) records of actual odometer
                readings used to estimate annual mileage accumulation at different
                vehicle ages and (2) CAFE compliance data used as the foundation for
                the ``analysis fleet'' containing, among other things, production
                volumes and fuel economy levels of specific configurations of specific
                vehicle models produced for sale in the U.S. Two examples of estimates
                include (1) forecasts of future GDP growth used, with other estimates,
                to forecast future vehicle sales volumes and (2) the ``retail price
                equivalent'' (RPE) factor used to estimate the ultimate cost to
                consumers of a given fuel-saving technology, given accompanying
                estimates of the technology's ``direct cost,'' as adjusted to account
                for estimated ``cost learning effects'' (i.e., the tendency that it
                will cost a manufacturer less to apply a technology as the manufacturer
                gains more experience doing so).
                 The agencies' analysis makes use of several models, some of which
                are actually integrated systems of multiple models. As discussed in the
                NPRM, the agencies' analysis of CAFE and CO2 standards
                involves two basic elements: First, estimating ways each manufacturer
                could potentially respond to a given set of standards in a manner that
                considers potential consumer response; and second, estimating various
                impacts of those responses. Estimating manufacturers' potential
                responses involves simulating manufacturers' decision-making processes
                regarding the year-by-year application of fuel-saving technologies to
                specific vehicles. Estimating impacts involves calculating resultant
                changes in new vehicle costs, estimating a
                [[Page 24217]]
                variety of costs (e.g., for fuel) and effects (e.g., CO2
                emissions from fuel combustion) occurring as vehicles are driven over
                their lifetimes before eventually being scrapped, and estimating the
                monetary value of these effects. Estimating impacts also involves
                consideration of the response of consumers--e.g., whether consumers
                will purchase the vehicles and in what quantities. Both of these basic
                analytical elements involve the application of many analytical inputs.
                 The agencies' analysis uses the CAFE Model to estimate
                manufacturers' potential responses to new CAFE and CO2
                standards and to estimate various impacts of those responses. The model
                may be characterized as an integrated system of models. For example,
                one model estimates manufacturers' responses, another estimates
                resultant changes in total vehicle sales, and still another estimates
                resultant changes in fleet turnover (i.e., scrappage). The CAFE model
                makes use of many inputs, values of which are developed outside of the
                model and not by the model. For example, the model applies fuel prices;
                it does not estimate fuel prices. The model does not determine the form
                or stringency of the standards; instead, the model applies inputs
                specifying the form and stringency of standards to be analyzed and
                produces outputs showing effects of manufacturers working to meet those
                standards, which become the basis for comparing between different
                potential stringencies.
                 The agencies also use EPA's MOVES model to estimate ``tailpipe''
                (a.k.a. ``vehicle'' or ``downstream'') emission factors for criteria
                pollutants,\81\ and use four DOE and DOE-sponsored models to develop
                inputs to the CAFE model, including three developed and maintained by
                DOE's Argonne National Laboratory. The agencies use the DOE Energy
                Information Administration's (EIA's) National Energy Modeling System
                (NEMS) to estimate fuel prices,\82\ and use Argonne's Greenhouse gases,
                Regulated Emissions, and Energy use in Transportation (GREET) model to
                estimate emissions rates from fuel production and distribution
                processes.\83\ DOT also sponsored DOE/Argonne to use Argonne's
                Autonomie full-vehicle modeling and simulation system to estimate the
                fuel economy impacts for roughly a million combinations of technologies
                and vehicle types.84 85 Section VI.B.3, below, and the
                accompanying final RIA document details of the agencies' use of these
                models. In addition, as discussed in the final EIS accompanying today's
                notice, DOT relied on a range of climate and photochemical models to
                estimate impacts on climate, air quality, and public health. The EIS
                discusses and documents the use of these models.
                ---------------------------------------------------------------------------
                 \81\ See https://www.epa.gov/moves. Today's final rule used
                version MOVES2014b, available at https://www.epa.gov/moves/latest-version-motor-vehicle-emission-simulator-moves.
                 \82\ See https://www.eia.gov/outlooks/aeo/info_nems_archive.php.
                Today's final rule uses fuel prices estimated using the Annual
                Energy Outlook (AEO) 2019 version of NEMS (see https://www.eia.gov/outlooks/aeo/data/browser/#/?id=3-AEO2019&cases=ref2019&sourcekey=0).
                 \83\ Information regarding GREET is available at https://greet.es.anl.gov/index.php. Today's notice uses the 2018 version of
                GREET.
                 \84\ As part of the Argonne simulation effort, individual
                technology combinations simulated in Autonomie were paired with
                Argonne's BatPAC model to estimate the battery cost associated with
                each technology combination based on characteristics of the
                simulated vehicle and its level of electrification. Information
                regarding Argonne's BatPAC model is available at http://www.cse.anl.gov/batpac/.
                 \85\ In addition, the impact of engine technologies on fuel
                consumption, torque, and other metrics was characterized using GT
                POWER simulation modeling in combination with other engine modeling
                that was conducted by IAV Automotive Engineering, Inc. (IAV). The
                engine characterization ``maps'' resulting from this analysis were
                used as inputs for the Autonomie full-vehicle simulation modeling.
                Information regarding GT Power is available at https://www.gtisoft.com/gt-suite-applications/propulsion-systems/gt-power-engine-simulation-software.
                ---------------------------------------------------------------------------
                 As further explained in the NPRM,\86\ to prepare for analysis
                supporting the proposal, DOT expanded the CAFE model to address EPA
                statutory and regulatory requirements through a year-by-year simulation
                of how manufacturers could comply with EPA's CO2 standards,
                including:
                ---------------------------------------------------------------------------
                 \86\ 83 FR 42986, 43003 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 Calculation of vehicle models' CO2 emission
                rates before and after application of fuel-saving (and, therefore,
                CO2-reducing) technologies;
                 Calculation of manufacturers' fleet average CO2
                emission rates;
                 Calculation of manufacturers' fleet average CO2
                emission rates under attribute-based CO2 standards;
                 Accounting for adjustments to average CO2
                emission rates reflecting reduction of air conditioner refrigerant
                leakage;
                 Accounting for the treatment of alternative fuel vehicles
                for CO2 compliance;
                 Accounting for production ``multipliers'' for PHEVs, BEVs,
                compressed natural gas (CNG) vehicles, and fuel cell vehicles (FCVs);
                 Accounting for transfer of CO2 credits between
                regulated fleets; and
                 Accounting for carried-forward (a.k.a. ``banked'')
                CO2 credits, including credits from model years earlier than
                modeled explicitly.
                 As further discussed in the NPRM, although EPA had previously
                developed a vehicle simulation tool (``ALPHA'') and a fleet compliance
                model (``OMEGA''), and had applied these in prior actions, having
                considered the facts before the Agency in 2018, EPA determined that,
                ``it is reasonable and appropriate to use DOE/Argonne's model for full-
                vehicle simulation, and to use DOT's CAFE model for analysis of
                regulatory alternatives.'' \87\
                ---------------------------------------------------------------------------
                 \87\ 83 FR 42986, 43000 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 As discussed below and in Section VI.B.3, some commenters--some
                citing deliberative EPA staff communications during NPRM development,
                and one submitting comments by a former EPA staff member closely
                involved in the origination of the above-mentioned OMEGA model--took
                strong exception to EPA's decision to rely on DOE/Argonne and DOT-
                originated models as the basis for analysis informing EPA's decisions
                regarding CO2 standards. Some commenters argued that the EPA
                Administrator must consider exclusively models and analysis originating
                with EPA staff, and that to do otherwise would be arbitrary and
                capricious. As explained below (and as explained in the NPRM), it is
                reasonable for the Administrator to consider analysis and information
                produced from many sources, including, in this instance, the DOE/
                Argonne and DOT models. The Administrator has the discretion to
                determine what information reasonably and appropriately informs
                decisions regarding emissions standards. Some commenters conflated
                models with decisions, suggesting that the former mechanically
                determine the latter. The CAA authorizes the EPA Administrator, not a
                model, to make decisions about emissions standards, just as EPCA
                provides similar authority to the Secretary. Models produce analysis,
                the results of which help to inform decisions. However, in making such
                decisions, the Administrator may and should consider other relevant
                information beyond the outputs of any models--including public
                comment--and, in all cases, must exercise judgment in establishing
                appropriate standards.
                 Some commenters conflated models with inputs and/or with results of
                the modeling. All of the models mentioned above rely on inputs,
                including not only data (i.e., facts), but also estimates (inputs about
                the future are estimates, not data). Given these inputs, the models
                produce estimates--ultimately, the agencies' reported estimates of the
                potential impacts of standards under
                [[Page 24218]]
                consideration. In other words, inputs do not define models; models use
                inputs. Therefore, disagreements about inputs do not logically extend
                to disagreements about models. Similarly, while models determine
                resulting outputs, they do so based on inputs. Therefore, disagreements
                about results do not necessarily imply disagreements about models; they
                may merely reflect disagreements about inputs. With respect to the
                Administrator's decisions regarding models underlying today's analysis,
                comments regarding inputs, therefore, are more appropriately addressed
                separately, which is done so below in Section VI.
                 The EPA Administrator's decision to continue relying on the DOE/
                Argonne Autonomie tool and DOT CAFE model rather than on the
                corresponding tools developed by EPA staff is informed by consideration
                of comments on results and on technical aspects of the models
                themselves. As discussed below, some commenters questioned specific
                aspects of the CAFE model's simulation of manufacturer's potential
                responses to CO2 standards. Considering these comments, the
                CAFE model applied in the final rule's analysis includes some revisions
                and updates. For example, the ``effective cost'' metric used to select
                among available opportunities to apply fuel-saving technologies now
                uses a ``cost per credit'' metric rather than the metric used for the
                NPRM. Also, the model's representation of sales ``multipliers'' EPA has
                included for CNG vehicles, PHEVs, BEVs, and FCVs reflects current EPA
                regulations or, as an input-selectable option, an alternative approach
                under consideration. On the other hand, some commenters questioning the
                CAFE model's approach to some CO2 program features appear to
                ignore the fact that prior analysis by EPA (using EPA's OMEGA) model
                likewise did not account for the same program features. For example,
                some stakeholders took issue with the CAFE model's approach to
                accounting for banked CO2 credits and, in particular,
                credits banked prior to the model years accounted for explicitly in the
                analysis. In the course of updating the basis for analysis fleet from
                model year 2016 to model year 2017, the agencies have since updated
                corresponding inputs. However, even though the ability to carry forward
                credits impacts outcomes, EPA's OMEGA model used in previous
                rulemakings never attempted to account for credit banking and, indeed,
                lacking a year-by-year structure, cannot account for credit banking.
                Therefore, at least with respect to this important CO2
                program flexibility, the CAFE model provides a more complete and
                realistic basis for estimating actual impacts of new CO2
                standards.
                 For its part, NHTSA remains confident that the combination of the
                Autonomie and CAFE models remains the best available for CAFE
                rulemaking analysis, and notes, as discussed below, that even the
                environmental group coalition stated that the CAFE model is aligned
                with EPCA requirements.\88\ In late 2001, after Congress discontinued
                an extended series of budget ``riders'' prohibiting work on CAFE
                standards, NHTSA and the DOT Volpe Center began development of a
                modeling system appropriate for CAFE rulemaking analysis, because other
                available models were not designed with this purpose in mind, and
                lacked capabilities important for CAFE rulemakings. For example,
                although NEMS had procedures to account for CAFE standards, those
                procedures did not provide the ability to account for specific
                manufacturers, as is especially relevant to the statutory requirement
                that NHTSA consider the economic practicability of any new CAFE
                standards. Also, as early as the first rulemaking making use of this
                early CAFE model, commenters stressed the importance of product
                redesign schedules, leading developers to introduce procedures to
                account for product cadence. In the 2003 notice regarding light truck
                standards for MYs 2005-2007, NHTSA stated that ``we also changed the
                methodology to recognize that capital costs require employment of
                technologies for several years, rather than a single year. . . . In our
                view, this makes the Volpe analysis more consistent with the [manually
                implemented] Stage analysis and better reflects actual conditions in
                the automotive industry.'' \89\ Since that time, NHTSA and the Volpe
                Center have significantly refined the CAFE model with each of
                rulemaking. For example, for the 2006 rulemaking regarding standards
                for MYs 2008-2011 light trucks, NHTSA introduced the ability to account
                for attribute-based standards, account for the social cost of
                CO2 emissions, estimate stringencies at which net benefits
                would be maximized, and perform probabilistic uncertainty analysis
                (i.e., Monte Carlo simulation).\90\ For the 2009 rulemaking regarding
                standards for MY 2011 passenger cars and light trucks, we introduced
                the ability to account for attribute-based passenger car standards, and
                the ability to apply ``synergy factors'' to estimate how some
                technology pairings impact fuel consumption,\91\ For the 2010
                rulemaking regarding standards for MYs 2012-2016, we introduced
                procedures to account for FFV credits, and to account for product
                planning as a multiyear consideration.\92\ For the 2012 rulemaking
                regarding standards for MYs 2017-2025, we introduced several new
                procedures, such as (1) accounting for electricity used to charge
                electric vehicles (EVs) and plug-in hybrid electric vehicles (PHEVs),
                (2) accounting for use of ethanol blends in flexible-fuel vehicles
                (FFVs), (3) accounting for costs (i.e., ``stranded capital'') related
                to early replacement of technologies, (4) accounting for previously-
                applied technology when determining the extent to which a manufacturer
                could expand use of the technology, (5) applying technology-specific
                estimates of changes in consumer value, (6) simulating the extent to
                which manufacturers might utilize EPCA's provisions regarding
                generation and use of CAFE credits, (7) applying estimates of fuel
                economy adjustments (and accompanying costs) reflecting increases in
                air conditioner efficiency, (8) reporting privately-valued benefits,
                (9) simulating the extent to which manufacturers might voluntarily
                apply technology beyond levels needed for compliance with CAFE
                standards, and (10) estimating changes in highway fatalities
                attributable to any applied reductions in vehicle mass.\93\ Also for
                the 2012 rulemaking, we began making use of Autonomie to estimate fuel
                consumption impacts of different combinations of technologies, using
                these estimates to specify inputs to the CAFE model.\94\ In 2016,
                providing analyses for both the draft TAR regarding light-duty CAFE
                standards and the final rule regarding fuel consumption standards for
                heavy-duty pickup trucks and vans, we greatly expanded the agency's use
                of Autonomie-based full vehicle simulations and introduced the ability
                to simulate compliance with attribute-based standards for heavy-duty
                pickups and vans.\95\ And, as discussed at length in the NPRM and
                below, for this rulemaking, we have, among other things, refined
                procedures to account for impacts on highway travel and safety,
                [[Page 24219]]
                added procedures to simulate compliance with CO2 standards,
                refined procedures to account for compliance credits, and added
                procedures to account for impacts on sales, scrappage, and employment.
                We have also significantly revised the model's graphical user interface
                (GUI) in order to make the model easier to operate and understand. Like
                any model, both Autonomie and the CAFE model benefit from ongoing
                refinement. However, NHTSA is confident that this combination of models
                produces a more realistic characterization of the potential impacts of
                new standards than would another combination of available models. Some
                stakeholders, while commenting on specific aspects of the inputs,
                models, and/or results, commended the agencies' exclusive reliance on
                the DOE/Argonne Autonomie tool and DOT CAFE model. With respect to
                CO2 standards, these stakeholders noted not only technical
                reasons to use these models rather than the EPA models, but also other
                reasons such as efficiency, transparency, and ease with which outside
                parties can exercise models and replicate the agencies' analysis. These
                comments are discussed below and in Section VI.
                ---------------------------------------------------------------------------
                 \88\ Environmental group coalition, NHTSA-2018-0067-12000,
                Appendix A, at 24-25.
                 \89\ 68 FR at 16885 (Apr. 7, 2003).
                 \90\ 71 FR at 17566 et seq. (Apr. 6, 2006).
                 \91\ 74 FR at 14196 et seq. (Mar. 30, 3009).
                 \92\ 75 FR at 25599 et seq. (May 7, 2010).
                 \93\ 77 FR 63009 et seq. (Oct. 15, 2012).
                 \94\ 77 FR at 62712 et seq. (Oct. 15, 2012).
                 \95\ 81 FR at 73743 et seq. (Oct. 25, 2016); Draft TAR,
                available at Docket No. NHTSA-2016-0068-0001, Chapter 13.
                ---------------------------------------------------------------------------
                 Nevertheless, some comments regarding the model's handling of CAFE
                and/or CO2 standards, and some comments regarding the
                model's estimation of resultant impacts, led the agencies to make
                changes to specific aspects of the model. Comments on and changes to
                the inputs and model are discussed below and in Section VI; results are
                discussed in Section VII and in the accompanying RIA; and the meaning
                of results in the context of the applicable statutory requirements is
                discussed in Section VIII.
                 As explained, the analysis is designed to reflect a number of
                statutory and regulatory requirements applicable to CAFE and tailpipe
                CO2 standard setting. EPCA contains a number of requirements
                governing the scope and nature of CAFE standard setting. Among these,
                some have been in place since EPCA was first signed into law in 1975,
                and some were added in 2007, when Congress passed EISA and amended
                EPCA. The CAA, as discussed elsewhere, provides EPA with very broad
                authority under Section 202(a), and does not contain EPCA/EISA's
                prescriptions. In the interest of harmonization, however, EPA has
                adopted some of the EPCA/EISA requirements into its tailpipe
                CO2 regulations, and NHTSA, in turn, has created some
                additional flexibilities by regulation not expressly envisioned by
                EPCA/EISA in order to harmonize better with some of EPA's programmatic
                decisions. EPCA/EISA requirements regarding the technical
                characteristics of CAFE standards and the analysis thereof include, but
                are not limited to, the following, and the analysis reflects these
                requirements as summarized:
                 Corporate Average Standards: 49 U.S.C. 32902 requires standards
                that apply to the average fuel economy levels achieved by each
                corporation's fleets of vehicles produced for sale in the U.S.\96\ CAA
                Section 202(a) does not preclude the EPA Administrator from expressing
                CO2 standards as de facto fleet average requirements, and
                EPA has adopted a similar approach in the interest of harmonization.
                The CAFE Model, used by the agencies to conduct the bulk of today's
                analysis, calculates the CAFE and CO2 levels of each
                manufacturer's fleets based on estimated production volumes and
                characteristics, including fuel economy levels, of distinct vehicle
                models that could be produced for sale in the U.S.
                ---------------------------------------------------------------------------
                 \96\ This differs from safety standards and traditional
                emissions standards, which apply separately to each vehicle. For
                example, every vehicle produced for sale in the U.S. must, on its
                own, meet all applicable federal motor vehicle safety standards
                (FMVSS), but no vehicle produced for sale must, on its own, federal
                fuel economy standards. Rather, each manufacturer is required to
                produce a mix of vehicles that, taken together, achieve an average
                fuel economy level no less than the applicable minimum level.
                ---------------------------------------------------------------------------
                 Separate Standards for Passenger Cars and Light Trucks: 49 U.S.C.
                32902 requires the Secretary of Transportation to set CAFE standards
                separately for passenger cars and light trucks. CAA Section 202(a) does
                not preclude the EPA Administrator from specifying CO2
                standards separately for passenger cars and light trucks, and EPA has
                adopted a similar approach. The CAFE Model accounts separately for
                passenger cars and light trucks, including differentiated standards and
                compliance.
                 Attribute-Based Standards: 49 U.S.C. 32902 requires the Secretary
                of Transportation to define CAFE standards as mathematical functions
                expressed in terms of one or more vehicle attributes related to fuel
                economy. This means that for a given manufacturer's fleet of vehicles
                produced for sale in the U.S. in a given regulatory class and model
                year, the applicable minimum CAFE requirement (i.e., the numerical
                value of the requirement) is computed based on the applicable
                mathematical function, and the mix and attributes of vehicles in the
                manufacturer's fleet. In the 2012 final rule that first established
                CO2 standards, EPA also adopted an attribute-based standard
                under its broad CAA Section 202(a) authority. The CAFE Model accounts
                for such functions and vehicle attributes explicitly.
                 Separately Defined Standards for Each Model Year: 49 U.S.C. 32902
                requires the Secretary to set CAFE standards (separately for passenger
                cars and light trucks) at the maximum feasible levels in each model
                year. CAA Section 202(a) allows EPA to establish CO2
                standards separately for each model year, and EPA has chosen to do so
                for this final rule, similar to the approach taken in the previous
                light-duty vehicle CO2 standard-setting rules. The CAFE
                Model represents each model year explicitly, and accounts for the
                production relationships between model years.\97\
                ---------------------------------------------------------------------------
                 \97\ For example, a new engine first applied to given vehicle
                model/configuration in model year 2020 will most likely be ``carried
                forward'' to model year 2021 of that same vehicle model/
                configuration, in order to reflect the fact that manufacturers do
                not apply brand-new engines to a given vehicle model every single
                year.
                ---------------------------------------------------------------------------
                 Separate Compliance for Domestic and Imported Passenger Car Fleets:
                49 U.S.C. 32904 requires the EPA Administrator to determine CAFE
                compliance separately for each manufacturers' fleets of domestic
                passenger cars and imported passenger cars, which manufacturers must
                consider as they decide how to improve the fuel economy of their
                passenger car fleets. CAA 202(a) does not preclude the EPA
                Administrator from determining compliance with CO2 standards
                separately for a manufacturer's domestic and imported car fleets, but
                EPA did not include such a distinction in either the 2010 or 2012 final
                rules, and EPA did not propose or ask for comment on taking such an
                approach in the proposal. The CAFE Model is able to account explicitly
                for this requirement when simulating manufacturers' potential responses
                to CAFE standards, but combines any given manufacturer's domestic and
                imported cars into a single fleet when simulating that manufacturer's
                potential response to CO2 standards.
                 Minimum CAFE Standards for Domestic Passenger Car Fleets: 49 U.S.C.
                32902 requires that domestic passenger car fleets achieve CAFE levels
                no less than 92 percent of the industry-wide average level required
                under the applicable attribute-based CAFE standard, as projected by the
                Secretary at the time the standard is promulgated. CAA 202(a) does not
                preclude the EPA Administrator from correspondingly requiring that
                domestic passenger car fleets achieve CO2 levels no greater
                than 108.7 percent (1/0.92 = 1.087) of the projected industry-wide
                average CO2
                [[Page 24220]]
                requirement under the attribute-based standard, but the GHG program
                that EPA designed in the 2010 and 2012 final rules did not include such
                a distinction, and EPA did not propose or seek comment on such an
                approach in the proposal. The CAFE Model is able to account explicitly
                for this requirement for CAFE standards, and sets this requirement
                aside for CO2 standards.
                 Civil Penalties for Noncompliance: 49 U.S.C. 32912 prescribes a
                rate (in dollars per tenth of a mpg) at which the Secretary is to levy
                civil penalties if a manufacturer fails to comply with a CAFE standard
                for a given fleet in a given model year, after considering available
                credits. Some manufacturers have historically demonstrated a
                willingness to treat CAFE noncompliance as an ``economic'' choice,
                electing to pay civil penalties rather than achieving full numerical
                compliance across all fleets. The CAFE Model calculates civil penalties
                for CAFE shortfalls and provides means to estimate that a manufacturer
                might stop adding fuel-saving technologies once continuing to do so
                would be effectively more ``expensive'' (after accounting for fuel
                prices and buyers' willingness to pay for fuel economy) than paying
                civil penalties. In contrast, the CAA does not authorize the EPA
                Administrator to allow manufacturers to sell noncompliant fleets and
                instead only pay civil penalties; manufacturers who choose to pay civil
                penalties for CAFE compliance tend to employ EPA's more-extensive
                programmatic flexibilities to meet tailpipe CO2 emissions
                standards. Thus, the CAFE Model does not allow civil penalty payment as
                an option for CO2 standards.
                 Dual-Fueled and Dedicated Alternative Fuel Vehicles: For purposes
                of calculating CAFE levels used to determine compliance, 49 U.S.C.
                32905 and 32906 specify methods for calculating the fuel economy levels
                of vehicles operating on alternative fuels to gasoline or diesel
                through MY 2020. After MY 2020, methods for calculating alternative
                fuel vehicle (AFV) fuel economy are governed by regulation. The CAFE
                Model is able to account for these requirements explicitly for each
                vehicle model. However, 49 U.S.C. 32902 requires that maximum feasible
                CAFE standards be set in a manner that does not presume manufacturers
                can respond by producing new dedicated alternative fuel vehicle (AFV)
                models. The CAFE model can be run in a manner that excludes the
                additional application of dedicated AFV technologies in model years for
                which maximum feasible standards are under consideration. As allowed
                under NEPA for analysis appearing in EISs informing decisions regarding
                CAFE standards, the CAFE Model can also be run without this analytical
                constraint. CAA 202(a) does not preclude the EPA Administrator adopting
                analogous provisions, but EPA has instead opted through regulation to
                ``count'' dual- and alternative fuel vehicles on a CO2 basis
                (and through MY 2026, to set aside emissions from electricity
                generation). The CAFE model accounts for this treatment of dual- and
                alternative fuel vehicles when simulating manufacturers' potential
                responses to CO2 standards. For natural gas vehicles, both
                dedicated and dual-fueled, EPA is establishing a multiplier of 2.0 for
                model years 2022-2026.
                 Creation and Use of Compliance Credits: 49 U.S.C. 32903 provides
                that manufacturers may earn CAFE ``credits'' by achieving a CAFE level
                beyond that required of a given fleet in a given model year, and
                specifies how these credits may be used to offset the amount by which a
                different fleet falls short of its corresponding requirement. These
                provisions allow credits to be ``carried forward'' and ``carried back''
                between model years, transferred between regulated classes (domestic
                passenger cars, imported passenger cars, and light trucks), and traded
                between manufacturers. However, these provisions also impose some
                specific statutory limits. For example, CAFE compliance credits can be
                carried forward a maximum of five model years and carried back a
                maximum of three model years. Also, EPCA/EISA caps the amount of credit
                that can be transferred between passenger car and light truck fleets,
                and prohibits manufacturers from applying traded or transferred credits
                to offset a failure to achieve the applicable minimum standard for
                domestic passenger cars. The CAFE Model explicitly simulates
                manufacturers' potential use of credits carried forward from prior
                model years or transferred from other fleets.\98\ 49 U.S.C. 32902
                prohibits consideration of manufacturers' potential application of CAFE
                compliance credits when setting maximum feasible CAFE standards. The
                CAFE Model can be operated in a manner that excludes the application of
                CAFE credits after a given model year. CAA 202(a) does not preclude the
                EPA Administrator adopting analogous provisions. EPA has opted to limit
                the ``life'' of compliance credits from most model years to 5 years,
                and to limit borrowing to 3 years, but has not adopted any limits on
                transfers (between fleets) or trades (between manufacturers) of
                compliance credits. The CAFE Model is able to account for the absence
                of limits on transfers of CO2 standards. Insofar as the CAFE
                model can be exercised in a manner that simulates trading of
                CO2 compliance credits, such simulations treat trading as
                unlimited.\99\ EPA has considered manufacturers' ability to use credits
                as part of its decisions on these final standards, and the CAFE model
                is now able to account for that.
                ---------------------------------------------------------------------------
                 \98\ As explained in Section VI, the CAFE Model does not
                explicitly simulate the potential that manufacturers would carry
                CAFE or CO2 credits back (i.e., borrow) from future model
                years, or acquire and use CAFE compliance credits from other
                manufacturers. At the same time, because EPA has elected to not
                limit credit trading, the CAFE Model can be exercised in a manner
                that simulates unlimited (a.k.a. ``perfect'') CO2
                compliance credit trading throughout the industry (or, potentially,
                within discrete trading ``blocs''). The agencies believe there is
                significant uncertainty in how manufacturers may choose to employ
                these particular flexibilities in the future: for example, while it
                is reasonably foreseeable that a manufacturer who over-complies in
                one year may ``coast'' through several subsequent years relying on
                those credits rather than continuing to make technology
                improvements, it is harder to assume with confidence that
                manufacturers will rely on future technology investments (that may
                not pan out as expected, as if market demand for ``target-beater''
                vehicles is lower than expected) to offset prior-year shortfalls, or
                whether/how manufacturers will trade credits with market competitors
                rather than making their own technology investments. Historically,
                carry-back and trading have been much less utilized than carry-
                forward, for a variety of reasons including higher risk and
                preference not to ``pay competitors to make fuel economy
                improvements we should be making'' (to paraphrase one manufacturer),
                although the agencies recognize that carry-back and trading are used
                more frequently when standards require more technology application
                than manufacturers believe their markets will bear. Given the
                uncertainty just discussed, and given also the fact that the
                agencies have yet to resolve some of analytical challenges
                associated with simulating use of these flexibilities, the agencies
                consider borrowing and trading to involve sufficient risk that it is
                prudent to support today's decisions with analysis that sets aside
                the potential that manufacturers could come to depend widely on
                borrowing and trading. While compliance costs in real life may be
                somewhat different from what is modeled today as a result of this
                analytical decision, that is broadly true no matter what, and the
                agencies do not believe that the difference would be so great that
                it would change the policy outcome.
                 \99\ To avoid making judgments (that would invariably turn out
                to be at least somewhat incorrect) about possible future trading
                activity, the model simulates trading by combining all manufacturers
                into a single entity, so that the most cost-effective choices are
                made for the fleet as a whole.
                ---------------------------------------------------------------------------
                 Statutory Basis for Stringency: 49 U.S.C. 32902 requires the
                Secretary to set CAFE standards at the maximum feasible levels,
                considering technological feasibility, economic practicability, the
                need of the Nation to conserve energy, and the impact of other
                government standards. EPCA/EISA authorizes the Secretary to interpret
                [[Page 24221]]
                these factors, and as the Department's interpretation has evolved,
                NHTSA has continued to expand and refine its qualitative and
                quantitative analysis. For example, as discussed below in Section
                VI.B.3, the Autonomie simulations reflect the agencies' judgment that
                it would not be economically practicable for a manufacturer to
                ``split'' an engine shared among many vehicle model/configurations into
                a myriad of versions each optimized to a single vehicle model/
                configuration. Also responding to evolving interpretation of these
                EPCA/EISA factors, the CAFE Model has been expanded to address
                additional impacts in an integrated manner. For example, the CAFE Model
                version used for the NPRM analysis included the ability to estimate
                impacts on labor utilization internally, rather than as an external
                ``off model'' or ``post processing'' analysis. In addition, NEPA
                requires the Secretary to issue an EIS that documents the estimated
                impacts of regulatory alternatives under consideration. The EIS
                accompanying today's notice documents changes in emission inventories
                as estimated using the CAFE model, but also documents corresponding
                estimates--based on the application of other models documented in the
                EIS, of impacts on the global climate, on tropospheric air quality, and
                on human health. Regarding CO2 standards, CAA 202(a)
                provides general authority for the establishment of motor vehicle
                emissions standards, and the final rule's analysis, like that
                accompanying the agencies' proposal, addresses impacts relevant to the
                EPA Administrator's decision making, such as technological feasibility,
                air quality impacts, costs to industry and consumers, and lead time
                necessary for compliance.
                 Other Factors: Beyond these statutory requirements applicable to
                DOT and/or EPA are a number of specific technical characteristics of
                CAFE and/or CO2 regulations that are also relevant to the
                construction of today's analysis. These are discussed at greater length
                in Section II.F. For example, EPA has defined procedures for
                calculating average CO2 levels, and has revised procedures
                for calculating CAFE levels, to reflect manufacturers' application of
                ``off-cycle'' technologies that increase fuel economy (and reduce
                CO2 emissions) in ways not reflected by the long-standing
                test procedures used to measure fuel economy. Although too little
                information is available to account for these provisions explicitly in
                the same way that the agencies have accounted for other technologies,
                the CAFE Model does include and makes use of inputs reflecting the
                agencies' expectations regarding the extent to which manufacturers may
                earn such credits, along with estimates of corresponding costs.
                Similarly, the CAFE Model includes and makes use of inputs regarding
                credits EPA has elected to allow manufacturers to earn toward
                CO2 levels (not CAFE) based on the use of air conditioner
                refrigerants with lower global warming potential (GWP), or on the
                application of technologies to reduce refrigerant leakage. In addition,
                EPA has elected to provide that through model year 2021, manufacturers
                may apply ``multipliers'' to plug-in hybrid electric vehicles,
                dedicated electric vehicles, fuel cell vehicles, and hydrogen vehicles,
                such that when calculating a fleet's average CO2 levels (not
                CAFE), the manufacturer may, for example, ``count'' each electric
                vehicle twice. The CAFE Model accounts for these multipliers, based on
                either current regulatory provisions or on alternative approaches.
                Although these are examples of regulatory provisions that arise from
                the exercise of discretion rather than specific statutory mandate, they
                can materially impact outcomes. Section VI.B explains in greater detail
                how today's analysis addresses them.
                Benefits of Analytical Approach
                 The agencies' analysis of CAFE and CO2 standards
                involves two basic elements: First, estimating ways each manufacturer
                could potentially respond to a given set of standards in a manner that
                considers potential consumer response; and second, estimating various
                impacts of those responses. Estimating manufacturers' potential
                responses involves simulating manufacturers' decision-making processes
                regarding the year-by-year application of fuel-saving technologies to
                specific vehicles. Estimating impacts involves calculating resultant
                changes in new vehicle costs, estimating a variety of costs (e.g., for
                fuel) and effects (e.g., CO2 emissions from fuel combustion)
                occurring as vehicles are driven over their lifetimes before eventually
                being scrapped, and estimating the monetary value of these effects.
                Estimating impacts also involves consideration of the response of
                consumers--e.g., whether consumers will purchase the vehicles and in
                what quantities. Both of these basic analytical elements involve the
                application of many analytical inputs.
                 As mentioned above, the agencies' analysis uses the CAFE model to
                estimate manufacturers' potential responses to new CAFE and
                CO2 standards and to estimate various impacts of those
                responses. DOT's Volpe National Transportation Systems Center (often
                simply referred to as the ``Volpe Center'') develops, maintains, and
                applies the model for NHTSA. NHTSA has used the CAFE model to perform
                analyses supporting every CAFE rulemaking since 2001, and the 2016
                rulemaking regarding heavy-duty pickup and van fuel consumption and
                CO2 emissions also used the CAFE model for analysis.\100\
                ---------------------------------------------------------------------------
                 \100\ While both agencies used the CAFE Model to simulate
                manufacturers' potential responses to standards, some model inputs
                differed EPA's and DOT's analyses, and EPA also used the EPA MOVES
                model to calculate resultant changes in emissions inventories. See
                81 FR 73478, 73743 (Oct. 25, 2016).
                ---------------------------------------------------------------------------
                 NHTSA recently arranged for a formal peer review of the model. In
                general, reviewers' comments strongly supported the model's conceptual
                basis and implementation, and commenters provided several specific
                recommendations. The agency agreed with many of these recommendations
                and has worked to implement them wherever practicable. Implementing
                some of the recommendations would require considerable further
                research, development, and testing, and will be considered going
                forward. For a handful of other recommendations, the agency disagreed,
                often finding the recommendations involved considerations (e.g., other
                policies, such as those involving fuel taxation) beyond the model
                itself or were based on concerns with inputs rather than how the model
                itself functioned. A report available in the docket for this rulemaking
                presents peer reviewers' detailed comments and recommendations, and
                provides DOT's detailed responses.\101\
                ---------------------------------------------------------------------------
                 \101\ Docket No. NHTSA-2018-0067-0055.
                ---------------------------------------------------------------------------
                 As also mentioned above, the agencies use EPA's MOVES model to
                estimate tailpipe emission factors, use DOE/EIA's NEMS to estimate fuel
                prices,\102\ and use Argonne's GREET model to estimate downstream
                emissions rates.\103\ DOT also sponsored DOE/Argonne to use the
                Autonomie full-vehicle modeling and simulation tool to estimate the
                fuel economy impacts for roughly a million
                [[Page 24222]]
                combinations of technologies and vehicle types.104 105
                ---------------------------------------------------------------------------
                 \102\ See https://www.eia.gov/outlooks/aeo/info_nems_archive.php. Today's notice uses fuel prices estimated
                using the Annual Energy Outlook (AEO) 2019 version of NEMS (see
                https://www.eia.gov/outlooks/archive/aeo19/ and https://www.eia.gov/outlooks/aeo/data/browser/#/?id=3-AEO2019&cases=ref2019&sourcekey=0).
                 \103\ Information regarding GREET is available at https://greet.es.anl.gov/index.php. Availability of NEMS is discussed at
                https://www.eia.gov/outlooks/aeo/info_nems_archive.php. Today's
                notice uses fuel prices estimated using the AEO 2019 version of
                NEMS.
                 \104\ As part of the Argonne simulation effort, individual
                technology combinations simulated in Autonomie were paired with
                Argonne's BatPAC model to estimate the battery cost associated with
                each technology combination based on characteristics of the
                simulated vehicle and its level of electrification. Information
                regarding Argonne's BatPAC model is available at http://www.cse.anl.gov/batpac/.
                 \105\ Furthermore, the impact of engine technologies on fuel
                consumption, torque, and other metrics was characterized using GT
                POWER simulation modeling in combination with other engine modeling
                that was conducted by IAV Automotive Engineering, Inc. (IAV). The
                engine characterization ``maps'' resulting from this analysis were
                used as inputs for the Autonomie full-vehicle simulation modeling.
                Information regarding GT Power is available at https://www.gtisoft.com/gt-suite-applications/propulsion-systems/gt-power-engine-simulation-software.
                ---------------------------------------------------------------------------
                 EPA developed two models after 2009, referred to as the ``ALPHA''
                and ``OMEGA'' models, which provide some of the same capabilities as
                the Autonomie and CAFE models. EPA applied the OMEGA model to conduct
                analysis of tailpipe CO2 emissions standards promulgated in
                2010 and 2012, and the ALPHA and OMEGA models to conduct analysis
                discussed in the above-mentioned 2016 Draft TAR and Proposed and 2017
                Initial Final Determinations regarding standards beyond 2021. In an
                August 2017 notice, the agencies requested comments on, among other
                things, whether EPA should use alternative methodologies and modeling,
                including DOE/Argonne's Autonomie full-vehicle modeling and simulation
                tool and DOT's CAFE model.\106\
                ---------------------------------------------------------------------------
                 \106\ 82 FR 39551, 39553 (Aug. 21, 2017).
                ---------------------------------------------------------------------------
                 Having reviewed comments on the subject and having considered the
                matter fully, the agencies have determined it is reasonable and
                appropriate to use DOE/Argonne's model for full-vehicle simulation, and
                to use DOT's CAFE model for analysis of regulatory alternatives. EPA
                interprets Section 202(a) of the CAA as giving the agency broad
                discretion in how it develops and sets CO2 emissions
                standards for light-duty vehicles. Nothing in Section 202(a) mandates
                that EPA use any specific model or set of models for analysis of
                potential CO2 standards for light-duty vehicles. EPA weighs
                many factors when determining appropriate levels for CO2
                standards, including the cost of compliance (see Section 202(a)(2)),
                lead time necessary for compliance (id.), safety (see NRDC v. EPA, 655
                F.2d 318, 336 n. 31 (D.C. Cir. 1981)) and other impacts on
                consumers,\107\ and energy impacts associated with use of the
                technology.\108\ Using the CAFE model allows consideration of a number
                of factors. The CAFE model explicitly evaluates the cost of compliance
                for each manufacturer, each fleet, and each model year; it accounts for
                lead time necessary for compliance by directly incorporating estimated
                manufacturer production cycles for every vehicle in the fleet, ensuring
                that the analysis does not assume vehicles can be redesigned to
                incorporate more technology without regard to lead time considerations;
                it provides information on safety effects associated with different
                levels of standards and information about many other impacts on
                consumers, and it calculates energy impacts (i.e., fuel saved or
                consumed) as a primary function, besides being capable of providing
                information about many other factors within EPA's broad CAA discretion
                to consider.
                ---------------------------------------------------------------------------
                 \107\ Since its earliest Title II regulations, EPA has
                considered the safety of pollution control technologies. See 45 FR
                14496, 14503 (1980).
                 \108\ See George E. Warren Corp. v. EPA, 159 F.3d 616, 623-624
                (D.C. Cir. 1998) (ordinarily permissible for EPA to consider factors
                not specifically enumerated in the Act).
                ---------------------------------------------------------------------------
                 Because the CAFE model simulates a wide range of actual constraints
                and practices related to automotive engineering, planning, and
                production, such as common vehicle platforms, sharing of engines among
                different vehicle models, and timing of major vehicle redesigns, the
                analysis produced by the CAFE model provides a transparent and
                realistic basis to show pathways manufacturers could follow over time
                in applying new technologies, which helps better assess impacts of
                potential future standards. Furthermore, because the CAFE model also
                accounts fully for regulatory compliance provisions (now including
                CO2 compliance provisions), such as adjustments for reduced
                refrigerant leakage, production ``multipliers'' for some specific types
                of vehicles (e.g., PHEVs), and carried-forward (i.e., banked) credits,
                the CAFE model provides a transparent and realistic basis to estimate
                how such technologies might be applied over time in response to CAFE or
                CO2 standards.
                 There are sound reasons for the agencies to use the CAFE model
                going forward in this rulemaking. First, the CAFE and CO2
                fact analyses are inextricably linked. Furthermore, the analysis
                produced by the CAFE model and DOE/Argonne's Autonomie addresses the
                agencies' analytical needs. The CAFE model provides an explicit year-
                by-year simulation of manufacturers' application of technology to their
                products in response to a year-by-year progression of CAFE standards
                and accounts for sharing of technologies and the implications for
                timing, scope, and limits on the potential to optimize powertrains for
                fuel economy. In the real world, standards actually are specified on a
                year-by-year basis, not simply some single year well into the future,
                and manufacturers' year-by-year plans involve some vehicles ``carrying
                forward'' technology from prior model years and some other vehicles
                possibly applying ``extra'' technology in anticipation of standards in
                ensuing model years, and manufacturers' planning also involves applying
                credits carried forward between model years. Furthermore, manufacturers
                cannot optimize the powertrain for fuel economy on every vehicle model
                configuration--for example, a given engine shared among multiple
                vehicle models cannot practicably be split into different versions for
                each configuration of each model, each with a slightly different
                displacement. The CAFE model is designed to account for these real-
                world factors.
                 Considering the technological heterogeneity of manufacturers'
                current product offerings, and the wide range of ways in which the many
                fuel economy-improving/CO2 emissions-reducing technologies
                included in the analysis can be combined, the CAFE model has been
                designed to use inputs that provide an estimate of the fuel economy
                achieved for many tens of thousands of different potential combinations
                of fuel-saving technologies. Across the range of technology classes
                encompassed by the analysis fleet, today's analysis involves more than
                a million such estimates. While the CAFE model requires no specific
                approach to developing these inputs, the National Academy of Sciences
                (NAS) has recommended, and stakeholders have commented, that full-
                vehicle simulation provides the best balance between realism and
                practicality. DOE/Argonne has spent several years developing, applying,
                and expanding means to use distributed computing to exercise its
                Autonomie full-vehicle modeling and simulation tool over the scale
                necessary for realistic analysis of CAFE or average tailpipe
                CO2 emissions standards. This scalability and related
                flexibility (in terms of expanding the set of technologies to be
                simulated) makes Autonomie well-suited for developing inputs to the
                CAFE model.
                 In addition, DOE/Argonne's Autonomie also has a long history of
                development and widespread application by a much wider range of users
                in government, academia, and industry. Many of these users apply
                [[Page 24223]]
                Autonomie to inform funding and design decisions. These real-world
                exercises have contributed significantly to aspects of Autonomie
                important to producing realistic estimates of fuel economy levels and
                CO2 emission rates, such as estimation and consideration of
                performance, utility, and driveability metrics (e.g., towing
                capability, shift business, frequency of engine on/off transitions).
                This steadily increasing realism has, in turn, steadily increased
                confidence in the appropriateness of using Autonomie to make
                significant investment decisions. Notably, DOE uses Autonomie for
                analysis supporting budget priorities and plans for programs managed by
                its Vehicle Technologies Office (VTO). Considering the advantages of
                DOE/Argonne's Autonomie model, it is reasonable and appropriate to use
                Autonomie to estimate fuel economy levels and CO2 emission
                rates for different combinations of technologies as applied to
                different types of vehicles.
                 Commenters have also suggested that the CAFE model's graphical user
                interface (GUI) facilitates others' ability to use the model quickly--
                and without specialized knowledge or training--and to comment
                accordingly.\109\ For the NPRM, NHTSA significantly expanded and
                refined this GUI, providing the ability to observe the model's real-
                time progress much more closely as it simulates year-by-year compliance
                with either CAFE or CO2 standards.\110\ Although the model's
                ability to produce realistic results is independent of the model's GUI,
                the CAFE model's GUI appears to have facilitated stakeholders'
                meaningful review and comment during the comment period.
                ---------------------------------------------------------------------------
                 \109\ From Docket Number EPA-HQ-OAR-2015-0827, see Comment by
                Global Automakers, Docket ID EPA-HQ-OAR-2015-0827-9728, at 34.
                 \110\ The updated GUI provides a range of graphs updated in real
                time as the model operates. These graphs can be used to monitor fuel
                economy or CO2 ratings of vehicles in manufacturers'
                fleets and to monitor year-by-year CAFE (or average CO2
                ratings), costs, avoided fuel outlays, and avoided CO2-
                related damages for specific manufacturers and/or specific fleets
                (e.g., domestic passenger car, light truck). Because these graphs
                update as the model progresses, they should greatly increase users'
                understanding of the model's approach to considerations such as
                multiyear planning, payment of civil penalties, and credit use.
                ---------------------------------------------------------------------------
                 The question of whether EPA's actions should consider and be
                informed by analysis using non-EPA-staff-developed modeling tools has
                generated considerable debate over time. Even prior to the NPRM,
                certain commenters had argued that EPA could not consider, in setting
                tailpipe CO2 emissions standards, any information derived
                from non-EPA-staff-developed modeling. Many of the pre-NPRM concerns
                focused on inputs used by the CAFE model for prior rulemaking
                analyses.111 112 113 Because inputs are exogenous to any
                model, they do not determine whether it would be reasonable and
                appropriate for EPA to use NHTSA's model for analysis. Other concerns
                focused on certain characteristics of the CAFE model that were
                developed to align the model better with EPCA and EISA. The model has
                been revised to accommodate both EPCA/EISA and CAA analysis, as
                explained further below. Some commenters also argued that use of any
                models other than ALPHA and OMEGA for CAA analysis would constitute an
                arbitrary and capricious delegation of EPA's decision-making authority
                to NHTSA, if NHTSA models are used for analysis instead.\114\ As
                discussed above, the CAFE Model--as with any model--is used to provide
                analysis, and does not result in decisions. Decisions are made by EPA
                in a manner that is informed by modeling outputs, sensitivity cases,
                public comments, any many other pieces of information.
                ---------------------------------------------------------------------------
                 \111\ For example, EDF previously stated that ``the data that
                NHTSA needs to input into its model is sensitive confidential
                business information that is not transparent and cannot be
                independently verified, . . .'' and it claimed ``the OMEGA model's
                focus on direct technological inputs and costs--as opposed to
                industry self-reported data--ensures the model more accurately
                characterizes the true feasibility and cost effectiveness of
                deploying greenhouse gas reducing technologies.'' EDF, EPA-HQ-OAR-
                2015-0827-9203, at 12. These statements are not correct, as nothing
                about either the CAFE or OMEGA model either obviates or necessitates
                the use of CBI to develop inputs.
                 \112\ As another example, CARB previously stated that ``another
                promising technology entering the market was not even included in
                the NHTSA compliance modeling'' and that EPA assumes a five-year
                redesign cycle, whereas NHTSA assumes a six to seven-year cycle.''
                CARB, EPA-HQ-OAR-2015-0827-9197, at 28. Though presented as
                criticisms of the models, these comments--at least with respect to
                the CAFE model--actually concern model inputs. NHTSA did not agree
                with CARB about the commercialization potential of the engine
                technology in question (``Atkinson 2'') and applied model inputs
                accordingly. Also, rather than applying a one-size-fits-all
                assumption regarding redesign cadence, NHTSA developed estimates
                specific to each vehicle model and applied these as model inputs.
                 \113\ As another example, NRDC has argued that EPA should not
                use the CAFE model because it ``allows manufacturers to pay civil
                penalties in lieu of meeting the standards, an alternative
                compliance pathway currently allowed under EISA and EPCA.'' NRDC,
                EPA-HQ-OAR-2015-0827-9826, at 37. While the CAFE model can simulate
                civil penalty payment, NRDC's comment appears to overlook the fact
                that this result depends on model inputs; the inputs can easily be
                specified such that the CAFE model will set aside civil penalty
                payment as an alternative to compliance.
                 \114\ See, e.g., CBD et al., NHTSA-2018-0067-12057, at 9.
                ---------------------------------------------------------------------------
                 Comments responding to the NPRM's use of the CAFE model and
                Autonomie rather than also (for CO2 standards) ALPHA and
                OMEGA were mixed. For example, the environmental group coalition stated
                that the CAFE model is aligned with EPCA requirements,\115\ but also
                argued (1) that EPA is legally prohibited from ``delegat[ing] technical
                decision-making to NHTSA;'' \116\ (2) that ``EPA must exercise its
                technical and scientific expertise'' to develop CO2
                standards and ``Anything less is an unlawful abdication of EPA's
                statutory responsibilities;'' \117\ (3) that EPA staff is much more
                qualified than DOT staff to conduct analysis relating to standards and
                has done a great deal of work to inform development of standards; \118\
                (4) that ``The Draft TAR and 2017 Final Determination relied
                extensively on use of sophisticated EPA analytic tools and
                methodologies,'' i.e., the ``peer reviewed simulation model ALPHA,''
                ``the agency's vehicle teardown studies,'' and the ``peer-reviewed
                OMEGA model to make reasonable estimates of how manufacturers could add
                technologies to vehicles in order to meet a fleet-wide [CO2
                emissions] standard;'' \119\ (5) that NHTSA had said in the MYs 2012-
                2016 rulemaking that the Volpe [CAFE] model was developed to support
                CAFE rulemaking and incorporates features ``that are not appropriate
                for use by EPA in setting [tailpipe CO2] standards;'' \120\
                (6) allegations that some EPA staff had disagreed with aspects of the
                NPRM analysis and had requested that ``EPA's name and logo should be
                removed from the DOT-NHTSA Preliminary Regulatory Impact Analysis
                document'' and stated that ``EPA is relying upon the technical analysis
                performed by DOT-NHTSA for the [NPRM];'' \121\ (7) that EPA had
                developed ``a range of relevant new analysis'' that the proposal
                ``failed to consider,'' including ``over a dozen 2017 and 2018 EPA peer
                reviewed SAE articles;'' \122\ (8) that EPA's OMEGA modeling undertaken
                during NPRM development ``found costs half that of NHTSA's findings,''
                ``Yet NHTSA did not correct the errors in its modeling and analysis,
                and the published proposal drastically overestimates the cost of
                complying . . . .;'' \123\ (9) that some EPA staff had requested that
                the technology ``HCR2'' be included in the NPRM analysis, ``Yet NHTSA
                overruled
                [[Page 24224]]
                EPA and omitted the technology;'' \124\ (10) that certain EPA staff had
                initially ``rejected use of the CAFE model for development of the
                proposed [tailpipe CO2] standards;'' \125\ (11) that there
                are ``many specific weaknesses of the modeling results derived in this
                proposal through use of the Volpe and Autonomie models'' and that the
                CAFE model is ``not designed in accordance with'' Section 202(a) of the
                CAA because (A) EPA ``is not required to demonstrate that standards are
                set at the maximum feasible level year-by-year,'' (B) because EPCA
                ``preclude[s NHTSA] from considering vehicles powered by fuels other
                than gas or diesel'' and EPA is not similarly bound, and (C) because
                the CAFE model assumed that the value of an overcompliance credit
                equaled $5.50, the value of a CAFE penalty.\126\ Because of all of
                these things, the environmental group coalition stated that the
                proposal was ``unlawful'' and that ``Before proceeding with this
                rulemaking, EPA must consider all relevant materials including these
                excluded insights, perform its own analysis, and issue a reproposal to
                allow for public comment.'' \127\
                ---------------------------------------------------------------------------
                 \115\ Environmental group coalition, NHTSA-2018-0067-12000,
                Appendix A, at 24-25.
                 \116\ Id. at 12.
                 \117\ Id. at 14.
                 \118\ Id. at 15-17.
                 \119\ Id. at 17.
                 \120\ Id. at 18.
                 \121\ Id. at 19.
                 \122\ Id. at 20.
                 \123\ Id. at 21.
                 \124\ Id. at 21-22.
                 \125\ Id. at 23.
                 \126\ Id. at 24-25.
                 \127\ Id. at 27.
                ---------------------------------------------------------------------------
                 Some environmental organizations and States contracted for external
                technical analyses augmenting general comments such as those summarized
                above. EDF engaged a consultant, Richard Rykowski, for a detailed
                review of the agencies' analysis.\128\ Among Mr. Rykowski's comments, a
                few specifically involve differences between these two models. Mr.
                Rykowski recommended NHTSA's CAFE model replace its existing
                ``effective cost'' metric (used to compare available options to add
                specific technologies to specific vehicles) with a ``ranking factor''
                used for the same purpose. As discussed below in Section VI.A, the
                model for today's final rule adopts this recommendation. He also states
                that (1) ``EPA has developed a better way to isolate and reject cost
                ineffective combinations of technologies . . . [and] includes only
                these 50 or so technology combinations in their OMEGA model runs;'' (2)
                ``NHTSA's arbitrary and rigid designation of leader-follower vehicles
                for engine, transmission and platform level technologies
                unrealistically slows the rollout of technology into the new vehicle
                fleet;'' (3) ``the Volpe Model is not capable of reasonably simulating
                manufacturers' ability to utilize CO2 credits to smooth the
                introduction of technology throughout their vehicle line-up;'' and (4)
                ``the Volpe Model is not designed to reflect the use of these [A/C
                leakage] technologies and refrigerants.'' \129\
                ---------------------------------------------------------------------------
                 \128\ EDF, NHTSA-2018-0067-12108, Appendix B. See also EPA, Peer
                Review of the Optimization Model for Reducing Emissions of
                Greenhouse Gases from Automobiles (OMEGA) and EPA's Response to
                Comments, EPA-420-R-09-016, September 2009.
                 \129\ EDF, op. cit., at 73-75.
                ---------------------------------------------------------------------------
                 Mr. Rogers's analysis focuses primarily on the agencies' published
                analysis, but mentions that some engine ``maps'' (estimates--used as
                inputs to full vehicle simulation--of engine fuel consumption under a
                wide range of engine operating conditions) applied in Autonomie show
                greater fuel consumption benefits of turbocharging than those applied
                previously by EPA to EPA's ALPHA model, and these benefits could have
                caused NHTSA's CAFE model to estimate an unrealistically great tendency
                toward turbocharged engines (rather than high compression ratio
                engines).\130\ Mr. Rogers also presents alternative examples of year-
                by-year technology application to specific vehicle models, contrasting
                these with year-by-year results from the agencies' NPRM analysis,
                concluding that ``that the use of logical, unrestricted technology
                pathways, with incremental benefits supported by industry-accepted
                vehicle simulation and dynamic system optimization and calibration,
                together with publicly-defensible costs, allows cost-effective
                solutions to achieve target fuel economy levels which meet MY 2025
                existing standards.'' \131\
                ---------------------------------------------------------------------------
                 \130\ Roush Industries, NHTSA-2018-0067-11984, at 17-21.
                 \131\ Roush Industries, NHTSA-2018-0067-11984, at 17-30.
                ---------------------------------------------------------------------------
                 Mr. Duleep's analysis also focuses primarily on the agencies'
                published analysis, but does mention that (1) ``the Autonomie modeling
                assumes no engine change when drag and rolling resistance reductions
                are implemented, as well as no changes to the transmission gear ratios
                and axle ratios, . . . [but] the EPA ALPHA model adjusts for this
                effect;'' (2) ``baseline differences in fuel economy [between two
                manufacturers' different products using similar technologies] are
                carried for all future years and this exaggerates the differences in
                technology adoption requirements and costs between manufacturers; (3)
                ``assumptions [that most technology changes are best applied as part of
                a vehicle redesign or freshening] result in unnecessary distortion in
                technology paths and may bias results of costs for different
                manufacturers;'' and (4) that for the sample results shown for the
                Chevrolet Equinox ``the publicly available EPA lumped parameter model
                (which was used to support the 2016 rulemaking) and 2016 TAR cost data
                . . . results in an estimate of attaining 52.2 mpg for a cost of $2110,
                which is less than half the cost estimated in the PRIA.'' \132\
                ---------------------------------------------------------------------------
                 \132\ H-D Systems, op. cit., at 48, et seq.
                ---------------------------------------------------------------------------
                 Beyond these comments regarding differences between EPA's models
                and the Argonne and DOT models applied for the NPRM, these and other
                technical reviewers had many specific comments about the agencies'
                analysis for the NPRM, and these comments are discussed in detail below
                in Section VI.B.
                 Manufacturers, on the other hand, supported the agencies' use of
                Autonomie and the CAFE model rather than, in EPA's case, the ALPHA and
                OMEGA models. Expressly identifying the distinction between models and
                model inputs, Global Automakers stated that:
                 The agencies provided a new, updated analysis based on the most
                up-to-date data, using a proven and long-developed modeling tool,
                known as the Volpe model, and offering numerous options to best
                determine the right regulatory and policy path for ongoing fuel
                efficiency improvements in our nation. Now, all stakeholders have an
                opportunity to come to the table as part of the public process to
                provide input, data, and information to help shape the final
                rule.\133\
                ---------------------------------------------------------------------------
                 \133\ Global Automakers, NHTSA-2018-0067-12032, at 2.
                ---------------------------------------------------------------------------
                 This NPRM's use of a single model to evaluate alternative
                scenarios for both programs provides consistency in the technical
                analysis, and Global Automakers supports the Volpe model's use as it
                has proven to be a transparent and user-friendly option in this
                current analysis. The use of the Volpe model has allowed for a broad
                range of stakeholders, with varying degrees of technical expertise,
                to review the data inputs to provide feedback on this proposed rule.
                The Volpe model's accompanying documentation has historically
                provided a clear explanation of all sources of input and constraints
                critical to a transparent modeling process. Other inputs have come
                from modeling that is used widely by other sources, specifically the
                Autonomie model, allowing for a robust validation, review and
                reassessment.\134\
                ---------------------------------------------------------------------------
                 \134\ Global Automakers, NHTSA-2018-0067-12032, Attachment A, at
                A-12.
                 The Alliance commented, similarly, that ``at least at this time,
                NHTSA's modeling systems are superior to EPA's'' and ``as such, we
                support the Agencies' decision to use NHTSA's modeling tools for this
                rulemaking and recommend that both Agencies continue on this path. We
                encourage Agencies to work together to provide input to the single
                common set of tools.'' \135\
                ---------------------------------------------------------------------------
                 \135\ Alliance, NHTSA-2018-0067-12073, at 134.
                ---------------------------------------------------------------------------
                [[Page 24225]]
                 Regarding the agencies' use of Argonne's Autonomie model rather
                than EPA's ALPHA model, the Alliance commented that (1) ``the benefits
                of virtually all technologies and their synergistic effects are now
                determined with full vehicle simulations;'' (2) ``vehicle categories
                have been increased to 10 to better recognize the range of 0-60
                performance characteristics within each of the 5 previous categories,
                in recognition of the fact that many vehicles in the baseline fleet
                significantly exceeded the previously assumed 0-60 performance metrics.
                This provides better resolution of the baseline fleet and more accurate
                estimates of the benefits of technology. . . .;'' (3) ``new
                technologies (like advanced cylinder deactivation) are included, while
                unproven combinations (like Atkinson engines with 14:1 compression,
                cooled EGR, and cylinder deactivation in combination) have been
                removed;'' (4) ``Consistent with the recommendation of the National
                Academy of Sciences and manufacturers, gradeability has been included
                as a performance metric used in engine sizing. This helps prevent the
                inclusion of small displacement engines that are not commercially
                viable and that would artificially inflate fuel savings;'' (5) ``the
                Alliance believes NHTSA's tools (Autonomie/Volpe) are superior to EPA's
                (APLHA[sic]/LPM/OMEGA). This is not surprising since NHTSA's tools have
                had a significant head start in development. . . .'' (6) ``the
                Autonomie model was developed at Argonne National Lab with funding from
                the Department of Energy going back to the PNGV (Partnership for Next
                Generation Vehicles) program in the 1990s. Autonomie was developed from
                the start to address the complex task of combining 2 power sources in a
                hybrid powertrain. It is a physics-based, forward looking, vehicle
                simulator, fully documented with available training,'' and (7) ``EPA's
                ALPHA model is also a physics-based, forward looking, vehicle
                simulator. However, it has not been validated or used to simulate
                hybrid powertrains. The model has not been documented with any
                instructions making it difficult for users outside of EPA to run and
                interpret the model.'' \136\
                ---------------------------------------------------------------------------
                 \136\ Id. at 135.
                ---------------------------------------------------------------------------
                 Regarding the use of NHTSA's CAFE model rather than EPA's OMEGA
                model, the Alliance stated that (1) NHTSA's model appropriately
                differentiate between domestic and imported automobiles; (2) in NHTSA's
                model, ``dynamic estimates of vehicle sales and scrappage in response
                to price changes replace unrealistic static sales and scrappage
                numbers;'' (3) NHTSA's model ``has new capability to perform
                [CO2 emissions] analysis with [tailpipe CO2]
                program flexibilities;'' (4) ``the baseline fleet [used in NHTSA's
                model] has been appropriately updated based on both public and
                manufacturer data to reflect the technologies already applied,
                particularly tire rolling resistance;'' and (5) ``some technologies
                have been appropriately restricted. For example, low rolling resistance
                tires are no longer allowed on performance vehicles, and aero
                improvements are limited to maximum levels of 15% for trucks and 10%
                for minivans.'' \137\ The Alliance continued, noting that ``NHTSA's
                Volpe model also predates EPA's OMEGA model. More importantly, the new
                Volpe model considers several factors that make its results more
                realistic.'' \138\ As factors leading the Volpe model to produce
                results that are more realistic than those produced by OMEGA, the
                Alliance commented that (1) ``The Volpe model includes estimates of the
                redesign and refresh schedules of vehicles based on historical trends,
                whereas the OMEGA model uses a fixed, and too short, time interval
                during which all vehicles are assumed to be fully redesigned. . . .;''
                (2) ``The Volpe model allows users to phase-in technology based on year
                of availability, platform technology sharing, phase-in caps, and to
                follow logical technology paths per vehicle. . . .;'' (3) ``The Volpe
                model produces a year-by year analysis from the baseline model year
                through many years in the future, whereas the OMEGA model only analyzes
                a fixed time interval. . . .;'' (4) ``The Volpe model recognizes that
                vehicles share platforms, engines, and transmissions, and that
                improvements to any one of them will likely extend to other vehicles
                that use them'' whereas ``The OMEGA model treats each vehicle as an
                independent entity. . . .;'' (5) ``The Volpe model now includes sales
                and scrappage effects;'' and (6) ``The Volpe model is now capable of
                analyzing for CAFE and [tailpipe CO2] compliance, each with
                unique program restrictions and flexibilities.'' \139\ The Alliance
                also incorporated by reference concerns it raised regarding EPA's
                OMEGA-based analysis supporting EPA's proposed and prior final
                determinations.\140\
                ---------------------------------------------------------------------------
                 \137\ Id. at 134.
                 \138\ Id. at 135.
                 \139\ Id. at 135-136.
                 \140\ Id. at 136.
                ---------------------------------------------------------------------------
                 The Alliance further stated that ``For all of the above reasons and
                to avoid duplicate efforts, the Alliance recommends that the Agencies
                continue to use DOT's Volpe and Autonomie modeling system, rather than
                continuing to develop two separate systems. EPA has demonstrated
                through supporting Volpe model code revisions and by supplying engine
                maps for use in the Autonomie model that their expertise can be
                properly represented in the rulemaking process without having to
                develop separate or new tools.'' \141\
                ---------------------------------------------------------------------------
                 \141\ Id. at 136.
                ---------------------------------------------------------------------------
                 Some individual manufacturers provided comments supporting and
                elaborating on the above comments by Global Automakers and the
                Alliance. For example, FCA commented that ``the modeling performed by
                the agencies should illuminate the differences between the CAFE and
                [tailpipe CO2 emissions] programs. This cannot be
                accomplished when each agency is using different tools and assumptions.
                Since we believe NHTSA possesses the better set of tools, we support
                both agencies using Autonomie for vehicle modeling and Volpe (CAFE) for
                fleet modeling.'' \142\
                ---------------------------------------------------------------------------
                 \142\ FCA, NHTSA-2018-0067-11943, at 82.
                ---------------------------------------------------------------------------
                 Honda stated that ``The current version of the CAFE model is
                reasonably accurate in terms of technology efficiency, cost, and
                overall compliance considerations, and reflects a notable improvement
                over previous agency modeling efforts conducted over the past few
                years. We found the CAFE model's characterization of Honda's
                ``baseline'' fleet--critical modeling minutiae that provide a technical
                foundation of the agencies' analysis--to be highly accurate. We commend
                NHTSA and Volpe Center staff on these updates, as well as on the
                overall transparency of the model. The model's graphical user interface
                (GUI) makes it easier to run, model functionality is thoroughly
                documented, and the use of logical, traceable input and output files
                accommodates easy tracking of results.'' \143\ Similarly, in an earlier
                presentation to the agencies, Honda included the following slide
                comparing EPA's OMEGA model to DOT's CAFE (Volpe) model, and making
                recommendations regarding future improvements to the latter: \144\
                ---------------------------------------------------------------------------
                 \143\ Honda, EPA-HQ-OAR-2018-0283, at 21-22.
                 \144\ Honda, NHTSA-2018-0067-12019, at 12.
                ---------------------------------------------------------------------------
                [[Page 24226]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.041
                 Toyota, in addition to arguing that the agencies' application of
                model inputs (e.g., an analysis fleet based on MY 2016 compliance data)
                produced more realistic results than in the draft TAR and in EPA's
                former proposed and final determinations, also stressed the importance
                of the CAFE model's year-by-year accounting for product redesigns,
                stating that this produces more realistic results than the OMEGA-based
                results shown previously by EPA:
                 The modeling now better accounts for factors that limit the rate
                at which new technologies enter and then diffuse through a
                manufacturer's fleet. Bringing new or improved vehicles and
                technologies to market is a several-year, capital-intensive
                undertaking. Once new designs are introduced, a period of stability
                is required so investments can be amortized. Vehicle and technology
                introductions are staggered over time to manage limited resources.
                Agency modeling now better recognizes the inherent constraints
                imposed by realities that dictate product cadence. We agree with the
                agencies' understanding that ``the simulation of compliance actions
                that manufacturers might take is constrained by the pace at which
                new technologies can be applied in the new vehicle market,'' and we
                are encouraged to learn that ``agency modeling can now account for
                the fact that individual vehicle models undergo significant
                redesigns relatively infrequently.'' The preamble correctly notes
                that manufacturers try to keep costs down by applying most major
                changes mainly during vehicle redesigns and more modest changes
                during product refresh, and that redesign cycles for vehicle models
                can range from six to ten years, and eight to ten-years for
                powertrains. This appreciation for standard business practice
                enables the modeling to more accurately capture the way vehicles
                share engines, transmissions, and platforms. There are now more
                realistic limits placed on the number of engines and transmissions
                in a powertrain portfolio which better recognizes manufacturers must
                manage limited engineering resources and control supplier,
                production, and service costs. Technology sharing and inheritance
                between vehicle models tends to limit the rate of improvement in a
                manufacturer's fleet.\145\
                ---------------------------------------------------------------------------
                 \145\ Toyota, NHTSA-2018-0067-12098, Attachment 1, at 3 et seq.
                 These comments urging EPA to use NHTSA's CAFE model echo comments
                provided in response to a 2018 peer review of the model. While
                identifying various opportunities for improvement, peer reviewers
                expressed strong overall support for the CAFE model's technical
                approach and execution. For example, one reviewer, after offering many
                ---------------------------------------------------------------------------
                specific technical recommendations, concluded as follows:
                 The model is impressive in its detail, and in the completeness
                of the input data that it uses. Although the model is complex, the
                reader is given a clear account of how variables are variously
                divided and combined to yield appropriate granularity and efficiency
                within the model. The model tracks well a simplified version of the
                real-world and manufacturing/design decisions. The progression of
                technology choices and cost benefit choices is clear and logical. In
                a few cases, the model simply explains a constraint, or a value
                assigned to a variable, without defending the choice of the value or
                commenting on real-world variability, but these are not substantive
                omissions. The model will lend itself well to future adaptation or
                addition of variables, technologies and pathways.\146\
                ---------------------------------------------------------------------------
                 \146\ NHTSA, CAFE Model Peer Review, DOT HS 812 590, Available
                at https://www.nhtsa.gov/document/cafe-model-peer-review, at 250.
                 Although the peer review charge focused solely on the CAFE model,
                another peer reviewer separately recommended that EPA ``consider
                opportunities for EPA to use the output from the Volpe Model in place
                of their OMEGA Model output'' \147\
                ---------------------------------------------------------------------------
                 \147\ Id. at 287-288 and 304.
                ---------------------------------------------------------------------------
                 More recently, in response to the NPRM, Dr. Julian Morris, an
                economist at George Washington University, commented extensively on the
                superiority of the agencies' NPRM analysis to previous analyses,
                offering the following overall assessment:
                 I have assessed the plausibility of the analyses undertaken by
                NHTSA and EPA in relation to the proposed SAFE rule. I found that
                the agencies have undertaken a thorough--one might even say
                exemplary--analysis, improving considerably on earlier analyses
                undertaken by the agencies of previous rules relating to CAFE
                standards and [tailpipe CO2] emission standards. Of
                particular note, the agencies included more realistic estimates of
                the rebound effect, developed a sophisticated model of the
                [[Page 24227]]
                scrappage effect, and better accounted for various factors affecting
                vehicle fatality rates.\148\
                ---------------------------------------------------------------------------
                 \148\ Morris, J., OAR-2018-0283-4028, at 6-11.
                 The agencies carefully considered these and other comments
                regarding which models to apply when estimating potential impacts of
                each of the contemplated regulatory alternatives. For purposes of
                estimating the impacts of CAFE standards, even the coalition of
                environmental advocates observed that the CAFE model reflects EPCA's
                requirements. As discussed below in Section VI.A, EPCA imposes specific
                requirements not only on how CAFE standards are to be structured (e.g.,
                including a minimum standard for domestic passenger cars), but also on
                how CAFE standards are to be evaluated (e.g., requiring that the
                potential to produce additional AFVs be set aside for the model years
                under consideration), and the CAFE model reflects these requirements,
                and the agencies consider the CAFE model to be the best available tool
                for CAFE rulemaking analysis. Regarding the use of Autonomie to
                construct fuel consumption (i.e., efficiency) inputs to the CAFE model,
                the agencies recognize that other vehicle simulation tools are
                available, including EPA's recently-developed ALPHA model. However, as
                also discussed in Section VI.B.3, Autonomie has a much longer history
                of development and refinement, and has been much more widely applied
                and validated. Moreover, Argonne experts have worked carefully for
                several years to develop methods for running large arrays of
                simulations expressly structured and calibrated for use in DOT's CAFE
                model. Therefore, the agencies consider Autonomie to be the best
                available tool for constructing such inputs to the CAFE model. While
                the agencies have also carefully considered potential specific model
                refinements, as well as the merits of potential changes to model inputs
                and assumptions, none of these potential refinements and input have led
                either agency to reconsider using the CAFE model and Autonomie for CAFE
                rulemaking analysis.
                 With respect to estimating the impacts of CO2 standards,
                even though Argonne and the agencies have adapted Autonomie and the
                CAFE model to support the analysis of CO2 standards,
                environmental groups, California, and other States would prefer that
                EPA use the models it developed during 2009-2018 for that purpose.\149\
                Arguments that EPA revert to its ALPHA and OMEGA models fall within
                three general categories: (1) Arguments that EPA's models would have
                selected what commenters consider better (i.e., generally more
                stringent) standards, (2) arguments that EPA's models are technically
                superior, and (3) arguments that the law requires EPA use its own
                models.
                ---------------------------------------------------------------------------
                 \149\ The last-finalized versions of EPA's OMEGA model and ALPHA
                tools were published in 2016 and 2017, respectively.
                ---------------------------------------------------------------------------
                 The first of these arguments--that EPA's models would have selected
                better standards--conflates the analytical tool used to inform
                decision-making with the action of making the decision. As explained
                elsewhere in this document and as made repeatedly clear over the past
                several rulemakings, the CAFE model (or, for that matter, any model)
                neither sets standards nor dictates where and how to set standards; it
                simply informs as to the potential effects of setting different levels
                of standards. In this rulemaking, EPA has made its own decisions
                regarding what CO2 standards would be appropriate under the
                CAA.
                 The third of these arguments--that EPA is legally required to use
                only models developed by its own staff--is also without merit. The CAA
                does not require the agency to create or use a specific model of its
                own creation in setting tailpipe CO2 standards. The fact
                that EPA's decision may be informed by non-EPA-created models does not,
                in any way, constitute a delegation of its statutory power to set
                standards or decision-making authority.\150\ Arguing to the contrary
                would suggest, for example, that EPA's decision would be invalid
                because it relied on EIA's Annual Energy Outlook for fuel prices for
                all of its regulatory actions rather than developing its own model for
                estimating future trends in fuel prices. Yet, all Federal agencies that
                have occasion to use forecasts of future fuel prices regularly (and
                appropriately) defer to EIA's expertise in this area and rely on EIA's
                NEMS-based analysis in the AEO, even when those same agencies are using
                EIA's forecasts to inform their own decision-making. Similarly, this
                argument would mean that the agencies could not rely on work done by
                contractors or other outside consultants, which is contrary to regular
                agency practice across the entirety of the Federal Government.
                ---------------------------------------------------------------------------
                 \150\ ``[A] federal agency may turn to an outside entity for
                advice and policy recommendations, provided the agency makes the
                final decisions itself.'' U.S. Telecom. Ass'n v. FCC, 359 F.3d 554,
                565-66 (D.C. Cir. 2004). To the extent commenters meant to suggest
                outside parties have a reliance interest in EPA using ALPHA and
                OMEGA to set standards, EPA and NHTSA do not agree a reliance
                interest is properly placed on an analytical methodology, which
                consistently evolves from rule to rule. Even if it were, all parties
                that closely examined ALPHA and OMEGA-based analyses in the past
                either also simultaneously closely examined CAFE and Autonomie-based
                analyses in the past, or were fully capable of doing so, and thus,
                should face no additional difficulty now they have only one set of
                models and inputs/outputs to examine.
                ---------------------------------------------------------------------------
                 The specific claim here that use of the CAFE model instead of ALPHA
                and OMEGA is somehow illegitimate is similarly unpersuasive. The CAFE
                and CO2 rules have, since Massachusetts v. EPA, all been
                issued as joint rulemakings, and, thus are the result of a
                collaboration between the two agencies. This was true when the
                rulemakings used separate models for the different programs and
                continues to be true in today's final rule, where the agencies take the
                next step in their collaborative approach by now using simply one model
                to simulate both programs. In 2007, immediately following this Supreme
                Court decision, the agencies worked together toward standards for model
                years 2011-2015, and EPA made use of the CAFE model for its work toward
                possible future CO2 standards. That the agencies would need
                to continue the unnecessary and inefficient process of using two
                separate combinations of models as the joint National Program continues
                to mature, therefore, runs against the idea that the agencies, over
                time, would best combine resources to create an efficient and robust
                regulatory program. For the reasons discussed throughout today's final
                rule, the agencies have jointly determined that Autonomie and the CAFE
                model have significant technical advantages, including important
                additional features, and are therefore the more appropriate models to
                use to support both analyses.
                 Further, the fact that Autonomie and CAFE models were initially
                developed by DOE/Argonne and NHTSA does not mean that EPA has no role
                in either these models or their inputs. That is, the development
                process for CAFE and CO2 standards inherently requires
                technical and policy examinations and deliberations between staff
                experts and decision-makers in both agencies. Such engagements are a
                healthy and important part of any rulemaking activity--and particularly
                so with joint rulemakings. The Supreme Court stated in Massachusetts v.
                EPA that, ``The two obligations [to set CAFE standards under EPCA and
                to set tailpipe CO2 emissions standards under the CAA] may
                overlap, but there is no reason to think the two agencies cannot both
                administer their obligations and yet
                [[Page 24228]]
                avoid inconsistency.'' \151\ When agency experts consider analytical
                issues and agency decision-makers decide on policy, which is informed
                (albeit not dictated) by the outcome of that work, they are working
                together as the Court appears to have intended in 2007, even if
                legislators' intentions have varied in the decades since EPCA and the
                CAA have been in place.\152\ Regulatory overlap necessarily involves
                deliberation, which can lead to a more balanced, reasonable, and
                improved analyses, and better regulatory outcomes. It did here. The
                existence of deliberation is not per se evidence of unreasonableness,
                even if some commenters believe a different or preferred policy outcome
                would or should have resulted.\153\
                ---------------------------------------------------------------------------
                 \151\ Massachusetts v. EPA, 549 U.S. 497, 532 (2007).
                 \152\ For example, when wide-ranging amendments to the CAA were
                being debated, S. 1630 contained provisions that, if enacted, would
                have authorized automotive CO2 emissions standards and
                prescribed specific average levels to be achieved by 1996 and 2000.
                In a letter to Senators, then-Administrator William K. Reilly noted
                that the Bill ``requires for the first time control of emissions of
                carbon dioxide; this is essentially a requirement to improve fuel
                efficiency'' and outlined four reasons the H.W. Bush Administration
                opposed the requirement, including that ``it is inappropriate to add
                this very complex issue to the Clean Air Act which is already full
                of complicated and controversial issues.'' Reilly, W., Letter to
                U.S. Senators (January 26, 1990). The CAA amendments ultimately
                signed into law did not contain these or any other provisions
                regarding regulation of CO2 emissions.
                 \153\ See, e.g., U.S. House of Representatives, Committee on
                Oversight and Government Reform, Staff Report, 112th Congress, ``A
                Dismissal of Safety, Choice, and Cost: The Obama Administration's
                New Auto Regulations,'' August 10, 2012, at 19-21 and 33-34.
                ---------------------------------------------------------------------------
                 Over the 44 years since EPCA established the requirement for CAFE
                standards, NHTSA, EPA and DOE career staff have discussed, collaborated
                on, and debated engineering, economic, and other aspects of CAFE
                regulation, through focused meetings and projects, informal exchanges,
                publications, conferences and workshops, and rulemakings.
                 Part of this expanded exchange has involved full vehicle
                simulation. While tools such as PSAT (the DOE-sponsored simulation tool
                that predated Autonomie) were in use prior to 2007, including for
                discrete engineering studies supporting inputs to CAFE rulemaking
                analyses, these tools' information and computing requirements were such
                that NHTSA had determined (and DOE and EPA had concurred) that it was
                impractical to more fully integrate full vehicle simulation into
                rulemaking analyses. Since that time, computing capabilities have
                advanced dramatically, and the agencies now agree that such integration
                of full vehicle simulation--such as the large-scale exercise of
                Autonomie to produce inputs to the CAFE Model--can make for more robust
                CAFE and CO2 rulemaking analysis. This is not to say,
                though, that experts always agree on all methods and inputs involved
                with full vehicle simulation. Differences in approach and inputs lead
                to differences in results. For example, compared to other publicly
                available tools that can be practicably exercised at the scale relevant
                to fleetwide analysis needed for CAFE and CO2 rulemaking
                analysis, DOE/Argonne's Autonomie model is more advanced, spans a wider
                range of fuel-saving technologies, and represents them in more specific
                detail, leaving fewer ``gaps'' to be filled with other models (risking
                inconsistencies and accompanying errors). These differences discussed
                in greater detail below in Section VI.B.3. Perhaps most importantly,
                the CAFE model considers fuel prices in determining both which
                technologies are applied and the total amount of technology applied, in
                the case where market forces demand fuel economy levels in excess of
                the standards. While OMEGA can apply technology in consideration of
                fuel prices, OMEGA will apply technology to reach the same level of
                fuel economy (or CO2 emissions) if fuel prices are 3, 5, or
                20 dollars, which violates the SAB's requirement that the analysis
                ``account for [. . .] future fuel prices .'' \154\ Furthermore, it
                produces a counterintuitive result. If fuel prices become exorbitantly
                high, we would expect consumers to place an emphasis on additional fuel
                efficiency as the potential for extra fuel savings is tremendous.
                ---------------------------------------------------------------------------
                 \154\ See SAB Report 10 (``Constructing each of the scenarios is
                challenging and involve extensive scientific, engineering, and
                economic uncertainties. Projecting the baseline requires the
                agencies to account for a wide range of variables including: The
                number of new vehicles sold, future fuel prices,. . . .'').
                ---------------------------------------------------------------------------
                 Moreover, DOE has for many years used Autonomie (and its precursor
                model, PSAT) to produce analysis supporting fuel economy-related
                research and development programs involving billions of dollars of
                public investment, and NHTSA's CAFE model with inputs from DOE/
                Argonne's Autonomie model has produced analysis supporting rulemaking
                under the CAA. In 2015, EPA proposed new tailpipe CO2
                standards for MY 2021-2027 heavy-duty pickups and vans, finalizing
                those standards in 2016. Supporting the NPRM and final rule, EPA relied
                on analysis implemented by NHTSA using NHTSA's CAFE model, and NHTSA
                used inputs developed by DOE/Argonne using DOE/Argonne's Autonomie
                model. CBD questioned this history, asserting that, ``EPA conducted a
                separate analysis using a different iteration of the CAFE model rather
                than rely on the version which NHTSA used, again resulting and parallel
                but corroborative modeling results.'' \155\ CBD's comment
                mischaracterizes EPA's actual use of the CAFE Model. As explained in
                the final rule, EPA's ``Method B'' analysis was developed as follows:
                ---------------------------------------------------------------------------
                 \155\ CBD, et al., 2018-0067-12000, Appendix A, at 27.
                 In Method B, the CAFE model from the NPRM was used to project a
                pathway the industry could use to comply with each regulatory
                alternative, along with resultant impacts on per-vehicle costs.
                However, the MOVES model was used to calculate corresponding changes
                in total fuel consumption and annual emissions for pickups and vans
                in Method B. Additional calculations were performed to determine
                corresponding monetized program costs and benefits.\156\
                ---------------------------------------------------------------------------
                 \156\ 81 FR 73478, 73506-07 (October 25, 2016).
                 In other words, a version of NHTSA's CAFE Model was used to perform
                the challenging part of the analysis--that is, the part that involves
                accounting for manufacturers' fleets, accounting for available fuel-
                saving technologies, accounting for standards under consideration, and
                estimating manufacturers' potential responses to new standards--EPA's
                MOVES model was used to perform ``downstream'' calculations of fuel
                consumption and tailpipe emissions, and used spreadsheets to calculate
                even more straightforward calculations of program costs and benefits.
                While some stakeholders perceive these differences as evidencing a
                meaningfully independent approach, in fact, the EPA staff's analysis
                was, at its core, wholly dependent on NHTSA's CAFE Model, and on that
                model's use of Autonomie simulations.
                 Given the above, the only remaining argument for EPA to revert to
                its previously-developed models rather than relying on Autonomie and
                the CAFE model would be that the former are so technically superior to
                the latter that even model refinements and input changes cannot lead
                Autonomie and the CAFE model to produce appropriate and reasonable
                results for CO2 rulemaking analysis. As discussed below,
                having considered a wide range of technical differences, the agencies
                find that the Autonomie and CAFE models currently provide the best
                analytical combination for CAFE and tailpipe CO2 emissions
                rulemaking analysis. As discussed
                [[Page 24229]]
                below in Section VI.B.3, Autonomie not only has a longer and wider
                history of development and application, but also DOE/Argonne's
                interaction with automakers, supplier and academies on continuous bases
                had made individual sub-models and assumptions more robust. Argonne has
                also been using research from DOE's Vehicle Technology Office (VTO) at
                the same time to make continuous improvements in Autonomie.\157\ Also,
                while Autonomie uses engine maps as inputs, and EPA developed engine
                maps that could have been used for today's analysis, EPA declined to do
                so, and those engine maps were only used in a limited capacity for
                reasons discussed below in Section VI.C.1.
                ---------------------------------------------------------------------------
                 \157\ U.S. DOE Benefits & Scenario Analysis publications is
                available at https://www.autonomie.net/publications/fuel_economy_report.html. Last accessed November 14, 2019.
                ---------------------------------------------------------------------------
                 As also discussed below in Section VI.A.4, the CAFE model accounts
                for some important CO2 provisions that EPA's OMEGA model
                cannot account for. For example, the CAFE model estimates the potential
                that any given manufacturer might apply CO2 compliance
                credits it has carried forward from some prior model year. While one
                commenter, Mr. Rykowski, takes issue with how the CAFE model handles
                credit banking, he does not acknowledge that EPA's OMEGA model, lacking
                a year-by-year representation of compliance, is altogether incapable of
                accounting for the earning and use of banked compliance credits. Also,
                although Mr. Rykowski's comments regarding A/C leakage and refrigerants
                are partially correct insofar as the CAFE model does not account for
                leakage-reducing technologies explicitly, the comment is as applicable
                to OMEGA as it is to the CAFE model and, in any event, data regarding
                which vehicles have which leakage-reducing technologies was not
                available for the MY 2016 fleet. Nevertheless, as discussed in Section
                VI.A.4, NHTSA has refined the CAFE model's accounting for the cost of
                leakage reduction technologies.
                 The agencies have also considered Mr. Rykowski's comments
                suggesting that using OMEGA would be preferable because, rather than
                selecting from hundreds of thousands of potential combinations of
                technologies, OMEGA includes only the ``50 or so'' combinations that
                EPA has already determined to be cost-effective. The ``better way'' of
                making this determination is also effectively a model, but the
                separation of this model from OMEGA is, as evidenced by manufacturers'
                comments, obfuscatory, especially in terms of revealing how specific
                vehicle model/configurations initial engineering properties are aligned
                with specific initial technology combinations. By using a full set of
                technology combinations, the CAFE model makes very clear how each
                vehicle model/configuration is assigned to a specific initial
                combination and, hence, how subsequently fuel consumption and cost
                changes are accounted for. Moreover, EPA's separation of ``thinning''
                process from OMEGA's main compliance simulation makes sensitivity
                analysis difficult to implement, much less follow. The agencies find,
                therefore, that the CAFE model's approach of retaining a full set of
                vehicle simulation results throughout the compliance simulation to be
                more realistic (e.g., more capable of reflecting manufacturer- and
                vehicle-specific factors), more responsive to changes in model inputs
                (e.g., changes to fuel prices, which could impact the relative
                attractiveness of different technologies), more transparent, and more
                amenable to independent corroboration the agencies' analysis.
                 Regarding comments by Messrs. Duleep, Rogers, and Rykowski
                suggesting that the CAFE model, by tying most technology application to
                planned vehicle redesigns and freshening, is too restrictive, the
                agencies disagree. As illustrated by manufacturers' comments cited
                above, as reinforced by both extensive product planning information
                provided to the agencies, and as further reinforced by extensive
                publicly available information, manufacturers tend to not make major
                changes to a specific vehicle model/configuration in one model year,
                and then make further major changes to the same vehicle model/
                configuration the next model year. There is ample evidence that
                manufacturers strive to avoid such discontinuity, complexity, and
                waste, and in the agencies' view, while it is impossible to represent
                every manufacturer's decision-making process precisely and with
                certainty, the CAFE model's approach of using estimated product design
                schedules provides a realistic basis for estimating what manufacturers
                could practicably do. Also, the relevant inputs are simply inputs to
                the CAFE model, and if it is actually more realistic to assume that a
                manufacturer can change major technology on all of its products every
                year, the CAFE model can easily be operated with every model year
                designated as a redesign year for every product, but as discussed
                throughout this document, the agencies consider this to be extremely
                unrealistic. While this means the CAFE model can be run without a year-
                by-year representation that carries forward technologies between model
                years, doing so would be patently unrealistic (as reflected in some
                stakeholders' comments in 2002 on the first version of the CAFE model).
                Conversely, the OMEGA model cannot be operated in a way that accounts
                for what the agencies consider to be very real product planning
                considerations.
                 However, having also considered Mr. Rykowski's comments about the
                CAFE model's ``effective cost'' metric, and having conducted side-by-
                side testing documented in the accompanying FRIA, the agencies are
                satisfied that an alternative ``cost per credit'' metric is also a
                reasonable metric to use for estimating how manufacturers might
                selected among available options to add specific fuel-saving
                technologies to specific vehicles.\158\ Therefore, NHTSA has revised
                the CAFE model accordingly, as discussed below in Section VI.A.4.
                ---------------------------------------------------------------------------
                 \158\ As discussed in the FRIA, results vary with model inputs,
                among manufacturers, and across model years, but compared to the
                NPRM's ``effective cost'' metric, the ``cost per credit'' metric
                appears to more frequently produce less expensive solutions than
                more expensive solutions, at least when simulating compliance with
                CO2 standards. Differences are more mixed when simulating
                compliance with CAFE standards, and even when simulating compliance
                with CO2 standards, results simulating ``perfect'' trading of
                CO2 compliance credits are less intuitive when the ``cost
                per credit metric.'' Nevertheless, and while less expensive
                solutions are not necessarily ``optimal'' solutions (e.g., if
                gasoline costs $7 per gallon and electricity is free, expensive
                electrification could be optimal), the agencies consider it
                reasonable to apply the ``cost per credit'' metric for the analysis
                supporting today's rulemaking.
                ---------------------------------------------------------------------------
                 Section VI.C.1 also addresses Mr. Rogers's comments on engine maps
                used as estimates to full vehicle simulation. In any event, because
                engine maps are inputs to full vehicle modeling and simulation, the
                relative merits of specific maps provide no basis to prefer one vehicle
                simulation modeling system over another. Similarly, Section VI.B.3 also
                addresses Mr. Duleep's comments preferring EPA's prior approach, using
                ALPHA, of effectively assuming that a manufacturer would incur no
                additional cost by reoptimizing every powertrain to extract the full
                fuel economy potential of even the smallest incremental changes to
                aerodynamic drag and tire rolling resistance. Mr. Duleep implies that
                Autonomie is flawed because the NPRM analysis did not apply Autonomie
                in a way that makes such assumptions. The agencies discuss powertrain
                sizing and calibration in Section VI.B.3, and note here that such
                assumptions are not inherent to
                [[Page 24230]]
                Autonomie; like engine maps, these are inputs to full vehicle
                simulation. Therefore, neither of these comments by Mr. Rogers and Mr.
                Duleep lead the agencies to find reason not to use Autonomie.
                 None of this is to say that Autonomie and the CAFE model as
                developed and applied for the NPRM left no room for improvement. In the
                NPRM and RIA, the agencies discussed plans to continue work in a range
                of specific technical areas, and invited comment on all aspects of the
                analysis. As discussed below in Chapter VI, the agencies received
                extensive comment on the published model, inputs, and analysis, both in
                response to the NPRM and, for newly-introduced modeling capabilities
                (estimation of sales, scrappage, and employment effects), in response
                to additional peer review conducted in 2019. The agencies have
                carefully considered these comments, refined various specific technical
                aspects of the CAFE model (like the ``effective cost'' metric mentioned
                above), and have also updated inputs to both Autonomie and the CAFE
                model. Especially given these refinements and updates, as discussed
                throughout this rule, EPA maintains that for CO2 rulemaking
                analysis, Autonomie and the CAFE model have advantages that warrant
                relying on them rather than on EPA's ALPHA and OMEGA models. Some
                examples of such advantages include: A longer history of ongong
                development and application for rulemaking, including by EPA;
                documentation and model operation stakeholders have found to be
                comparatively clear and enabling of independent replication of agency
                analyses; a mechanism to explicitly reflect the fact that
                manufacturers' product decisions are likely to be informed by fuel
                prices; better integration of various model functions, enabling
                efficient sensitivity analysis; and an annual time step that makes it
                possible to conduct report results on both a calendar year and model
                year basis, to estimate accruing impacts on vehicle sales and
                scrappage, and to account for the fact that not every vehicle can be
                designed in every model year; and other advantages discussed throughout
                today's notice. Therefore, recognizing that models inform but do not
                make regulatory decisions, EPA has elected to rely solely on the
                Autonomie and CAFE models to produce today's analysis of regulatory
                alternatives for CO2 standards.
                 The following sections provide a brief technical overview of the
                CAFE model, including changes NHTSA made to the model since 2012, and
                differences between the current analysis, the analysis for the 2016
                Draft TAR and for the 2017 Proposed Determination/2018 Final
                Determination, and the 2018 NPRM, before discussing inputs to the model
                and then diving more deeply into how the model works. For more
                information on the latter topic, see the CAFE model documentation,
                available in the docket for this rulemaking and on NHTSA's website.
                1. What assumptions have changed since the 2012 final rule?
                 Any analysis of regulatory actions that will be implemented several
                years in the future, and whose benefits and costs accrue over decades,
                requires a large number of assumptions. Over such time horizons, many,
                if not most, of the relevant assumptions in such an analysis are
                inevitably uncertain.\159\ The 2012 CAFE/CO2 rule considered
                regulatory alternatives for model years through MY 2025 (17 model years
                after the 2008 market information that formed the basis of the
                analysis) that accrued costs and benefits into the 2060s. Not only was
                the new vehicle market in 2025 unlikely to resemble the market in 2008,
                but so, too, were fuel prices. It is natural, then, that each
                successive CAFE/CO2 analysis should update assumptions to
                reflect better the current state of the world and the best current
                estimates of future conditions.\160\ However, beyond the issue of
                unreliable projections about the future, a number of agency assertions
                have proven similarly problematic. In fact, Securing America's Future
                Energy (SAFE) stated in their comments on the NPRM:
                ---------------------------------------------------------------------------
                 \159\ As often stated, ``It's difficult to make predictions,
                especially about the future.'' See, e.g., https://quoteinvestigator.com/2013/10/20/no-predict/.
                 \160\ See, e.g., 77 FR 62785 (Oct. 15, 2012) (``If EPA initiates
                a rulemaking [to revise standards for MYs 2022-2025], it will be a
                joint rulemaking with NHTSA. . . . NHTSA's development of its
                proposal in that later rulemaking will include the making of
                economic and technology analyses and estimates that are appropriate
                for those model years and based on then-current information.'').
                 Although the agencies argue ``circumstances have changed'' and
                ``analytical methods and inputs have been updated,'' a thorough
                analysis should provide a side-by-side comparison of the changing
                circumstances, methods, and inputs used to arrive at this
                determination . . . They represent a rapid, dramatic departure from
                the agencies' previous analyses, without time for careful review and
                consideration.\161\
                ---------------------------------------------------------------------------
                 \161\ Securing America's Energy Future, NHTSA-2018-0067-12172,
                at 39.
                 We describe in detail (below) the changes to critical assumptions,
                perspectives, and modeling techniques that have created substantive
                differences between the current analysis and the analysis conducted in
                2012 to support the final rule. To the greatest extent possible, we
                have calculated the impacts of these changes on the 2012 analysis.
                a) The Value of Fuel Savings
                 The value of fuel savings associated with the preferred alternative
                in the 2012 final rule is primarily a consequence of two assumptions:
                \162\ The fuel price forecast and the assumed growth in fuel economy in
                the baseline alternative against which savings are measured. Therefore,
                as the value of fuel savings accounted for nearly 80 percent of the
                total benefits of the 2012 rule, each of these assumptions is
                consequential. With a lower fuel price projection and an expectation
                that new vehicle buyers respond to fuel prices, the 2012 rule would
                have shown much smaller fuel savings attributable to the more stringent
                standards. Projected fuel prices are considerably lower today than in
                2012, the agencies now understand new vehicle buyers to be at least
                somewhat responsive to fuel prices, and the agencies have therefore
                updated corresponding model inputs to produce an analysis the agencies
                consider to be more realistic.
                ---------------------------------------------------------------------------
                 \162\ The value of fuel savings is also affected by the rebound
                effect assumption, assumed lifetime VMT accumulation, and the
                simulated penetration of alternative fuel technologies. However,
                each of these ancillary factors is small compared to the impact of
                the two factors discussed in this subsection.
                ---------------------------------------------------------------------------
                 The first of these assumptions, fuel prices, was simply an artifact
                of the timing of the rule. Following recent periodic spikes in the
                national average gasoline price and continued volatility after the
                great recession, the fuel price forecast then produced by EIA (as part
                of AEO 2011) showed a steady march toward historically high, sustained
                gasoline prices in the United States. However, the actual series of
                fuel prices has skewed much lower. As it has turned out, the observed
                fuel price in the years between the 2012 final rule and this rule has
                frequently been lower than the ``Low Oil Price'' sensitivity case in
                the 2011 AEO, even when adjusted for inflation. The following graph
                compares fuel prices underlying the 2012 final rule to fuel prices
                applied in the analysis reported in today's notice, expressing both
                projections in 2010 dollars. The differences are clear and significant:
                [[Page 24231]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.042
                 The discrepancy in fuel prices is important to the discussion of
                differences between the current rule and the 2012 final rule, because
                that discrepancy leads in turn to differences in analytical outputs and
                thus to differences in what the agencies consider in assessing what
                levels of standards are reasonable, appropriate, and/or maximum
                feasible. As an example, the agencies discuss in Sections VI.D.3
                Simulating Environmental Impacts of Regulatory Alternatives and
                VIII.A.3 EPA's Conclusion that the Final CO2 Standards are
                Appropriate and Reasonable that fuel price projections from the 2012
                rule were one assumption, among others, that could have led to
                overestimates of the health benefits that resulted from reducing
                criteria pollutant emissions. Yet the agencies caution readers not to
                interpret this discrepancy as a reflection of negligence on the part of
                the agencies, or on the part of EIA. Long-term predictions are
                challenging and the fuel price projections in the 2012 rule were within
                the range of conventional wisdom at the time. However, it does suggest
                that fuel economy and tailpipe CO2 regulations set almost
                two decades into the future are vulnerable to surprises, in some ways,
                and reinforces the value of being able to adjust course when critical
                assumptions are proven inaccurate. This value was codified in
                regulation when EPA bound itself to the mid-term evaluation process as
                part of the 2012 final rule.\163\
                ---------------------------------------------------------------------------
                 \163\ See 40 CFR 86-1818-12(h).
                ---------------------------------------------------------------------------
                 To illustrate this point clearly, substituting the current (and
                observed) fuel price forecast for the forecast used in the 2012 final
                rule creates a significant difference in the value of fuel savings.
                Even under identical discounting methods (see Section 2, below), and
                otherwise identical inputs in the 2012 version of the CAFE Model, the
                current (and historical) fuel price forecast reduces the value of fuel
                savings by $150 billion--from $525 billion to $375 billion (in 2009
                dollars).
                 The second assumption employed in the 2012 (as well as the 2010)
                final rule, that new vehicle fuel economy never improves unless
                manufacturers are required to increase fuel economy in the new vehicle
                market by increasingly stringent regulations, is more problematic.
                Despite the extensive set of recent academic studies showing, as
                discussed in Section VI.D.1.a)(2), that consumers value at least some
                portion, and in some studies nearly all, of the potential fuel savings
                from higher levels of fuel economy at the time they purchase vehicles,
                the agencies assumed in past rulemakings that buyers of new vehicles
                would never purchase, and manufacturers would never supply, vehicles
                with higher fuel economy than those in the baseline (MY 2016 in the
                2012 analysis), regardless of technology cost or prevailing fuel prices
                in future model years. In calendar year 2025, the 2012 final rule
                assumed gasoline would cost nearly $4.50/gallon in today's dollars, and
                continue to rise in subsequent years. Even recognizing that higher
                levels of fuel economy would be achieved under the augural/existing
                standards than without them, the assertion that fuel economy and
                CO2 emissions would not improve beyond 2016 levels in the
                presence of nearly $5/gallon gasoline is not supportable. This is
                highlighted by the observed increased consumer demand for higher-fuel-
                economy vehicles during the gas price spike of 2008, when average U.S.
                prices briefly broke $4/gallon. In the 2012 final rule, this
                assumption--that fuel economy and emissions would never improve absent
                regulation--created a persistent gap in fuel economy between
                [[Page 24232]]
                the baseline and augural standards that grew to 13 mpg (at the industry
                average, across all vehicles) by MY 2025. In the 2016 Draft TAR,
                NHTSA's analysis included the assumption that manufacturers would
                deploy, and consumers would demand, any technology that recovered its
                own cost in the first year of ownership through avoided fuel costs.
                However, in both the Draft TAR and the Proposed and Final Determination
                documents, EPA's analysis assumed that the fuel economy levels achieved
                to reach compliance with MY 2021 standards would persist indefinitely,
                regardless of fuel prices or technology costs.
                 By substituting the conservative assumption that consumers are
                willing to purchase fuel economy improvements that pay for themselves
                with avoided fuel expenditures over the first 2.5 years \164\
                (identical to the assumption in this final rule's central analysis) the
                gap in industry average fuel economy between the baseline and augural
                scenarios narrows from 13 mpg in 2025 to 6 mpg in 2025. As a corollary,
                acknowledging that fuel economy would continue to improve in the
                baseline under the fuel price forecast used in the final rule erodes
                the value of fuel savings attributable to the preferred alternative.
                While each gallon is still worth as much as was assumed in 2012, fewer
                gallons are consumed in the baseline due to higher fuel economy levels
                in new vehicles. In particular, the number of gallons saved by the
                preferred alternative selected in 2012 drops from about 180 billion to
                50 billion once we acknowledge the existence of even a moderate market
                for fuel economy.\165\ The value of fuel savings is similarly eroded,
                as higher fuel prices lead to correspondingly higher demand for fuel
                economy even in the baseline--reducing the value of fuel savings from
                $525 billion to $190 billion.
                ---------------------------------------------------------------------------
                 \164\ Greene, D.L. and Welch, J.G., ``Impacts of fuel economy
                improvements on the distribution of income in the U.S.,'' Energy
                Policy, Volume 122, November 2018, pp. 528-41 (``Four nationwide
                random sample surveys conducted between May 2004 and January 2013
                produced payback period estimates of approximately three years,
                consistent with the manufacturers' perceptions.'') (The 2018 article
                succeeds Greene and Welch's 2017 publication titled ``The Impact of
                Increased Fuel Economy for Light-Duty Vehicles on the Distribution
                of Income in the U.S.: A Retrospective and Prospective Analysis,''
                Howard H. Baker Jr. Center for Public Policy, March 2017, which
                Consumers Union, CFA, and ACEEE comments include as Attachment 4,
                Docket NHTSA-2018-0067-11731).
                 \165\ Readers should note that this is not an estimate of the
                total amount of fuel that will be consumed or not consumed by the
                fleet as a whole, but simply the amount of fuel that will be
                consumed or not consumed as a direct result of the regulation. As
                illustrated in Section VII, light-duty vehicles in the U.S. would
                continue to consume considerable quantities of fuel and emit
                considerable quantities of CO2 even under the baseline/
                augural standards, and agencies' analysis shows that the standards
                finalized today will likely increase fuel consumption and
                CO2 emissions by a small amount.
                ---------------------------------------------------------------------------
                 The magnitude of the fuel economy improvement in the baseline is a
                consequence of both the fuel prices assumed in the 2012 rule (already
                discussed as being higher than both subsequent observed prices and
                current projections) and the assumed technology costs. In 2012, a
                number of technologies were assumed to have negative incremental
                costs--meaning that applying those technologies to existing vehicles
                would both improve their fuel economy and reduce the cost to produce
                them. Asserting that the baseline would experience no improvement in
                fuel economy without regulation is equivalent to asserting that
                manufacturers, despite their status as profit maximizing entities,
                would not apply these cost-saving technologies unless forced to do so
                by regulation. While this issue is discussed in greater detail in
                Section VI.B the combination of inexpensive (or free) technology and
                high fuel prices created a logically inconsistent perspective in the
                2012 rule--where consumers never demanded additional fuel economy,
                despite high fuel costs, and manufacturers never supplied additional
                fuel economy, despite the availability of inexpensive (or cost saving)
                technology to do so.
                 Many commenters on earlier rules supported the assumption that fuel
                economy would not improve at all in the absence of standards. In fact,
                some commenters still support this position. For example, EDF commented
                to the NPRM that, ``NHTSA set the Volpe model to project that, with
                CAFE standards remaining flat at MY 2020 levels through MY 2026,
                automakers would over-comply with the MY 2020 standards by 9 grams/mile
                of CO2 for cars and 15 g/mi of CO2 for light
                trucks during the 2029-2032 timeframe, plus 1%/year improvements beyond
                MY 2032. This assumption unreasonably obscures the impacts of the
                rollback and is not reflective of historical compliance performance.''
                \166\
                ---------------------------------------------------------------------------
                 \166\ EDF, NHTSA-2018-0067-11996, Comments to DEIS, at 4.
                ---------------------------------------------------------------------------
                 EDF is mistaken in two different ways: (1) By acknowledging the
                existence of a well-documented market for fuel economy, rather than
                erroneously inflating the benefits associated with increasing
                standards, this assumption serves to isolate the benefits actually
                attributable to each regulatory alternative, and (2) it is, indeed,
                reflective of historical compliance performance. While the agencies
                rely on the academic literature (and comments from companies that build
                and sell automobiles) to defend the assertion that a market for fuel
                economy exists, the industry's historical CAFE compliance performance
                is a matter of public record.\167\ As shown in Figure IV-3, Figure IV-
                4, and Figure IV-5 for more than a decade, the industry average CAFE
                has exceeded the standard for each regulatory class--by several mpg
                during periods of high fuel prices.
                ---------------------------------------------------------------------------
                 \167\ Data from CAFE Public Information Center (PIC), https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Home.htm, last accessed 10/08/2019.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24233]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.043
                [[Page 24234]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.044
                BILLING CODE 4910-59-C
                 While this rulemaking has shown the impact of deviations from the
                2012 rule assumptions individually, these two assumptions affect the
                value of fuel savings jointly. Replacing the fuel price forecast with
                the observed historical and current projected prices, and including any
                technology that pays for itself in the first 2.5 years of ownership
                through avoided fuel expenditures, reduces the value of fuel savings
                from $525 billion in the 2012 rule to $140 billion, all else equal.
                Interestingly, this reduction in the value of fuel savings is smaller
                than the result when assuming only that the desired payback period is
                nonzero. While it may seem counterintuitive, it is entirely consistent.
                 The number of gallons saved under the preferred alternative is
                actually higher when modifying both assumptions, compared to only
                modifying the payback period. Updating both assumptions leads to about
                100 billion gallons saved by the preferred alternative in 2012,
                compared to only 50 billion from changing only the payback period, and
                180 billion in the 2012 analysis. This occurs because the fuel economy
                in the baseline is lower when updating both the fuel price and the
                payback period--the gap between the augural standards and the baseline
                grows to 9 mpg, rather than only 6 mpg when updating only the payback
                period. Despite the existence of inexpensive
                [[Page 24235]]
                technology in both cases, with lower fuel prices there are fewer
                opportunities to apply technology that will pay back quickly. As a
                consequence, the number of gallons saved by the preferred alternative
                in 2012 increases--but each gallon saved is worth less because the
                price of fuel is lower.
                b) Technology Cost
                 While the methods used to identify cost-effective technologies to
                improve fuel economy in new vehicles have continuously evolved since
                2012 (as discussed further in Section IV.B.1), as have the estimated
                cost of individual technologies, the inclusion of a market response in
                all scenarios (including the baseline) has changed the total technology
                cost associated with a given alternative. As also discussed in Section
                VI.B, acknowledging the existence of a market for fuel economy leads to
                continued application of the most cost-effective technologies in the
                baseline--and in other less stringent alternatives--up to the point at
                which there are no remaining technologies whose cost is fully offset by
                the value of fuel saved in the first 30 months of ownership. The
                application of this market-driven technology has implications for fuel
                economy levels under lower stringencies (as discussed earlier), but
                also for the incremental technology cost associated with more stringent
                alternatives. As lower stringency alternatives (including the 2012
                baseline) accrue more technology, the incremental cost of more
                stringent alternatives decreases.
                 By including a modest market for fuel economy, and preserving all
                other assumptions from the 2012 final rule, the incremental cost of
                technology attributable to the preferred alternative decreases from
                about $140 billion to about $72 billion. This significant reduction in
                technology cost is somewhat diminished by the associated reduction in
                the value of fuel savings (a decrease of $385 billion) when
                acknowledging the existence of a market for fuel economy. Another
                consequence of these changes is that the incremental cost of fuel
                economy technology is responsive to fuel price, as it should be. Under
                higher prices (as were assumed in 2012), consumers demand higher fuel
                economy in the new vehicle market. Under lower prices (as have occurred
                since the 2012 rule) consumers demand less fuel economy than would have
                been consistent with the fuel price assumptions in 2012.\168\ Including
                a market response in the analysis ensures that, in each case, the cost
                of fuel economy technology within an alternative is consistent with
                those assumptions. Using the same fuel price forecast that supports
                this rule, and the same estimate of market demand for fuel economy, the
                incremental cost of technology in the preferred alternative would rise
                back up to about $110 billion.
                ---------------------------------------------------------------------------
                 \168\ This is why dozens of studies examining the ability of
                fuel taxes (and carbon taxes, which produce the same result for
                transportation fuels) to reduce CO2 emissions have found
                cost-effective opportunities available for those pricing mechanisms.
                ---------------------------------------------------------------------------
                c) The Social Cost of Carbon (SCC) Emissions
                 As discussed extensively in the NPRM, the agencies' perspective
                regarding the social cost of carbon has narrowed in focus. While the
                2012 final rule considered the net present value of global damages
                resulting from carbon emitted by vehicles sold in the U.S. between MY
                2009 and MY 2025, the NPRM (and this final rule) consider only those
                damages that occur to the United States and U.S. territories. As a
                result of this change in perspective, the value of estimated damages
                per-ton of carbon is correspondingly smaller. Had the 2012 final rule
                utilized the same perspective on the social cost of carbon, the
                benefits associated with the preferred alternative would have been
                about $11 billion, rather than $53 billion. However, the savings
                associated with carbon damages are a consequence of both the assumed
                cost per-ton of damages and the number of gallons of fuel saved. As
                discussed above, the gallons saved in the 2012 final rule were likely
                inflated as a result of both fuel price forecasts and the assumption
                that no market exists for fuel economy improvements. Correcting the
                estimate of gallons saved from the preferred alternative in the 2012
                rule and considering only the domestic social cost of carbon further
                reduces the savings in carbon damages to $6 billion.
                d) Safety Neutrality
                 In the 2012 final rule, the agencies showed a ``safety neutral''
                compliance solution; that is, a compliance solution that produced no
                net increase in on-road fatalities for MYs 2017-2025 vehicles as a
                result of technology changes associated with the preferred alternative.
                In practice, safety neutrality was achieved by expressly limiting the
                availability of mass reduction technology to only those vehicles whose
                usage causes fewer fatalities with decreased mass. This result was
                discussed as one possible solution, where manufacturers chose
                technology solutions that limited the amount of mass reduction applied,
                and concentrated the application on vehicles that improve the safety of
                other vehicles on the roads (primarily by reducing the mass
                differential in collisions). However, it implicitly assumed that each
                and every manufacturer would leave cost-effective technologies unused
                on entire market segments of vehicles in order to preserve a safety
                neutral outcome at the fleet level for a given model year (or set of
                model years) whose useful lives stretched out as far as the 2060s.
                Removing these restrictions tells a different story.
                 When mass reduction technology, determined in the model to be a
                cost-effective solution (particularly in later model years, when more
                advanced levels of mass reduction were expected to be possible), is
                unrestricted in its application, the 2012 version of the CAFE Model
                chooses to apply it to vehicles in all segments. This has a small
                effect on technology costs, increasing compliance costs in the earliest
                years of the program by a couple billion dollars, and reducing
                compliance costs for MYs 2022--2025 by a couple billion dollars.
                However, the impact on safety outcomes is more pronounced.
                 Also starting with the model and inputs used for the 2012 final
                rule (and, as an example, focusing on that rule's 2008-based market
                forecast), removing the restrictions on the application of mass
                reduction technology results in an additional 3,400 fatalities over the
                full lives of MYs 2009-2025 vehicles in the baseline,\169\ and another
                6,900 fatalities over those same vehicle lives under the preferred
                alternative. The result, a net increase of 3,500 fatalities under the
                preferred alternative relative to the baseline, also produces a net
                social cost of $18 billion. The agencies' current treatment of both
                mass reduction technology, which can greatly improve the effectiveness
                of certain technology packages by reducing road load, and estimated
                fatalities and now account for both general exposure (omitted in the
                2012 final rule modeling) and fatality risk by age of the vehicle,
                further changes the story around mass reduction technology application
                for compliance and its relationship to on-road safety.
                ---------------------------------------------------------------------------
                 \169\ Relative to the continuation of vehicle mass from the 2008
                model year carried forward into the future.
                ---------------------------------------------------------------------------
                2. What methods have changed since the 2012 final rule?
                 Simulating how manufacturers might respond to CAFE/CO2
                standards
                [[Page 24236]]
                requires information about existing products being offered for sale, as
                well as information about the costs and effectiveness of technologies
                that could be applied to those vehicles to bring the fleets in which
                they reside into compliance with a given set of standards. Following
                extensive additional work and consideration since the 2012 analysis,
                both agencies now use the CAFE Model to simulate these compliance
                decisions. This has several practical implications which are discussed
                in greater detail in Section VI.A. Briefly, this change represents a
                shift toward including a number of real-world production constraints--
                such as component sharing across a manufacturer's portfolio--and
                product cadence, where only a subset of vehicles in a given model year
                are redesigned (and thus eligible to receive fuel economy technology).
                Furthermore, the year-by-year accounting ensures a continuous evolution
                of a manufacturer's product portfolio that begins with the market data
                of an initial model year (model year 2017, in this analysis) and
                continues through the last year for which compliance is simulated.
                Finally, the modeling approach has migrated from one that relied on the
                simple product of single values to estimate technology effectiveness to
                a model that relies on full vehicle simulation to determine the
                effectiveness of any combination of fuel economy technologies. The
                combination of these changes has greatly improved the realism of
                simulated vehicle fuel economy for combinations of technologies across
                vehicle systems and classes.
                 In addition to these changes to the portions of the analysis that
                represent the supply of fuel economy (by manufacturer, fleet, and model
                year) in the new vehicle market, this analysis contains changes to the
                representation of consumer demand for fuel economy. One such measure
                was discussed above--the notion that consumers will demand some amount
                of fuel economy improvement over time, consistent with technology costs
                and fuel prices. However, another deviation from the 2012 final rule
                analysis reflects overall demand for new vehicles. Across ten
                alternatives, ranging from the baseline (freezing future standards at
                2016 levels) to scenarios that increased stringency by seven percent
                per year, from 2017 through 2025, the 2012 analysis showed no response
                in new vehicle sales, down to the individual model level. This implied
                that, regardless of changes to vehicle cost or attributes driven by
                stringency increases, no fewer (or possibly more) units of any single
                model would be sold in any year, in any alternative. Essentially, that
                analysis asserted that the new vehicle market does not respond, in any
                way, to average new vehicle prices across the alternatives--regardless
                of whether the incremental cost is $1,600/vehicle (as it was estimated
                to be under the preferred alternative) or nearly $4,000/vehicle (as it
                was in under the 7 percent alternative). Both the NPRM and this final
                rule, while not employing pricing models or full consumer choice models
                to address differentiated demand within brands or manufacturer
                portfolios, have incorporated a modeled sales response that seeks to
                quantify what was not quantified in previous rulemakings.
                 An important accounting method has also changed since the 2012
                final rule was published. At the time of that rule, the agencies used
                an approach to discounting that combined attributes of a private
                perspective and a social perspective in their respective benefit cost
                analyses. This approach was logically inconsistent, and further
                reinforced some of the exaggerated estimates of fuel savings, social
                benefits (from reduced externalities), and technology costs described
                above. The old method discounted the value of all incremental
                quantities, whether categorized as benefits or costs, to the model year
                of the vehicle to which they accrued. This approach is largely
                acceptable for use in a private benefit cost analysis, where the costs
                and benefits accrue to the buyer of a new vehicle (in the case of this
                policy) who weighs their discounted present values at the time of
                purchase. However, the private perspective would not include any costs
                or benefits that are external to the buyer (e.g., congestion or the
                social cost of carbon emissions). For an analysis that compares
                benefits and costs from the social perspective, attempting to estimate
                the relative value of a policy to all of society rather than just
                buyers of new vehicles, this approach is more problematic.
                 The discounting approach in the 2012 final rule was particularly
                distortionary for a few reasons. The fact that benefits and costs
                occurred over long time periods in the 2012 rule, and the standards
                isolated the most aggressive stringency increases in the latter years
                of the program, served to allow benefits that occurred in 2025 (for
                example) to enter the accounting without being discounted, provided
                that they accrued to the affected vehicles during their first year of
                ownership. In a setting where numerous inputs (e.g., fuel price and
                social cost of carbon) increase over time, benefits were able to grow
                faster than the discount rate in some cases--essentially making them
                infinite. The interpretation of discounting for externalities was
                equally problematic. For example, the discounting approach in the 2012
                final rule would have counted a ton of CO2 not emitted in CY
                2025 in multiple ways, despite the fact that the social cost of carbon
                emissions was inherently tied to the calendar year in which the
                emissions occurred. Were the ton avoided by a MY 2020 vehicle, which
                would have been five years old in CY 2025, the value of that ton would
                have been the social cost of carbon times 0.86, but would have been
                undiscounted if that same ton had been saved by a MY 2025 vehicle in
                its initial year of usage.
                 This approach was initially updated in the 2016 Draft TAR to be
                consistent with common economic practice for benefit-cost analysis, and
                this analysis continues that approach. In the social perspective, all
                benefits and costs are discounted back to the decision year based on
                the calendar year in which they occur. Had the agencies utilized such
                an approach in the 2012 final rule, net benefits would have been
                reduced by about 20 percent, from $465 billion to $374 billion--not
                accounting for any of the other adjustments discussed above.
                3. How have conditions changed since the 2012 final rule was published?
                 The 2012 final rule relied on market and compliance information
                from model year 2008 to establish standards for model years 2017-2025.
                However, in the intervening years, both the market and the industry's
                compliance positions have evolved. The industry has undergone a
                significant degree of change since the MY 2008 fleet on which the
                2012FR was based. Entire brands (Pontiac, Oldsmobile, Saturn, Hummer,
                Mercury, etc.) and companies (Saab, Suzuki, Lotus) have exited the U.S.
                market, while others (most notably Tesla) have emerged. Several dozen
                nameplates have been retired and dozens of other created in that time.
                Overall, the industry has offered a diverse set of vehicle models that
                have generally higher fuel economy than the prior generation, and an
                ever-increasing set of alternative fuel powertrains.
                 As Table IV-1 shows, alternative powertrains have steadily
                increased under CAFE/CO2 regulations. Under the standards
                between 2011 and 2018, the number of electric vehicle offerings in the
                market has increased from 1 model to 57 models (inclusive of all plug-
                in vehicles that are rated for use on the highway), and hybrids (like
                the Toyota Prius) have increased from 20 models to
                [[Page 24237]]
                43 models based on data from DOE's Alternative Fuels Data Center. Fuel
                efficient diesel vehicles have similarly been on the rise in that
                period, more than doubling the number of offerings. Flexible fuel
                vehicles (FFVs), capable of operating on both gasoline and E85 remain
                readily available in the market, but have been excluded from the table
                due to both their lower fuel economy and demonstrated consumer
                reluctance to operate FFVs on E85. They have historically been used to
                improve a manufacturer's compliance position, rather than other
                alternative fuel systems that reduce fuel consumption and save buyers
                money.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.045
                 Not only have alternative powertrain options proliferated since the
                2012 FR, the average fuel economy of new vehicles within each body
                style has increased. However, the more dramatic effect may lie in the
                range of fuel economies available within each body style. Figure IV-6
                shows the distribution of new vehicle fuel economy (in miles per gallon
                equivalent) by body style for MYs 2008, 2016, and 2020 (simulated).
                Each box represents the 25th and 75th percentiles, where 25 and 75
                percent of new models offered are less fuel efficient than that level.
                Not only has the median fuel economy improved (the median shows the
                point at which 50 percent of new models are less efficient) under the
                CAFE/CO2 programs, but the range of available fuel economies
                (determined by the length of the boxes and their whiskers) has
                increased as well. For example, the 25th percentile of pickup truck
                fuel economy in 2020 is expected to be significantly more efficient
                than 75 percent of the pickups offered in 2008. In MY 2008, there were
                only a few SUVs offered with rated fuel economies above 34MPG. By MY
                2020 almost half of the SUVs offered will have higher fuel economy
                ratings--with almost 20 percent of offerings exceeding 40MPG.
                 The improvement in passenger car styles has been no less dramatic.
                As with the other styles, the range of available fuel economies has
                increased under the CAFE/CO2 programs and the distribution
                of available fuel economies skewed higher--with 40 percent of MY 2020
                models exceeding 40MPG. The attribute-based standards are designed to
                encourage manufacturers to improve vehicle fuel economy across their
                portfolios, and they have clearly done so. Not only have the higher
                ends of the distributions increased, the lower ends in all body styles
                have improved as well, where the least fuel efficient 25 percent of
                vehicles offered in MY 2016 (and simulated in 2020) are more fuel
                efficient than the most efficient 25 percent of vehicles offered in MY
                2008.
                BILLING CODE 4910-59-P
                [[Page 24238]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.046
                BILLING CODE 4910-59-C
                 Some commenters have argued that consumers will be harmed by any
                set of standards lower than the baseline (augural) standards because
                buyers of new vehicles will be forced to spend more on fuel than they
                would have under the augural standards. However, as Figure IV-6 shows,
                the range of fuel economies available in the new market is already
                sufficient to suit the needs of buyers who desire greater fuel economy
                rather than interior volume or some other attributes. Full size pickup
                trucks are now available with smaller turbocharged engines paired with
                8 and 10-speed transmissions and some mild electrification. Buyers
                looking to transport a large family can choose to purchase a plug-in
                hybrid minivan. There were 57 electric models available in 2018, and
                hybrid powertrains are no longer limited to compact cars (as they once
                were). Buyers can choose hybrid SUVs with all-wheel and four-wheel
                drive. While these kinds of highly efficient options were largely
                absent from some body styles in MY 2008, this is no longer the case.
                Given that high-MPG vehicles are widely available, consumers must also
                value other vehicle attributes (e.g., acceleration and load-carrying
                capacity) that can can also be improved with the same technologies that
                can be used to improve fuel economy.
                ---------------------------------------------------------------------------
                 \170\ Circles represent specific outlying vehicle models.
                ---------------------------------------------------------------------------
                 Manufacturers have accomplished a portfolio-wide improvement by
                improving the combustion efficiency of engines (through direct
                injection and
                [[Page 24239]]
                turbocharging), migrating from four and five speed transmissions to 8
                and 10 speed transmissions, and electrifying to varying degrees. All of
                this has increased both production costs and fuel efficiency during a
                period of economic expansion and low energy prices. While the vehicles
                offered for sale have increased significantly in efficiency between MY
                2008 and MY 2020, the sales-weighted average fuel economy has achieved
                less improvement. Despite stringency increases of about five percent
                (year-over-year) between 2012 and 2016, the sales-weighted average fuel
                economy increased marginally. Figure IV-7 shows an initial increase in
                average new vehicle fuel economy (the heavy solid line, shown in mpg as
                indicated on the left y axis), followed by relative stagnation as fuel
                prices (the light dashed lines, shown in dollars per gallon as
                indicated on the right y axis) fell and remained low.\171\ It is worth
                noting that average new vehicle fuel economy observed a brief spike
                during the year that the Tesla Model 3 was introduced (as a consequence
                of strong initial sales volumes, as pre-orders were satisfied, and fuel
                economy ratings that are significantly higher than the industry
                average), and settled around 27.5 MPG in Fall 2019. Average fuel
                economy receded further over the next several months to 26.6 MPG in
                February 2020.\172\
                ---------------------------------------------------------------------------
                 \171\ Ward's Automotive, https://www.wardsauto.com/industry/fuel-economy-index-shows-slow-improvement-april. Last accessed Dec.
                13, 2019.
                 \172\ Ward's Automotive, https://wardsintelligence.informa.com/WI964622/Fuel-Economy-Slightly-Down-in-February. Last accessed Mar.
                9, 2020.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.047
                 In their NPRM comments, manufacturers expressed concern that CAFE
                standards had already increased to the point where the price increases
                necessary to recoup manufacturers' increased costs for providing
                further increases in fuel economy outweigh the value of fuel
                savings.173 174 The agencies do not agree that this point
                has already been reached by previous stringency increases, but
                acknowledge the reality of diminishing marginal returns to improvements
                in fuel economy. A driver with a 40MPG vehicle uses about 300 gallons
                of fuel per year. Increasing the fuel economy of that vehicle to 50MPG,
                a 25 percent increase, would likely be over $1000 in additional
                technology cost. However, that driver would only save 25 percent of
                their annual fuel consumption, or 75 gallons out of 300 gallons. Even
                at $3/gallon, higher than the current national average, that represents
                $225 per year in fuel savings. That means that the buyer's $1000
                investment in additional fuel economy pays back in just under 4.5 years
                (undiscounted). The agencies' respective programs have created greater
                access to high MPG vehicles in all classes and encouraged the
                proliferation of alternative fuels and powertrains. But if the value of
                the fuel savings is insufficient to motivate buyers to invest in ever
                greater levels of fuel economy, manufacturers will face challenges in
                the market.
                ---------------------------------------------------------------------------
                 \173\ NHTSA-2018-0067-12064-25.
                 \174\ NHTSA-2018-0067-12073-2.
                ---------------------------------------------------------------------------
                 While Figure IV-3 through Figure IV-5 illustrate the trends in
                historical CAFE compliance for the entire industry, the figures contain
                another relevant fact. After several consecutive years of increasing
                standards, the achieved and required levels converge. When the
                standards began increasing again for passenger cars in 2011, the prior
                year had industry CAFE levels 5.6 mpg and 7.7 mpg in excess of their
                standards for domestic cars and imported cars, respectively. Yet, by
                2016, the consecutive year-over-year increases had eroded the levels of
                over-compliance. Light trucks similarly exceeded their standard prior
                to increasing standards, which began in 2005. Yet, by 2011, after
                several consecutive years of stringency increases, the industry light-
                truck average CAFE was merely compliant with the rising standard.
                 This is largely due to the fact that stringency requirements have
                increased at a faster rate than achieved fuel
                [[Page 24240]]
                economy levels for several years. The attribute-based standards took
                effect in 2011 for all regulatory classes, although light truck CAFE
                standards had been increasing since 2005. Since 2004, light truck
                stringency has increased an average of 2.7 percent per year, while
                light truck's compliance fuel economy has increased by an average of
                1.7 percent over the same period.\175\ For the passenger classes, a
                similar story unfolds over a shorter period of time. Year over year
                stringency increases have averaged 4.7 percent per year for domestic
                cars (though increases in the first two years were about 8 percent--
                with lower subsequent increases), but achieved fuel economy increases
                averaged only 2.2 percent per year over the same period. Imported
                passenger cars were similar to domestic cars, with average annual
                stringency increases of 4.4 percent but achieved fuel economy levels
                increasing an average of only 1.4 percent per year from 2011 through
                2017. Given that each successive percent increase in stringency is
                harder to achieve than the previous one, long-term discrepancies
                between required and achieved year-over-year increases cannot be offset
                indefinitely with existing credit banks, as they have been so far.
                ---------------------------------------------------------------------------
                 \175\ Both the standards and these calculations are defined in
                consumption space--gallons per mile--which also translates directly
                into CO2 based on the carbon content of the fuel
                consumed.
                ---------------------------------------------------------------------------
                 With the fuel price increases fresh in the minds of consumers, and
                the great recession only recently passed, the CAFE stringency increases
                that began in 2011 (and subsequent CAFE/CO2 stringency
                increases after EPA's program was first enforced in MY 2012) had
                something of a head start. As Figure IV-3 through Figure IV-5
                illustrate, the standards were not binding in MY 2011--even
                manufacturers that had historically paid civil penalties were earning
                credits for overcompliance. It took two years of stringency increase to
                catch up to the CAFE levels already present in MY 2011. However, seven
                consecutive years of increases for passenger cars and a decade of
                increases for light trucks has changed the credit situation. Figure IV-
                8 shows CAFE credit performance for regulated fleets--the solid line
                represents the number of fleets generating shortfalls and the dashed
                line represents the number of fleets earning credits in each model
                year.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.048
                 Fewer than half as many fleets earned surplus credits for over-
                compliance in MY 2017 compared to MY 2011--and this trend is
                persistent. The story varies from one manufacturer to another, but it
                seems sufficient to state the obvious--when the agencies conducted the
                analysis to establish standards through MY 2025 back in 2012, most (if
                not all) manufacturers had healthy credit positions. That is no longer
                the case, and each successive increase requires many fleets to not only
                achieve the new level from the resulting increase, but to resolve
                deficits from the prior year as well. The large sums of credits, which
                last five years under both programs, have allowed most manufacturers to
                resolve shortfalls. But the light truck fleet, in particular, has a
                dwindling supply of credits available for purchase or trade. The
                CO2 program has a provision that allows credits earned
                during the early years of over-compliance to be applied through MY
                2021. This has reduced the compliance burden in the last several years,
                as intended, but will not mitigate the compliance challenges some OEMs
                would face if the baseline standards remained in place and energy
                prices persisted at current levels.
                BILLING CODE 4910-59-P
                [[Page 24241]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.049
                BILLING CODE 4910-59-C
                 Table IV-2 shows the credits earned by each manufacturer over
                time.\176\ As the table shows, when the agencies considered future
                standards in 2012, most manufacturers were earning credits in at least
                one fleet. However, the bold values show years with deficits and even
                some manufacturers who started out in strong positions, such as Ford's
                passenger car fleet, have seen growing deficits in recent years. While
                [[Page 24242]]
                the initial banks for early-action years eases the burden of
                CO2 compliance for many OEMs, the year-to-year compliance
                story is similar to CAFE, see Table IV-3.
                ---------------------------------------------------------------------------
                 \176\ MY 2017 values represent estimated earned credits based on
                MY 2017 final compliance data.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.050
                 Credit position and shortfall rates clearly illustrate
                manufacturers' fleet performance relative to the standards. Recognizing
                that manufacturers plan compliance over several model years at any
                given time, sporadic shortfalls may not be evidence of undue
                difficulty, but sustained, widespread, growing shortfalls should
                probably be viewed as evidence that standards previously believed to be
                manageable might no longer be so. While NHTSA is prohibited by statute
                from considering availability of credits (and thus, size of credit
                banks) in determining maximum feasible standards, it does consider
                shortfalls as part of its determination. EPA has no such prohibition
                under the CAA and is free to consider both credits and shortfalls.
                 These increasing credit shortfalls are occurring at a time that the
                industry is deploying more technology than the agencies anticipated
                when establishing future standards in 2012. The agencies' projections
                of transmission technologies were mixed. While the agencies expected
                the deployment of 8-speed transmissions to about 25 percent of the
                market by MY 2018, transmissions with eight or more gears account for
                almost 30 percent of the market. However, the agencies projected no CVT
                transmissions in future model years, instead projecting high
                penetration of DCTs. However, CVTs currently make up more than 20
                percent of new transmissions. The tradeoff between advanced engines and
                electrification was also underestimated. While the agencies projected
                penetration rates of turbocharged engines that are higher than we've
                observed in the market (45 percent compared to 30 percent), the
                estimated penetration of electric technologies was significantly lower.
                The agencies projected a couple percent of strong hybrids--which we've
                seen--but virtually no PHEVs or EVs. While the volumes of those
                vehicles are still only a couple percent of the new vehicle market,
                they are heavily credited under both programs and can significantly
                improve compliance positions even at smaller volumes. Even lower-level
                electrification technologies, like stop-start systems, are
                significantly more prevalent than we anticipated (stop-start systems
                were projected to be in about 2 percent of the market, compared to over
                20 percent in the 2018 fleet). Despite technology deployment that is
                comparable to 2012 projections, and occasionally more aggressive,
                passenger car and light truck fleets have slightly lower fuel economy
                than projected. As fleet volumes have shifted along the footprint
                curve, the standards have decreased as well (relative to the
                expectation in 2012), but less so. While compliance deficits have been
                modest, they have been accompanied by record sales for several years.
                This has not only depleted existing credit banks, but created
                significant shortfalls that may be more difficult to overcome if sales
                recede from record levels.
                V. Regulatory Alternatives Considered
                 Agencies typically consider regulatory alternatives in proposals as
                a way of evaluating the comparative effects of different potential ways
                of accomplishing their desired goal. NEPA
                [[Page 24243]]
                requires agencies (in this case, NHTSA, but not EPA) to compare the
                potential environmental impacts of their proposed actions to those of a
                reasonable range of alternatives. Executive Orders 12866 and 13563 and
                OMB Circular A-4 also encourage agencies to evaluate regulatory
                alternatives in their rulemaking analyses. Alternatives analysis begins
                with a ``no-action'' alternative, typically described as what would
                occur in the absence of any regulatory action. This final rule, like
                the proposal, includes a no-action alternative, described below, as
                well as seven ``action alternatives.'' The final standards may, in
                places, be referred to as the ``preferred alternative,'' which is NEPA
                parlance, but NHTSA and EPA intend ``final standards'' and ``preferred
                alternative'' to be used interchangeably for purposes of this
                rulemaking.
                 In the proposal, NHTSA and EPA defined the different regulatory
                alternatives (other than the no-action alternative) in terms of
                percent-increases in CAFE and CO2 stringency from year to
                year. Percent increases in stringency referred to changes in the
                standards year over year--as in, standards that become 1 percent more
                stringent each year. Readers should recognize that those year-over-year
                changes in stringency are not measured in terms of mile per gallon or
                CO2 gram per mile differences (as in, 1 percent more
                stringent than 30 miles per gallon in one year equals 30.3 miles per
                gallon in the following year), but in terms of shifts in the footprint
                functions that form the basis for the actual CAFE and CO2
                standards (as in, on a gallon or gram per mile basis, the CAFE and
                CO2 standards change by a given percentage from one model
                year to the next). Under some alternatives, the rate of change was the
                same for both passenger cars and light trucks; under others, the rate
                of change differed. Like the no-action alternative, all of the
                alternatives considered in the proposal were more stringent than the
                preferred alternative.
                 Alternatives considered in the proposal also varied in other
                significant ways. Alternatives 3 and 7 in the proposal involved a
                gradual discontinuation of CAFE and average CO2 adjustments
                reflecting the use of technologies that improve air conditioner
                efficiency or otherwise improve fuel economy under conditions not
                represented by long-standing fuel economy test procedures (off-cycle
                adjustments, described in further detail in Section IX, although the
                proposal itself would have retained these flexibilities. Commenters
                responding to the request for comment on phasing out these
                flexibilities generally supported maintaining the existing program, as
                proposed. Some commenters suggested changes to the existing program
                that were not discussed in the NPRM. Such changes would be beyond the
                scope of this rulemaking and were not considered. Section IX contains a
                more thorough summary of these comments and the issues they raise, as
                well as the agencies' responses. Consistent with the decision to retain
                these flexibilities in the final rule, alternatives reflecting their
                phase-out have not been considered in this final rule.
                 Additionally, in the NPRM for this rule, EPA proposed to exclude
                the option for manufacturers partially to comply with tailpipe
                CO2 standards by generating CO2-equivalent
                emission adjustments associated with air conditioning refrigerants and
                leakage after MY 2020. This approach was proposed in the interest of
                improved harmonization between the CAFE and tailpipe CO2
                emissions programs because this optional flexibility cannot be
                available in the CAFE program.\177\ Alternatives 1 through 8 excluded
                this option. EPA requested comment ``on whether to proceed with [the]
                proposal to discontinue accounting for A/C leakage, methane emissions,
                and nitrous oxide emissions as part of the CO2 emissions
                standards to provide for better harmony with the CAFE program, or
                whether to continue to consider these factors toward compliance and
                retain that as a feature that differs between the programs.'' \178\ EPA
                stated that if EPA were to proceed with excluding A/C refrigerant
                credits as proposed, ``EPA would consider whether it is appropriate to
                initiate a new rulemaking to regulate these programs independently . .
                . .'' \179\ EPA also stated that ``[i]f the agency decides to retain
                the A/C leakage . . . provisions for CO2 compliance, it
                would likely re-insert the current A/C leakage offset and increase the
                stringency levels for CO2 compliance by the offset amounts
                described above (i.e., 13.8 g/mi equivalent for passenger cars and 17.2
                g/mi equivalent for light trucks). EPA received comments from a wide
                range of stakeholders, most of whom opposed the elimination of these
                flexibility provisions.
                ---------------------------------------------------------------------------
                 \177\ For the CAFE program, carbon-based tailpipe emissions
                (including CO2, HC, and CO) are measured, and fuel
                economy is calculated using a carbon balance equation. EPA also uses
                carbon-based emissions (CO2, HC, and CO, the same as for
                CAFE) to calculate tailpipe CO2 for use in determining
                compliance with its standards. In addition, under the no-action
                alternative for the proposal and under all alternatives in the final
                rule, in determining compliance, EPA includes on a CO2
                equivalent basis (using Global Warming Potential (GWP) adjustment)
                A/C refrigerant leakage credits, at the manufacturer's option, and
                nitrous oxide and methane emissions. EPA also has separate emissions
                standards for methane and nitrous oxides. The CAFE program does not
                include or account for A/C refrigerant leakage, nitrous oxide and
                methane emissions because they do not impact fuel economy. Under
                Alternatives 1-8 in the proposal, the standards were closely aligned
                for gasoline powered vehicles because compliance with the fleet
                average standard for such vehicles is based on tailpipe
                CO2, HC, and CO for both programs and not emissions
                unrelated to fuel economy, although diesel and alternative fuel
                vehicles would have continued to be treated differently between the
                CAFE and CO2 programs. While such an approach would have
                significantly improved harmonization between the programs, standards
                would not have been fully aligned because of the small fraction of
                the fleet that uses diesel and alternative fuels (as described in
                the proposal, such vehicles made up approximately four percent of
                the MY 2016 fleet), as well as differences involving EPCA/EISA
                provisions EPA has not adopted, such as minimum standards for
                domestic passenger cars and limits on credit transfers between
                regulated fleets. The proposal to eliminate flexibilities associated
                with A/C refrigerants and leakage was not adopted for this final
                rule, and the reasons for and implications of that decision are
                discussed further below.
                 \178\ 83 FR at 43193 (Aug. 24, 2018).
                 \179\ Id. at 43194.
                ---------------------------------------------------------------------------
                 Specifically, the two major trade organizations of auto
                manufacturers, as well as some individual automakers, supported
                retaining these provisions. Global Automakers commented that ``[a]ir
                conditioning refrigerant leakage . . . should be included for
                compliance with the EPA standards for all MYs, even if it means a
                divergence from the NHTSA standards.'' \180\ Global provides several
                detailed reasons for their comments, including that the existing
                provisions are ``. . . important to maintaining regulatory flexibility
                through real [CO2] emission reductions and would prevent the
                potential for additional bifurcated, separate programs at the state
                level.'' \181\ The Alliance similarly commented that it ``supports
                continuation of the full air conditioning refrigerant leakage credits
                under the [CO2] standards.'' \182\ Some individual
                [[Page 24244]]
                manufacturers, including General Motors,\183\ Fiat Chrysler,\184\ and
                BMW,\185\ also commented in support of maintaining the current A/C
                refrigerant and leakage credits.
                ---------------------------------------------------------------------------
                 \180\ Global, NHTSA-2018-0067-12032, Appendix A at A-5.
                 \181\ Id. Global also stated that excluding A/C leakage credits
                would ``. . . greatly limit the ability [of manufacturers] to select
                the most cost-effective approach for technology improvements and
                result in a costlier, separate set of regulations that actually
                relate to the overall GHG standards.'' Global also expressed concern
                that issuing separate regulations for A/C leakage could take too
                long and create a gap in which States might act to separately
                regulate or even ban refrigerants, and supported continued inclusion
                of A/C leakage and refrigerant regulation in EPA's GHG program to
                avoid risk of an ensuing patchwork. Global argued that manufacturers
                had already invested to meet the existing program, and that ``the
                proposed phase-out also creates another risk that manufacturers will
                have stranded capital in technologies that are not fully
                amortized.'' Global Automakers, EPA-HQ-OAR-2018-0283-5704,
                Attachment A, at A.43-44.
                 \182\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 12.
                Alliance also expressed concern about stranded capital and risk of
                patchwork of state regulation if MAC direct credits were not
                retained in the Federal GHG program. Id. at 80-81.
                 \183\ General Motors, NHTSA-2018-0067-11858, Appendix 4, at 1
                (``General Motors supports the extensive comments from the Alliance
                of Automobile Manufacturers regarding flexibility mechanisms, and
                incorporates them by reference. In particular, the Alliance cites
                the widening gap between the regulatory standards and actual
                industry-wide new vehicle average fuel economy that has become
                evident since 2016, despite the growing use of improvement `credits'
                from various flexibility mechanisms, such as off-cycle technology
                credits, mobile air conditioner efficiency credits, mobile air
                conditioner refrigerant leak reduction credits and credits from
                electrified vehicles.'')
                 \184\ FCA, NHTSA-2018-0067-11943, at 8. FCA also expressed
                concern about patchwork in the absence of a federal rule. Id.
                 \185\ BMW, EPA-HQ-OAR-2018-4204, at 3. BMW stated that ``Today's
                rules allow flexibilities to be used by the motor vehicle
                manufacturers for fuel saving technologies and efficiency gains
                which are not covered in the applicable test procedures. To enhance
                the future use of these technologies and to reward motor vehicle
                manufacturer's investments taken for future innovations, the
                agencies should consider the continuation of current flexibilities
                for the model years 2021 to 2026.''
                ---------------------------------------------------------------------------
                 Auto manufacturing suppliers who addressed A/C refrigerant and
                leakage credits also generally supported retaining the existing
                provisions. MEMA commented that ``It is essential for supplier
                investment and jobs, and continuous innovation and improvements in the
                technologies that the credit programs continue and expand to broaden
                the compliance pathways. MEMA urges the agencies to continue the
                current credit and incentives programs . . . . '' \186\ DENSO also
                supported maintaining the current provisions.\187\ However, BorgWarner
                supported the proposed removal of A/C refrigerant credits ``for
                harmonization reasons,'' while encouraging EPA to regulate A/C
                refrigerants and leakage separately from the CO2
                standards.\188\
                ---------------------------------------------------------------------------
                 \186\ MEMA, available at https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf, comment at p. 2. MEMA also expressed
                concern about stranded capital investments by suppliers and supplier
                jobs if the direct MAC credits were not available; stated that the
                credits were an important compliance flexibility and ``one of the
                highest values of any credit offered in the EPA program;'' and
                stated that ``Harmonizing the programs does not require making them
                identical or equivalent. Rather, harmonization can be achieved by
                better coordinating the two programs to the extent feasible while
                allowing each agency to implement its separate and distinct
                mandate.'' Id. at 15-16.
                 \187\ DENSO, NHTSA-2018-0067-11880, at 8.
                 \188\ BorgWarner, NHTSA-2018-0067-11895, at 10.
                ---------------------------------------------------------------------------
                 The two producers of a lower GWP refrigerant, Chemours and
                Honeywell, commented extensively in support of continuing to allow A/C
                refrigerant and leakage credits for CO2 compliance, making
                both economic and legal arguments. Both Chemours and Honeywell stated
                that A/C refrigerant and leakage credits were a highly cost-effective
                way for OEMs to comply with the CO2 standards,\189\ with
                Chemours suggesting that OEM compliance strategies are based on the
                assumption that these credits will be available for CO2
                compliance \190\ and that any increase in stringency above the proposal
                effectively necessitates that the credits remain part of the
                program.\191\ Honeywell stated that all OEMs (and a variety of other
                parties) supported retaining the credits for CO2
                compliance,\192\ and Chemours, Honeywell, and CBD et al. all noted that
                OEMs are already using the credits for low GWP refrigerants in more
                than 50 percent of the MY 2018 vehicles produced for sale in the
                U.S.\193\ The American Chemistry Council also stated that the ``auto
                industry widely supports the credits, and U.S. chemical manufacturers
                are at a loss as to why EPA would propose to eliminate such a
                successful flexible compliance program.'' \194\ In response to NPRM
                statements expressing concern that the A/C refrigerant and leakage
                credits could be market distorting, both Chemours and Honeywell
                disagreed,\195\ arguing that the credits were simply a highly cost-
                effective means of complying with the CO2 standards,\196\
                and that removal of the credits at this point would, itself, distort
                the market for refrigerants. Honeywell argued that eliminating the A/C
                credit program from CO2 compliance would put the U.S. at a
                competitive disadvantage with other countries, and would risk U.S.
                jobs.\197\
                ---------------------------------------------------------------------------
                 \189\ Chemours at 1 (``MVAC credits many times offer the `least
                cost' approach to compliance . . .'') and 9; Honeywell at 6.
                 \190\ Chemours at 6-7; both Chemours and Honeywell expressed
                concern about OEM reliance on the expectation that HFC credits would
                continue to be part of the CO2 program (Chemours at 31;
                Honeywell at 16-20) and that investments in alternative refrigerants
                would be stranded (Chemours at 1, 3, 4-6; Honeywell at 2, 7-8).
                 \191\ Chemours at 7.
                 \192\ Honeywell at 8-11.
                 \193\ Chemours at 4; Honeywell at 6-7; CBD et al. at 46-47.
                 \194\ American Chemistry Council, EPA-HQ-OAR-2018-0283-1415, at
                9-10 (comments similar to Chemours and Honeywell).
                 \195\ Chemours at 1; Honeywell at 13.
                 \196\ Chemours at 29-30; Honeywell at 13-14.
                 \197\ Honeywell at 20-21.
                ---------------------------------------------------------------------------
                 Regarding the NPRM's statements that removing the A/C refrigerant
                and leakage credits from CO2 compliance would promote
                harmonization with the CAFE program, these commenters argued that
                harmonization was not a valid basis for that aspect of the proposal.
                Chemours, Honeywell, and CBD et al. all argued that Section 202(a)
                creates no obligation to harmonize the [CO2] program with
                the CAFE program.\198\ Chemours further argued that to the extent
                disharmonization between the programs existed, it should be addressed
                via stringency changes (i.e., reducing CAFE stringency relative to
                CO2 stringency) rather than ``dropping low-cost compliance
                options.'' \199\
                ---------------------------------------------------------------------------
                 \198\ Chemours at 23-24; Honeywell at 11-12; CBD et al. at 47.
                 \199\ Chemours at 9-11.
                ---------------------------------------------------------------------------
                 These commenters also expressed concern that the proposal
                constituted an EPA decision not to regulate HFC emissions from motor
                vehicles at all. Commenters argued that the NPRM provided no legal
                analysis or reasoned explanation for stopping regulation of HFCs,\200\
                and that Massachusetts v. EPA requires any final rule to regulate all
                greenhouse gases from motor vehicles and not CO2 alone,\201\
                suggesting that there was a high likelihood of a lapse in regulation
                because EPA had not yet proposed a new way of regulating HFC
                emissions.\202\ Because the NPRM provided no specific information about
                how EPA might regulate non-CO2 emissions separately,
                commenters argued that the lack of clarity was inherently disruptive to
                OEMs.\203\ CBD et al. argued that any lapse in regulation is ``illegal
                on its face'' and that even creating a separate standard for HFC
                emissions would be ``illegal'' because it ``would increase costs to
                manufacturers and result in environmental detriment by removing any
                incentive to use the most aggressive approaches to curtail emissions of
                these highly potent GHGs.'' \204\
                ---------------------------------------------------------------------------
                 \200\ Chemours at 1-2; Honeywell at 11.
                 \201\ Chemours at 18-19; Honeywell at 14-16.
                 \202\ Chemours at 6; Honeywell at 16.
                 \203\ Chemours at 21; Honeywell at 16; ICCT at I-39.
                 \204\ CBD et al. at 46.
                ---------------------------------------------------------------------------
                 Environmental organizations,\205\ other NGOs, academic
                institutions, consumer organizations, and state governments \206\ also
                commented in support of continuing the existing provisions.
                ---------------------------------------------------------------------------
                 \205\ ICCT, NHTSA-2018-0067-11741, Full Comments, at 4
                (describing ``air conditioning GHG-reduction technologies [as]
                available, cost-effective, and experiencing increased deployment by
                many companies due to the standards.''); CBD et al., Appendix A, at
                45-47.
                 \206\ CARB, NHTSA-2018-0067-11873, Detailed Comments, at 120-
                121; Washington State Department of Ecology, NHTSA-2018-0067-11926,
                at 6 (HFCs are an important GHG; compliance flexibility is
                important).
                ---------------------------------------------------------------------------
                 EPA has considered its proposed approach to A/C refrigerant and
                leakage
                [[Page 24245]]
                credits in light of these comments. EPA believes that maintaining this
                element of its program is consistent with EPA's authority under Section
                202(a) to establish standards for reducing emissions from LDVs. Thus,
                maintaining the optional HFC credit program is appropriate. In
                addition, EPA recognizes the value of regulatory flexibility and
                compliance options, and has concluded that the advantages from
                retaining the existing A/C refrigerant/leakage credit program and
                associated offset between the CO2 and CAFE standards--in
                terms of providing for a more-comprehensive regulation of emissions
                from light-duty vehicles--outweigh the disadvantages resulting from the
                lack of harmonization.
                 Regarding the comment from BorgWarner about how having a separate
                A/C refrigerant and leakage regulation would allow for better
                harmonization between the programs, the agencies accept this to be an
                accurate statement, but believe the benefits of continued refrigerant
                regulation as an option for CO2 compliance outweigh the
                problems associated with lack of harmonization with the CAFE program.
                 For these reasons, EPA is not finalizing the proposed provisions,
                and is making no changes in the A/C refrigerant and leakage-related
                provisions of the current program. In light of this conclusion, EPA
                does not need to address the legal arguments made by CBD et al. and
                CARB about regulating refrigerant-related emissions separately, or
                potential lapses in regulation of refrigerant emissions while such a
                program could be developed.
                 As with A/C refrigerant and leakage credits, EPA proposed to
                exclude nitrous oxide and methane from average performance calculations
                after model year 2020, thereby removing these optional program
                flexibilities. Alternatives 1 through 8 excluded this option. EPA
                sought comment on whether to remove those aspects of the program that
                allow a manufacturer to use nitrous oxide and methane emissions
                reductions for compliance with its CO2 average fleet
                standards because such a flexibility is not allowed in the NHTSA CAFE
                program, or whether to retain the flexibilities as a feature that
                differs between the programs. Further, EPA sought comment on whether to
                change the existing methane and nitrous oxide standards. Specifically,
                EPA requested information from the public on whether the existing
                standards are appropriate, or whether they should be revised to be less
                stringent or more stringent based on any updated data.
                 The Alliance in its comments may have misunderstood EPA's proposal
                to mean that EPA was proposing to eliminate regulation of methane and
                nitrous oxide emissions altogether. The Alliance commented in support
                of such a proposal as they understood it, to eliminate the standards to
                provide better harmony between the two compliance programs.\207\ The
                Alliance commented that ``[n]ot only is emission of these two
                substances from vehicles a relatively minor contribution to GHG
                emissions, the Alliance has continuing concern regarding measurement
                and testing technologies for nitrous oxide.'' \208\ The Alliance
                commented further that if ``EPA decides instead to continue to regulate
                methane and nitrous oxide, the Alliance recommends that EPA re-assess
                whether the levels of the standards remain appropriate and to retain
                the current compliance flexibilities. Furthermore, in this scenario,
                the Alliance also recommends that methane and nitrous oxide standards
                be assessed as a fleet average and as the average of FTP and HFET test
                cycles.'' \209\ Several individual manufacturers submitted similar
                comments, including Ford,\210\ FCA,\211\ Volvo,\212\ and Mazda.\213\
                Ford also commented that it does not support the proposal to maintain
                the existing N2O/CH4 standards while removing the
                program flexibilities.\214\
                ---------------------------------------------------------------------------
                 \207\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 13.
                 \208\ Id.
                 \209\ Id.
                 \210\ Ford, EPA-HQ-OAR-2018-0283-5691, at 4.
                 \211\ FCA, NHTSA-2018-0067-11943, at 9.
                 \212\ Volvo, NHTSA-2018-0067-12036, at 5.
                 \213\ Mazda, NHTSA-2018-0067-11727, at 3 (``In reality, these
                emissions are at deminimis levels and have very little, if any,
                impact on global warming. So, the need to regulate these emissions
                as part of the GHG program, or separately, is unclear. Although most
                current engines can comply with the existing requirements, there are
                some existing and upcoming new technologies that may not be able to
                fully comply. These technologies can provide substantial
                CO2 reductions.'').
                 \214\ Ford, at 4 (``Finally, without the ability to incorporate
                exceedances into CREE, each vehicle will need to employ hardware
                solutions if they do not comply. We do not believe it was EPA's
                intent in the original rulemaking to require additional after-
                treatment, with associated cost increases, explicitly for the
                control and reduction of an insignificant contributor to GHG
                emissions. Therefore, we do not support the proposal to maintain the
                existing N2O/CH4 standards while removing the
                CREE exceedance pathway.'').
                ---------------------------------------------------------------------------
                 The Alliance further commented that ``data from the 2016 EPA report
                on light-duty vehicle emissions supports the position that
                CH4 and N2O have minimal impact on total GHG
                emissions, reporting only 0.045 percent in exceedance of the standard.
                This new information makes it apparent that CH4 and
                N2O contribute a de minimis amount to GHG emissions.
                Additionally, gasoline CH4 and N2O performance is
                within the current standards. Finally, the main producers of
                CH4 and N2O emissions are flex fuel (E85) and
                diesel vehicles, and these vehicles have been declining in sales as
                compared to gasoline-fueled vehicles.'' \215\ The Alliance also
                commented that CH4 and N2O have minimal
                opportunities to be catalytically treated, as N2O is
                generated in the catalyst and CH4 has a low conversion
                efficiency compared to other emissions. EPA did not intend that
                additional hardware should be required to comply with the
                CH4 or N2O standards on any vehicle.'' \216\
                ---------------------------------------------------------------------------
                 \215\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 43.
                 \216\ Id. at 44.
                ---------------------------------------------------------------------------
                 Global Automakers commented in support of continuing inclusion of
                nitrous oxide and methane emissions standards for all MYs, even if it
                means a divergence from the NHTSA standards for these program elements
                in the regulations, ``because they are complementary to EPA's program,
                and are better managed through a coordinated federal policy. They are
                also important to maintaining regulatory flexibility through real
                [CO2] emission reductions and would prevent the potential
                for additional bifurcated, separate programs at the state level.''
                \217\ Global Automakers recommended that they remain in place per the
                existing program but continued to support that the N2O
                testing is not necessary. Global Automakers commented that it
                ``strongly recommends reducing the need for N2O testing or
                eliminating these test requirements in their entirety. It should be
                sufficient to allow manufacturers to attest to compliance with the
                N2O capped standards based upon good engineering judgment,
                development testing, and correlation to NOX emissions. EPA
                could, however, maintain the option to request testing to be performed
                for new technologies only, which could have unknown impacts on
                N2O emissions.'' \218\ Hyundai \219\ and Kia \220\ submitted
                similar comments.
                ---------------------------------------------------------------------------
                 \217\ Global, NHTSA-2018-0067-12032, at 4, 5.
                 \218\ Global, Appendix A, NHTSA-2018-0067-12032, at A-44, fn.
                89.
                 \219\ Hyundai, EPA-HQ-OAR-2018-0283-4411, at 7.
                 \220\ Kia, EPA-HQ-OAR-2018-0283-4195, at 8-9.
                ---------------------------------------------------------------------------
                 Others commented in support of retaining the existing program. MECA
                commented that it supports the existing standards for methane and
                nitrous oxide because catalyst technologies provided by MECA members
                that reduce these climate forcing gases are readily
                [[Page 24246]]
                available and cost-effective.\221\ MECA also commented that the ability
                to trade reductions in these pollutants in exchange for CO2
                gives vehicle manufacturers the flexibilities they need to comply with
                the emission limits by the most cost-effective means.\222\ CBD et al.
                commented that the alternative compliance mechanisms currently
                available in the program exist to provide cost-effective options for
                compliance, and were considered by manufacturers to be a necessary
                element of the program for certain types of vehicles.\223\ CBD et al.
                further argued that ``[e]liminating these flexibilities consequently
                imposes costs on manufacturers without discernible environmental
                benefits,'' and suggested that harmonization with the CAFE program was
                not a relevant decision factor for EPA.\224\ Several other parties
                commented generally in support of retaining the existing program for A/
                C leakage credits, discussed above, and N2O and
                CH4 standards.\225\
                ---------------------------------------------------------------------------
                 \221\ MECA, NHTSA-2018-0067-11994, at 12.
                 \222\ Id.
                 \223\ CBD et al. at 48.
                 \224\ Id.
                 \225\ Washington State Department of Ecology, NHTSA-2018-0067-
                11926, at 6.
                ---------------------------------------------------------------------------
                 After considering these comments, EPA is retaining the regulatory
                provisions related to the N2O and CH4 standards
                with no changes, specifically including the existing flexibilities that
                accompany those standards. EPA is not adopting its proposal to exclude
                nitrous oxide and methane emissions from average performance
                calculations after model year 2020 or any other changes to the program.
                The standards continue to serve their intended purpose of capping
                emissions of those pollutants and providing for more-comprehensive
                regulation of emissions from light-duty vehicles. The standards were
                intended to prevent future emissions increases, and these standards
                were generally not expected to result in the application of new
                technologies or significant costs for manufacturers using current
                vehicle designs.\226\ The program flexibilities are working as intended
                and all manufacturers are successfully complying with the standards.
                Most vehicle models are well below the standards and for those that are
                above the standards, manufacturers have used the flexibilities to
                offset exceedances with CO2 improvements to demonstrate
                compliance. EPA did not receive any data in response to its request for
                comments supporting potential alternative levels of stringency.
                ---------------------------------------------------------------------------
                 \226\ 77 FR 62624, at 62799 (Oct 15, 2012).
                ---------------------------------------------------------------------------
                 While the Alliance and several individual manufacturers recommended
                eliminating the standards altogether, EPA did not propose to eliminate
                the standards, but to eliminate the optional flexibilities, and
                solicited comment on adjusting the standards to be more or less
                stringent. Thus, EPA does not believe it would be appropriate to
                eliminate completely the standards in this final rule without providing
                an opportunity for comment on that idea. Furthermore, as noted above,
                EPA believes the standards are continuing to serve their intended
                purpose of capping emissions and remain appropriate. Manufacturers have
                been subject to the standards for several years, and the Alliance
                acknowledges in their comments that the exceedance of the standards,
                which is offset by manufacturers using compliance flexibilities, is
                very small and that most vehicles meet the standards. Regarding the
                Alliance comments that the standards should be based on a fleet average
                approach, EPA notes that the purpose of the standards is to cap
                emissions, not to achieve fleet-wide reductions.\227\ The fleet average
                emissions for N2O and CH4 are well below the
                numerical level of the cap standards and therefore the existing cap
                standards would not be an appropriate fleet average standard. Adopting
                a fleet average approach using the same numerical level as the
                established cap standards would not achieve the intended goal of
                capping emissions at current levels. If technologies lead to
                exceedances of the caps, automakers have the opportunity to apply
                appropriate flexibilities under the current program to achieve GHG
                emission neutrality. EPA is not aware of any manufacturer that has been
                prevented from bringing a technology to the marketplace because of the
                current cap levels or approach. EPA believes it would need to consider
                all options further, with an opportunity for public comment, before
                adopting such a significant change to the program.
                ---------------------------------------------------------------------------
                 \227\ Relatedly, the Alliance and Global Automakers raised
                concerns in their comments regarding N2O measurement and
                testing burden. EPA did not propose any changes in testing
                requirements and at this time EPA is not adopting any changes.
                Manufacturers have been measuring N2O emissions and have
                successfully certified vehicles to the N2O standards for
                several years and EPA does not believe N2O measurement is
                an issue needing regulatory change. EPA continues to believe direct
                measurement is the best way for manufacturers to demonstrate
                compliance with the N2O standards and is more appropriate
                than an engineering statement without direct measurement.
                ---------------------------------------------------------------------------
                 As explained above, the agencies have changed the alternatives
                considered for the final rule, partly in response to comments. The
                basic form of the standards represented by the alternatives--footprint-
                based, defined by particular mathematical functions--remains the same
                and as described in the NPRM. For the EPA program, EPA has chosen in
                this final rule to retain the existing program for regulation of A/C
                refrigerant leakage, nitrous oxide, and methane emissions as part of
                the CO2 standard. This allows manufacturers to continue to
                rely on this flexibility which they describe as extremely important for
                compliance, although it results in continued differences between EPA's
                and NHTSA's programs. This approach also avoids the possibility of gaps
                in the regulation of HFCs, CH4, and N2O while EPA
                developed a different way of regulating the non-CO2
                emissions as part of or concurrent with the NPRM, and thereby allows
                EPA to continue to regulate GHE emissions from light-duty vehicles on a
                more-comprehensive basis. Thus, all alternatives considered in this
                final rule reflect inclusion of CH4, N2O, and HFC
                in EPA's overall ``CO2'' (more accurately, CO2-
                equivalent, or CO2e) requirements. Besides this change, the
                alternatives considered for the final rule differ from the NPRM in two
                additional ways: First, alternatives reflecting the phase-out of the A/
                C efficiency and off-cycle programs have been dropped in response to
                certain comments and in recognition of the potential real-world
                benefits of those programs. And second, the preferred alternative for
                this final rule reflects a 1.5 percent year-over-year increase for both
                passenger cars and light trucks. These changes will be discussed
                further below, following a brief discussion of the form of the
                standards.
                A. Form of the Standards
                 As in the CAFE and CO2 rulemakings in 2010 and 2012,
                NHTSA and EPA proposed in the NPRM to set attribute-based CAFE and
                CO2 standards defined by a mathematical function of vehicle
                footprint, which has observable correlation with fuel economy and
                vehicle emissions. EPCA, as amended by EISA, expressly requires that
                CAFE standards for passenger cars and light trucks be based on one or
                more vehicle attributes related to fuel economy and be expressed in the
                form of a mathematical function.\228\ While the CAA includes no
                specific requirements regarding CO2 regulation, EPA has
                chosen to adopt attribute-based CO2 standards consistent
                with NHTSA's EPCA/EISA requirements in the interest of harmonization
                and simplifying compliance. Such an approach is permissible under
                section 202(a) of the
                [[Page 24247]]
                CAA, and EPA has used the attribute-based approach in issuing standards
                under analogous provisions of the CAA. Thus, both the proposed and
                final standards take the form of fuel economy and CO2
                targets expressed as functions of vehicle footprint (the product of
                vehicle wheelbase and average track width). Section V.A.2 below
                discusses the agencies' continued reliance on footprint as the relevant
                attribute.
                ---------------------------------------------------------------------------
                 \228\ 49 U.S.C. 32902(a)(3)(A).
                ---------------------------------------------------------------------------
                 Under the footprint-based standards, the function defines a
                CO2 or fuel economy performance target for each unique
                footprint combination within a car or truck model type. Using the
                functions, each manufacturer thus will have a CAFE and CO2
                average standard for each year that is almost certainly unique to each
                of its fleets,\229\ based upon the footprints and production volumes of
                the vehicle models produced by that manufacturer. A manufacturer will
                have separate footprint-based standards for cars and for trucks. The
                functions are mostly sloped, so that generally, larger vehicles (i.e.,
                vehicles with larger footprints) will be subject to lower CAFE mpg
                targets and higher CO2 grams/mile targets than smaller
                vehicles. This is because, generally speaking, smaller vehicles are
                more capable of achieving higher levels of fuel economy/lower levels of
                CO2 emissions, mostly because they tend not to have to work
                as hard (and therefore require as much energy) to perform their driving
                task. Although a manufacturer's fleet average standards could be
                estimated throughout the model year based on the projected production
                volume of its vehicle fleet (and are estimated as part of EPA's
                certification process), the standards to which the manufacturer must
                comply are determined by its final model year production figures. A
                manufacturer's calculation of its fleet average standards as well as
                its fleets' average performance at the end of the model year will thus
                be based on the production-weighted average target and performance of
                each model in its fleet.\230\
                ---------------------------------------------------------------------------
                 \229\ EPCA/EISA requires NHTSA to separate passenger cars into
                domestic and import passenger car fleets whereas EPA combines all
                passenger cars into one fleet.
                 \230\ As discussed in prior rulemakings, a manufacturer may have
                some vehicle models that exceed their target and some that are below
                their target. Compliance with a fleet average standard is determined
                by comparing the fleet average standard (based on the production-
                weighted average of the target levels for each model) with fleet
                average performance (based on the production-weighted average of the
                performance of each model).
                ---------------------------------------------------------------------------
                 For passenger cars, consistent with prior rulemakings, NHTSA is
                defining fuel economy targets as follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.051
                where:
                TARGETFE is the fuel economy target (in mpg) applicable to a
                specific vehicle model type with a unique footprint combination,
                a is a minimum fuel economy target (in mpg),
                b is a maximum fuel economy target (in mpg),
                c is the slope (in gallons per mile per square foot, or gpm, per
                square foot) of a line relating fuel consumption (the inverse of
                fuel economy) to footprint, and
                d is an intercept (in gpm) of the same line.
                 Here, MIN and MAX are functions that take the minimum and maximum
                values, respectively, of the set of included values. For example,
                MIN[40,35] = 35 and MAX(40, 25) = 40, such that MIN[MAX(40, 25), 35] =
                35.
                 For light trucks, also consistent with prior rulemakings, NHTSA is
                defining fuel economy targets as follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.052
                where:
                TARGETFE is the fuel economy target (in mpg) applicable to a
                specific vehicle model type with a unique footprint combination,
                a, b, c, and d are as for passenger cars, but taking values specific
                to light trucks,
                e is a second minimum fuel economy target (in mpg),
                f is a second maximum fuel economy target (in mpg),
                g is the slope (in gpm per square foot) of a second line relating
                fuel consumption (the inverse of fuel economy) to footprint, and
                h is an intercept (in gpm) of the same second line.
                 Although the general model of the target function equation is the
                same for each vehicle category (passenger cars and light trucks) and
                each model year, the parameters of the function equation differ for
                cars and trucks. For MYs 2020-2026, the parameters are unchanged,
                resulting in the same stringency in each of those model years.
                 Mathematical functions defining the CO2 targets are
                expressed as functions that are similar, with coefficients a-h
                corresponding to those listed above.\231\ For passenger cars, EPA is
                defining CO2 targets mathematically equivalent to the
                following:
                ---------------------------------------------------------------------------
                 \231\ EPA regulations use a different but mathematically
                equivalent approach to specify targets. Rather than using a function
                with nested minima and maxima functions, EPA regulations specify
                requirements separately for different ranges of vehicle footprint.
                Because these ranges reflect the combined application of the listed
                minima, maxima, and linear functions, it is mathematically
                equivalent and more efficient to present the targets as in this
                Section.
                ---------------------------------------------------------------------------
                TARGETCO2 = MIN[b, MAX[a, c x FOOTPRINT + d]]
                where:
                TARGETCO2 is the is the CO2 target (in grams per mile, or
                g/mi) applicable to a specific vehicle model configuration,
                a is a minimum CO2 target (in g/mi),
                b is a maximum CO2 target (in g/mi),
                c is the slope (in g/mi, per square foot) of a line relating
                CO2 emissions to footprint, and
                d is an intercept (in g/mi) of the same line.
                 For light trucks, CO2 targets are defined as follows:
                TARGETCO2 = MIN[MIN[b, MAX[a, c x FOOTPRINT + d]], MIN[f, MAX[e, g x
                FOOTPRINT + h]]
                [[Page 24248]]
                where:
                TARGETCO2 is the is the CO2 target (in g/mi) applicable
                to a specific vehicle model configuration,
                a, b, c, and d are as for passenger cars, but taking values specific
                to light trucks,
                e is a second minimum CO2 target (in g/mi),
                f is a second maximum CO2 target (in g/mi),
                g is the slope (in g/mi per square foot) of a second line relating
                CO2 emissions to footprint, and
                h is an intercept (in g/mi) of the same second line.
                 To be clear, as has been the case since the agencies began
                establishing attribute-based standards, no vehicle need meet the
                specific applicable fuel economy or CO2 targets, because
                compliance with either CAFE or CO2 standards is determined
                based on corporate average fuel economy or fleet average CO2
                emission rates. In this respect, CAFE and CO2 standards are
                unlike, for example, safety standards and traditional vehicle emissions
                standards. CAFE and CO2 standards apply to the average fuel
                economy levels and CO2 emission rates achieved by
                manufacturers' entire fleets of vehicles produced for sale in the U.S.
                Safety standards apply on a vehicle-by-vehicle basis, such that every
                single vehicle produced for sale in the U.S. must, on its own, comply
                with minimum FMVSS. Similarly, criteria pollutant emissions standards
                are applied on a per-vehicle basis, such that every vehicle produced
                for sale in the U.S. must, on its own, comply with all applicable
                emissions standards. When first mandating CAFE standards in the 1970s,
                Congress specified a more flexible averaging-based approach that allows
                some vehicles to ``under comply'' (i.e., fall short of the overall flat
                standard, or fall short of their target under attribute-based
                standards) as long as a manufacturer's overall fleet is in compliance.
                 The required CAFE level applicable to a given fleet in a given
                model year is determined by calculating the production-weighted
                harmonic average of fuel economy targets applicable to specific vehicle
                model configurations in the fleet, as follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.053
                where:
                CAFErequired is the CAFE level the fleet is required to achieve,
                i refers to specific vehicle model/configurations in the fleet,
                PRODUCTIONi is the number of model configuration i produced for sale
                in the U.S., and
                TARGETFE,i the fuel economy target (as defined above) for model
                configuration i.
                 Similarly, the required average CO2 level applicable to
                a given fleet in a given model year is determined by calculating the
                production-weighted average (not harmonic) of CO2 targets
                applicable to specific vehicle model configurations in the fleet, as
                follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.054
                where:
                CO2required is the average CO2 level the fleet is
                required to achieve,
                i refers to specific vehicle model/configurations in the fleet,
                PRODUCTIONi is the number of model configuration i produced for sale
                in the U.S., and
                TARGETCO2,i is the CO2 target (as defined above) for
                model configuration i.
                 Section VI.A.1 describes the advantages of attribute standards,
                generally. Section VI.A.2 explains the agencies' specific decision to
                use vehicle footprint as the attribute over which to vary stringency
                for past and current rules. Section VI.A.3 discusses the policy
                considerations in selecting the specific mathematical function. Section
                VI.A.4 discusses the methodologies used to develop current attribute-
                based standards, and the agencies' current proposal to continue to do
                so for MYs 2021-2026. Section VI.A.5 discusses the methodologies used
                to reconsider the mathematical function for the proposed standards.
                1. Why attribute-based standards, and what are the benefits?
                 Under attribute-based standards, every vehicle model has fuel
                economy and CO2 targets, the levels of which depend on the
                level of that vehicle's determining attribute (for the MYs 2021-2026
                standards, footprint is the determining attribute, as discussed below).
                The manufacturer's fleet average CAFE performance is calculated by the
                harmonic production-weighted average of those targets, as defined
                below:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.055
                 Here, i represents a given model \232\ in a manufacturer's
                fleet, Productioni represents the U.S. production of that model, and
                Targeti represents the target as defined by the attribute-based
                standards. This means no vehicle is required to meet its target;
                instead, manufacturers are free to balance improvements however they
                deem best within (and, given credit transfers, at least partially
                across) their fleets.
                 \232\ If a model has more than one footprint variant, here each
                of those variants is treated as a unique model, i, since each
                footprint variant will have a unique target.
                ---------------------------------------------------------------------------
                 Because CO2 is on a gram per mile basis rather a mile
                per gallon basis,
                [[Page 24249]]
                harmonic averaging is not necessary when calculating required
                CO2 levels:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.056
                 The idea is to select the shape of the mathematical function
                relating the standard to the fuel economy-related attribute to reflect
                the trade-offs manufacturers face in producing more of that attribute
                over fuel efficiency (due to technological limits of production and
                relative demand of each attribute). If the shape captures these trade-
                offs, every manufacturer is more likely to continue adding fuel-
                efficient technology across the distribution of the attribute within
                their fleet, instead of potentially changing the attribute--and other
                correlated attributes, including fuel economy--as a part of their
                compliance strategy. Attribute-based standards that achieve this have
                several advantages.
                 First, assuming the attribute is a measurement of vehicle size,
                attribute-based standards help to at least partially reduce the
                incentive for manufacturers to respond to CAFE and CO2
                standards by reducing vehicle size in ways harmful to safety, as
                compared to ``flat,'' non-attribute based standards.\233\ Larger
                vehicles, in terms of mass and/or crush space, generally consume more
                fuel and produce more carbon dioxide emissions, but are also generally
                better able to protect occupants in a crash.\234\ Because each vehicle
                model has its own target (determined by a size-related attribute),
                properly fitted attribute-based standards reduce the incentive to build
                smaller vehicles simply to meet a fleet-wide average, because smaller
                vehicles are subject to more stringent compliance targets.
                ---------------------------------------------------------------------------
                 \233\ The 2002 NAS Report described at length and quantified the
                potential safety problem with average fuel economy standards that
                specify a single numerical requirement for the entire industry. See
                Transportation Research Board and National Research Council. 2002.
                Effectiveness and Impact of Corporate Average Fuel Economy (CAFE)
                Standards, Washington, DC: The National Academies Press (``2002 NAS
                Report'') at 5, finding 12, available at https://www.nap.edu/catalog/10172/effectiveness-and-impact-of-corporate-average-fuel-economy-cafe-standards (last accessed June 15, 2018). Ensuing
                analyses, including by NHTSA, support the fundamental conclusion
                that standards structured to minimize incentives to downsize all but
                the largest vehicles will tend to produce better safety outcomes
                than flat standards.
                 \234\ Bento, A., Gillingham, K., & Roth, K. (2017). The Effect
                of Fuel Economy Standards on Vehicle Weight Dispersion and Accident
                Fatalities. NBER Working Paper No. 23340. Available at http://www.nber.org/papers/w23340 (last accessed June 15, 2018).
                ---------------------------------------------------------------------------
                 Second, attribute-based standards, if properly fitted, provide
                automakers with more flexibility to respond to consumer preferences
                than do single-valued standards. As discussed above, a single-valued
                standard encourages a fleet mix with a larger share of smaller vehicles
                by creating incentives for manufacturers to use downsizing the average
                vehicle in their fleet (possibly through fleet mixing) as a compliance
                strategy, which may result in manufacturers building vehicles for
                compliance reasons that consumers do not want. Under a size-related,
                attribute-based standard, reducing the size of the vehicle for
                compliance's sake is a less-viable strategy because smaller vehicles
                have more stringent regulatory targets. As a result, the fleet mix
                under such standards is more likely to reflect aggregate consumer
                demand for the size-related attribute used to determine vehicle
                targets.
                 Third, attribute-based standards provide a more equitable
                regulatory framework across heterogeneous manufacturers who may each
                produce different shares of vehicles along attributes correlated with
                fuel economy.\235\ An industry-wide single-value CAFE standard imposes
                disproportionate cost burden and compliance challenges on manufacturers
                who produce more vehicles with attributes inherently correlated with
                lower fuel economy--i.e. manufacturers who produce, on average, larger
                vehicles. As discussed above, retaining flexibility for manufacturers
                to produce vehicles which respect heterogeneous market preferences is
                an important consideration. Since manufacturers may target different
                markets as a part of their business strategy, ensuring that these
                manufacturers do not incur a disproportionate share of the regulatory
                cost burden is an important part of conserving consumer choices within
                the market.
                ---------------------------------------------------------------------------
                 \235\ 2002 NAS Report at 4-5, finding 10.
                ---------------------------------------------------------------------------
                 Industry commenters generally supported attribute-based standards,
                while other commenters questioned their benefits. IPI argued that
                preserving the current vehicle mix was not necessarily desirable or
                necessary for consumer welfare, and suggested that some vehicle
                downsizing in the fleet might be beneficial both for safety and for
                compliance.\236\ IPI also argued that compliance credit trading would
                ``help smooth out any disproportionate impacts on certain
                manufacturers'' and ``ensure that manufacturers with relatively
                efficient fleets still have an incentive to continue improving fuel
                economy (in order to generate credits)'' \237\ Similarly, citing Ito
                and Sallee, Kathryn Doolittle commented that ``. . . Ito and Sallee
                (2018) have found ABR [``attribute-based regulations''] inefficient in
                cost when juxtaposed with flat standard with compliance trading.''
                \238\
                ---------------------------------------------------------------------------
                 \236\ IPI, NHTSA-2018-0067-12362, at 14-15.
                 \237\ IPI, NHTSA-2018-0067-12362, at 14.
                 \238\ Doolittle, K, NHTSA-2018-0067-7411. See also Ito, K and
                Sallee, J. ``The Economics of Attribute-Based Regulation: Theory and
                Evidence from Fuel Economy Standards.'' The Review of Economics and
                Statistics (2018), 100(2), pp. 319-36.
                ---------------------------------------------------------------------------
                 The agencies have considered these comments. IPI incorrectly
                characterizes the agencies' prior statements as claims that it is
                important to preserve the current vehicle mix. EPA and NHTSA have never
                claimed, and are not today claiming that it is important to preserve
                the current fleet mix. The agencies have said, and are today
                reiterating, that it is reasonable to expect that reducing the tendency
                of standards to distort the market should reduce at least part of the
                tendency of standards to reduce consumer welfare. Or, more concisely,
                it is better to work with the market than against it. Single-value (aka
                flat) CAFE standards in place from the 1970s through 2010 were clearly
                distortionary. Recognizing this, the National Academy of Sciences
                recommended in 2002 that NHTSA adopt attribute-based CAFE standards.
                NHTSA did so in 2006, for light trucks produced starting MY 2008. As
                mentioned above, in 2007, Congress codified the requirement for
                attribute-based passenger car and light truck CAFE standards. Agreeing
                with this history, premise, and motivation, EPA has also adopted
                attribute-based CO2 standards. None of this is to say the
                agencies consider it important to hold fleet mix constant. Rather, the
                agencies expect that, compared to flat standards, attribute-based
                standards can allow the market--including fleet mix--to better
                [[Page 24250]]
                follow its natural course, and all else equal, consumer acceptance is
                likely to be greater if the market does so.
                 The agencies also disagree with comments implying that compliance
                credit trading can address all of the market distortion that flat
                standards would entail. Evidence thus far suggests that trading is
                fragmented, with some manufacturers apparently willing to trade only
                with some other specific manufacturers. The Ito and Sallee article
                cited by one commenter is a highly idealized theoretical construction,
                with the authors noting, inter alia, that their model ``assumes perfect
                competition.'' \239\ Its findings regarding comparative economic
                efficiency of flat- and attribute-based standards are, therefore,
                merely hypothetical, and the agencies find little basis in recent
                transactions to suggest the compliance credit trading market reflects
                the authors' idealized assumptions. Even if the agencies did expect
                credit trading markets to operate as in an idealized textbook example,
                basing the structure of standards on the presumption of perfect trading
                would not be appropriate. FCA commented that ``. . . when flexibilities
                are considered while setting targets, they cease to be flexibilities
                and become simply additional technology mandates,'' and the Alliance
                commented, similarly, that ``the Agencies should keep `flexibilities'
                as optional ways to comply and not unduly assume that each flexibility
                allows additional stringency of footprint-based standards.'' \240\
                Perhaps recognizing this reality, Congress has barred NHTSA from
                considering manufacturers' ability to use compliance credits (even
                credits earned and used by the same OEM, much less credits traded
                between OEMs). As discussed further in Section VIII.A.2, EPA believes
                that while credit trading may be a useful flexibility to reduce the
                overall costs of the program, it is important to set standards in a way
                that does not rely on credit purchasing availability as a compliance
                mechanism.
                ---------------------------------------------------------------------------
                 \239\ Ito and Sallee, op. cit., Supplemental Appendix, at A-15,
                available at https://www.mitpressjournals.org/doi/suppl/10.1162/REST_a_00704/suppl_file/REST_a_00704-esupp.pdf (accessed October 29,
                2019).
                 \240\ FCA, NHTSA-2018-0067-11943, at 6; Alliance, NHTSA-2018-
                0067-12073, Full Comment Set, at 40, fn. 82.
                ---------------------------------------------------------------------------
                 Considering these comments and realities, considering EPCA's
                requirement for attribute-based CAFE standards, and considering the
                benefits of regulatory harmonization, the agencies are, again,
                finalizing attribute-based CAFE and CO2 standards rather
                than, for either program, finalizing flat standards.
                Why footprint as the attribute?
                 It is important that the CAFE and CO2 standards be set
                in a way that does not unnecessarily incentivize manufacturers to
                respond by selling vehicles that are less safe. Vehicle size is highly
                correlated with vehicle safety--for this reason, it is important to
                choose an attribute correlated with vehicle size (mass or some
                dimensional measure). Given this consideration, there are several
                policy and technical reasons why footprint is considered to be the most
                appropriate attribute upon which to base the standards, even though
                other vehicle size attributes (notably, curb weight) are more strongly
                correlated with fuel economy and tailpipe CO2 emissions.
                 First, mass is strongly correlated with fuel economy; it takes a
                certain amount of energy to move a certain amount of mass. Footprint
                has some positive correlation with frontal surface area, likely a
                negative correlation with aerodynamics, and therefore fuel economy, but
                the relationship is less deterministic. Mass and crush space
                (correlated with footprint) are both important safety considerations.
                As discussed below and in the accompanying PRIA, NHTSA's research of
                historical crash data indicates that holding footprint constant, and
                decreasing the mass of the largest vehicles, will result in a net
                positive safety impact to drivers overall, while holding footprint
                constant and decreasing the mass of the smallest vehicles will result
                in a net decrease in fleetwide safety. Properly fitted footprint-based
                standards provide little, if any, incentive to build smaller footprint
                vehicles to meet CAFE and CO2 standards, and therefore help
                minimize the impact of standards on overall fleet safety.
                 Second, it is important that the attribute not be easily
                manipulated in a manner that does not achieve the goals of EPCA or
                other goals, such as safety. Although weight is more strongly
                correlated with fuel economy than footprint, there is less risk of
                artificial manipulation (i.e., changing the attribute(s) to achieve a
                more favorable target) by increasing footprint under footprint-based
                standards than there would be by increasing vehicle mass under weight-
                based standards. It is relatively easy for a manufacturer to add enough
                weight to a vehicle to decrease its applicable fuel economy target a
                significant amount, as compared to increasing vehicle footprint, which
                is a much more complicated change that typically takes place only with
                a vehicle redesign.
                 Further, some commenters on the MY 2011 CAFE rulemaking were
                concerned that there would be greater potential for such manipulation
                under multi-attribute standards, such as those that also depend on
                weight, torque, power, towing capability, and/or off-road capability.
                As discussed in NHTSA's MY 2011 CAFE final rule,\241\ it is anticipated
                that the possibility of manipulation is lowest with footprint-based
                standards, as opposed to weight-based or multi-attribute-based
                standards. Specifically, standards that incorporate weight, torque,
                power, towing capability, and/or off-road capability in addition to
                footprint would not only be more complex, but by providing degrees of
                freedom with respect to more easily adjusted attributes, they could
                make it less certain that the future fleet would actually achieve the
                projected average fuel economy and CO2 levels. This is not
                to say that a footprint-based system eliminates manipulation, or that a
                footprint-based system eliminates the possibility that manufacturers
                will change vehicles in ways that compromise occupant protection, but
                footprint-based standards achieve the best balance among affected
                considerations.
                ---------------------------------------------------------------------------
                 \241\ See 74 FR at 14359 (Mar. 30, 2009).
                ---------------------------------------------------------------------------
                 Several stakeholders commented on whether vehicular footprint is
                the most suitable attribute upon which to base standards. IPI commented
                that ``. . . footprint-based standards may be unnecessary to respect
                consumer preferences, may negatively impact safety, and may be overall
                inefficient. Several arguments call into question the footprint-based
                approach, but a particularly important one is that large vehicles can
                impose a negative safety externality on other drivers.'' \242\ IPI
                commented, further, that the agencies should consider the relative
                merits of other vehicle attributes, including vehicle fuel type,
                suggesting that it would be more difficult for manufacturers to
                manipulate a flatter standard or one ``differentiated by fuel type.''
                \243\ Similarly, Michalek and Whitefoot recommended ``that the agencies
                reexamine automaker response to the footprint-based standards to
                determine if adjustments should be made to avoid inducing increases to
                vehicle size.'' \244\
                ---------------------------------------------------------------------------
                 \242\ IPI, NHTSA-2018-0067-12362, at 12.
                 \243\ IPI, NHTSA-2018-0067-12362, at 13 et seq.
                 \244\ Michalek, J. and Whitefoot, K., NHTSA-2018-0067-11903, at
                13.
                ---------------------------------------------------------------------------
                [[Page 24251]]
                 Conversely, ICCT commented that ``the switch to footprint-based
                CAFE and [CO2] standards has been widely credited with
                diminishing safety concerns with efficiency standards. Footprint
                standards encourage larger vehicles with wider track width, which
                reduces rollovers, and longer wheelbase, which increases the crush
                space and reduces deceleration forces for both vehicles in a two-
                vehicle collision.'' \245\ Similarly, BorgWarner commented that ``the
                use of a footprint standard not only provides greater incentive for
                mass reduction, but also encourages a larger footprint for a given
                vehicle mass, thus providing increased safety for a given mass
                vehicle,'' \246\ and the Aluminum Association commented footprint based
                standards drive ``fuel-efficiency improvement across all vehicle
                classes,'' ``eliminate the incentive to shift fleet volume to smaller
                cars which has been shown to slightly decrease safety in vehicle-to-
                vehicle collisions,'' and provide ``an incentive for reducing weight in
                the larger vehicles, where weight reduction is of the most benefit for
                societal safety,'' citing Ford's aluminum-intensive F150 pickup truck
                as an example.\247\ NADA urged the agencies to continue basing
                standards on vehicle footprint, as doing so ``serves both to require
                and allow OEMs to build more fuel-efficient vehicles across the
                broadest possible light-duty passenger car and truck spectrum,'' \248\
                and UCS commented that footprint-based standards ``increase consumer
                choice, ensuring that the vehicles available for purchase in every
                vehicle class continue to get more efficient.'' \249\ Furthermore,
                regarding concerns that footprint-based standards may be susceptible to
                manipulation, the Alliance commented that ``the data above [from
                Novation Analytics] shows there are no systemic footprint increases (or
                any type of target manipulation) occurring.'' \250\ While FCA's
                comments supported this Alliance comment, FCA commented further that,
                lacking some utility-related vehicle attributes such as towing
                capability, 4-wheel-drive, and ride height, ``it is clear the footprint
                standard does not fully account for pickup truck capability and the
                components needed such as larger powertrains, greater mass and frontal
                area,'' and requested the agencies ``correct LDT standards to reflect
                the current market preference for capability over efficiency, and
                introduce mechanisms into the regulation that can adjust for efficiency
                and capability tradeoffs that footprint standards currently ignore.''
                \251\
                ---------------------------------------------------------------------------
                 \245\ ICCT, NHTSA-2018-0067-11741, at B-4.
                 \246\ BorgWarner, NHTSA-2018-0067-11893, at 10.
                 \247\ Aluminum Association, NHTSA-2018-0067-11952, at 3.
                 \248\ NADA, NHTSA-2018-0067-12064, at 13.
                 \249\ UCS, UCS, NHTSA-2018-0067-12039, at 46.
                 \250\ Alliance, NHTSA-2018-0067-12073, at 123.
                 \251\ FCA, NHTSA-2018-0067-11943, at 49.
                ---------------------------------------------------------------------------
                 When first electing to adopt footprint-based standards, NHTSA
                carefully considered other alternatives, including vehicle mass and
                ``shadow'' (overall width multiplied by overall length). Compared to
                both of these other alternatives, footprint is much less susceptible to
                gaming, because while there is some potential to adjust track width,
                wheelbase is more expensive to change, at least outside a planned
                vehicle redesign. EPA agreed with NHTSA's assessment, nothing has
                changed the relative merits of at least these three potential
                attributes, and nothing in the evolution of the fleet demonstrates that
                footprint-based standards are leading manufacturers to increase the
                footprint of specific vehicle models by more than they would in
                response to customer demand. Also, even if footprint-based standards
                are encouraging some increases in vehicle size, NHTSA continues to
                maintain, and EPA to agree, that such increases should tend to improve
                overall highway safety rather than degrading it. Regarding FCA's
                request that the agencies adopt an approach that accounts for a wider
                range of vehicle attributes related to both vehicle fuel economy and
                customer-facing vehicle utility, the agencies are concerned that doing
                so could further complicate already-complex standards and also lead to
                unintended consequences. For example, it is not currently clear how a
                multi-attribute approach would appropriately balance emphasis between
                vehicle attributes (e.g., how much relative fuel consumption should be
                attributed to, respectively, vehicle footprint, towing capacity, drive
                type, and ground clearance). Also, basing standards on, in part, ground
                clearance would encourage manufacturers to increase ride height,
                potentially increasing the frequency of vehicle rollover crashes.
                Regarding IPI's recommendation that fuel type be included as a vehicle
                attribute for attribute-based standards, the agencies note that both
                CAFE and CO2 standards already account for fuel type in the
                procedures for measuring fuel economy levels and CO2
                emission rates, and for calculating fleet average CAFE and
                CO2 levels.
                 Therefore, having considered public comments on the choice of
                vehicle attributes for CAFE and CO2 standards, the agencies
                are finalizing standards that, as proposed, are defined in terms of
                vehicle footprint.
                3. What mathematical function should be used to specify footprint-based
                standards?
                 In requiring NHTSA to ``prescribe by regulation separate average
                fuel economy standards for passenger and non-passenger automobiles
                based on 1 or more vehicle attributes related to fuel economy and
                express each standard in the form of a mathematical function,'' EPCA/
                EISA provides ample discretion regarding not only the selection of the
                attribute(s), but also regarding the nature of the function. The CAA
                provides no specific direction regarding CO2 regulation, and
                EPA has continued to harmonize this aspect of its CO2
                regulations with NHTSA's CAFE regulations. The relationship between
                fuel economy (and CO2 emissions) and footprint, though
                directionally clear (i.e., fuel economy tends to decrease and
                CO2 emissions tend to increase with increasing footprint),
                is theoretically vague, and quantitatively uncertain; in other words,
                not so precise as to a priori yield only a single possible curve.
                 The decision of how to specify this mathematical function therefore
                reflects some amount of judgment. The function can be specified with a
                view toward achieving different environmental and petroleum reduction
                goals, encouraging different levels of application of fuel-saving
                technologies, avoiding any adverse effects on overall highway safety,
                reducing disparities of manufacturers' compliance burdens, and
                preserving consumer choice, among other aims. The following are among
                the specific technical concerns and resultant policy tradeoffs the
                agencies have considered in selecting the details of specific past and
                future curve shapes:
                 Flatter standards (i.e., curves) increase the risk that
                both the size of vehicles will be reduced, potentially compromising
                highway safety, and reducing any utility consumers would have gained
                from a larger vehicle.
                 Steeper footprint-based standards may create incentives to
                upsize vehicles, potentially oversupplying vehicles of certain
                footprints beyond what consumers would naturally demand, and thus
                increasing the possibility that fuel savings and CO2
                reduction benefits will be forfeited artificially.
                 Given the same industry-wide average required fuel economy
                or CO2 standard, flatter standards tend to place greater
                compliance burdens on full-line manufacturers.
                 Given the same industry-wide average required fuel economy
                or CO2
                [[Page 24252]]
                standard, dramatically steeper standards tend to place greater
                compliance burdens on limited-line manufacturers (depending of course,
                on which vehicles are being produced).
                 If cutpoints are adopted, given the same industry-wide
                average required fuel economy, moving small-vehicle cutpoints to the
                left (i.e., up in terms of fuel economy, down in terms of
                CO2 emissions) discourages the introduction of small
                vehicles, and reduces the incentive to downsize small vehicles in ways
                that could compromise overall highway safety.
                 If cutpoints are adopted, given the same industry-wide
                average required fuel economy, moving large-vehicle cutpoints to the
                right (i.e., down in terms of fuel economy, up in terms of
                CO2 emissions) better accommodates the design requirements
                of larger vehicles--especially large pickups--and extends the size
                range over which downsizing is discouraged.
                4. What mathematical functions have been used previously, and why?
                 Notwithstanding the aforementioned discretion under EPCA/EISA, data
                should inform consideration of potential mathematical functions, but
                how relevant data is defined and interpreted, and the choice of
                methodology for fitting a curve to that data, can and should include
                some consideration of specific policy goals. This section summarizes
                the methodologies and policy concerns that were considered in
                developing previous target curves (for a complete discussion see the
                2012 FRIA).
                 As discussed below, the MY 2011 final curves followed a constrained
                logistic function defined specifically in the final rule.\252\ The MYs
                2012-2021 final standards and the MYs 2022-2025 augural standards are
                defined by constrained linear target functions of footprint, as shown
                below: \253\
                ---------------------------------------------------------------------------
                 \252\ See 74 FR 14196, 14363-14370 (Mar. 30, 2009) for NHTSA
                discussion of curve fitting in the MY 2011 CAFE final rule.
                 \253\ The right cutpoint for the light truck curve was moved
                further to the right for MYs 2017-2021, so that more possible
                footprints would fall on the sloped part of the curve. In order to
                ensure that, for all possible footprints, future standards would be
                at least as high as MY 2016 levels, the final standards for light
                trucks for MYs 2017-2021 is the maximum of the MY 2016 target curves
                and the target curves for the give MY standard. This is defined
                further in the 2012 final rule. See 77 FR 62624, at 62699-700 (Oct.
                15, 2012).
                [GRAPHIC] [TIFF OMITTED] TR30AP20.057
                 Here, Target is the fuel economy target applicable to vehicles
                of a given footprint in square feet (Footprint). The upper
                asymptote, a, and the lower asymptote, b, are specified in mpg; the
                reciprocal of these values represent the lower and upper asymptotes,
                respectively, when the curve is instead specified in gallons per
                mile (gpm). The slope, c, and the intercept, d, of the linear
                portion of the curve are specified as gpm per change in square feet,
                ---------------------------------------------------------------------------
                and gpm, respectively.
                 The min and max functions will take the minimum and maximum values
                within their associated parentheses. Thus, the max function will first
                find the maximum of the fitted line at a given footprint value and the
                lower asymptote from the perspective of gpm. If the fitted line is
                below the lower asymptote it is replaced with the floor, which is also
                the minimum of the floor and the ceiling by definition, so that the
                target in mpg space will be the reciprocal of the floor in mpg space,
                or simply, a. If, however, the fitted line is not below the lower
                asymptote, the fitted value is returned from the max function and the
                min function takes the minimum value of the upper asymptote (in gpm
                space) and the fitted line. If the fitted value is below the upper
                asymptote, it is between the two asymptotes and the fitted value is
                appropriately returned from the min function, making the overall target
                in mpg the reciprocal of the fitted line in gpm. If the fitted value is
                above the upper asymptote, the upper asymptote is returned is returned
                from the min function, and the overall target in mpg is the reciprocal
                of the upper asymptote in gpm space, or b.
                 In this way curves specified as constrained linear functions are
                specified by the following parameters:
                a = upper limit (mpg)
                b = lower limit (mpg)
                c = slope (gpm per sq.ft.)
                d = intercept (gpm)
                 The slope and intercept are specified as gpm per sq. ft. and gpm
                instead of mpg per sq. ft. and mpg because fuel consumption and
                emissions appear roughly linearly related to gallons per mile (the
                reciprocal of the miles per gallon).
                a) NHTSA in MY 2008 and MY 2011 CAFE (Constrained Logistic)
                 For the MY 2011 CAFE rule, NHTSA estimated fuel economy levels by
                footprint from the MY 2008 fleet after normalization for differences in
                technology,\254\ but did not make adjustments to reflect other vehicle
                attributes (e.g., power-to-weight ratios). Starting with the
                technology-adjusted passenger car and light truck fleets, NHTSA used
                minimum absolute deviation (MAD) regression without sales weighting to
                fit a logistic form as a starting point to develop mathematical
                functions defining the standards. NHTSA then identified footprints at
                which to apply minimum and maximum values (rather than letting the
                standards extend without limit) and transposed these functions
                vertically (i.e., on a gallons-per-mile basis, uniformly downward) to
                produce the promulgated standards. In the preceding rule, for MYs 2008-
                2011 light truck standards, NHTSA examined a range of potential
                functional forms, and concluded that, compared to other considered
                forms, the constrained logistic form provided the expected and
                appropriate trend (decreasing fuel economy as footprint increases), but
                avoided creating ``kinks'' the agency was concerned would provide
                distortionary incentives for vehicles with neighboring footprints.\255\
                ---------------------------------------------------------------------------
                 \254\ See 74 FR 14196, 14363-14370 (Mar. 30, 2009) for NHTSA
                discussion of curve fitting in the MY 2011 CAFE final rule.
                 \255\ See 71 FR 17556, 17609-17613 (Apr. 6, 2006) for NHTSA
                discussion of ``kinks'' in the MYs 2008-2011 light truck CAFE final
                rule (there described as ``edge effects''). A ``kink,'' as used
                here, is a portion of the curve where a small change in footprint
                results in a disproportionally large change in stringency.
                ---------------------------------------------------------------------------
                b) MYs 2012-2016 Standards (Constrained Linear)
                 For the MYs 2012-2016 rule, potential methods for specifying
                mathematical functions to define fuel economy and CO2
                standards were reevaluated. These methods were fit to the same MY 2008
                data as the MY 2011 standard. Considering these further specifications,
                the constrained logistic form, if applied to post-MY 2011 standards,
                would likely contain a steep mid-section that would provide undue
                incentive to increase the footprint of midsize passenger cars.\256\ A
                range of
                [[Page 24253]]
                methods to fit the curves would have been reasonable, and a minimum
                absolute deviation (MAD) regression without sales weighting on a
                technology-adjusted car and light truck fleet was used to fit a linear
                equation. This equation was used as a starting point to develop
                mathematical functions defining the standards. Footprints were then
                identified at which to apply minimum and maximum values (rather than
                letting the standards extend without limit). Finally, these
                constrained/piecewise linear functions were transposed vertically
                (i.e., on a gpm or CO2 basis, uniformly downward) by
                multiplying the initial curve by a single factor for each MY standard
                to produce the final attribute-based targets for passenger cars and
                light trucks described in the final rule.\257\ These transformations
                are typically presented as percentage improvements over a previous MY
                target curve.
                ---------------------------------------------------------------------------
                 \256\ 75 FR at 25362.
                 \257\ See generally 74 FR at 49491-96; 75 FR at 25357-62.
                ---------------------------------------------------------------------------
                c) MYs 2017 and Beyond Standards (Constrained Linear)
                 The mathematical functions finalized in 2012 for MYs 2017 and
                beyond changed somewhat from the functions for the MYs 2012-2016
                standards. These changes were made both to address comments from
                stakeholders, and to consider further some of the technical concerns
                and policy goals judged more preeminent under the increased uncertainty
                of the impacts of finalizing and proposing standards for model years
                further into the future.\258\ Recognizing the concerns raised by full-
                line OEMs, it was concluded that continuing increases in the stringency
                of the light truck standards would be more feasible if the light truck
                curve for MYs 2017 and beyond was made steeper than the MY 2016 truck
                curve and the right (large footprint) cut-point was extended only
                gradually to larger footprints. To accommodate these considerations,
                the 2012 final rule finalized the slope fit to the MY 2008 fleet using
                a sales-weighted, ordinary least-squares regression, using a fleet that
                had technology applied to make the technology application across the
                fleet more uniform, and after adjusting the data for the effects of
                weight-to-footprint. Information from an updated MY 2010 fleet was also
                considered to support this decision. As the curve was vertically
                shifted (with fuel economy specified as mpg instead of gpm or
                CO2 emissions) upwards, the right cutpoint was progressively
                moved for the light truck curves with successive model years, reaching
                the final endpoint for MY 2021.
                ---------------------------------------------------------------------------
                 \258\ The MYs 2012-2016 final standards were signed April 1st,
                2010--putting 6.5 years between its signing and the last affected
                model year, while the MYs 2017-2021 final standards were signed
                August 28th, 2012--giving just more than nine years between signing
                and the last affected final standards.
                ---------------------------------------------------------------------------
                5. Reconsidering the Mathematical Functions for Today's Rulemaking
                a) Why is it important to reconsider the mathematical functions?
                 By shifting the developed curves by a single factor, it is assumed
                that the underlying relationship of fuel consumption (in gallons per
                mile) to vehicle footprint does not change significantly from the model
                year data used to fit the curves to the range of model years for which
                the shifted curve shape is applied to develop the standards. However,
                it must be recognized that the relationship between vehicle footprint
                and fuel economy is not necessarily constant over time; newly developed
                technologies, changes in consumer demand, and even the curves
                themselves could influence the observed relationships between the two
                vehicle characteristics. For example, if certain technologies are more
                effective or more marketable for certain types of vehicles, their
                application may not be uniform over the range of vehicle footprints.
                Further, if market demand has shifted between vehicle types, so that
                certain vehicles make up a larger share of the fleet, any underlying
                technological or market restrictions which inform the average shape of
                the curves could change. That is, changes in the technology or market
                restrictions themselves, or a mere re-weighting of different vehicles
                types, could reshape the fit curves.
                 For the above reasons, the curve shapes were reconsidered in the
                proposal using the newest available data from MY 2016. With a view
                toward corroboration through different techniques, a range of
                descriptive statistical analyses were conducted that do not require
                underlying engineering models of how fuel economy and footprint might
                be expected to be related, and a separate analysis that uses vehicle
                simulation results as the basis to estimate the relationship from a
                perspective more explicitly informed by engineering theory was
                conducted as well. Despite changes in the new vehicle fleet both in
                terms of technologies applied and in market demand, the underlying
                statistical relationship between footprint and fuel economy has not
                changed significantly since the MY 2008 fleet used for the 2012 final
                rule; therefore, EPA and NHTSA proposed to continue to use the curve
                shapes fit in 2012. The analysis and reasoning supporting this decision
                follows.
                b) What statistical analyses did EPA and NHTSA consider?
                 In considering how to address the various policy concerns discussed
                above, data from the MY 2016 fleet was considered, and a number of
                descriptive statistical analyses (i.e., involving observed fuel economy
                levels and footprints) using various statistical methods, weighting
                schemes, and adjustments to the data to make the fleets less
                technologically heterogeneous were performed. There were several
                adjustments to the data that were common to all of the statistical
                analyses considered.
                 With a view toward isolating the relationship between fuel economy
                and footprint, the few diesels in the fleet were excluded, as well as
                the limited number of vehicles with partial or full electric
                propulsion; when the fleet is normalized so that technology is more
                homogenous, application of these technologies is not allowed. This is
                consistent with the methodology used in the 2012 final rule.
                 The above adjustments were applied to all statistical analyses
                considered, regardless of the specifics of each of the methods,
                weights, and technology level of the data, used to view the
                relationship of vehicle footprint and fuel economy. Table V-1, below,
                summarizes the different assumptions considered and the key attributes
                of each. The analysis was performed considering all possible
                combinations of these assumptions, producing a total of eight footprint
                curves.
                [[Page 24254]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.058
                (1) Current Technology Level Curves
                 The ``current technology'' level curves exclude diesels and
                vehicles with electric propulsion, as discussed above, but make no
                other changes to each model year fleet. Comparing the MY 2016 curves to
                ones built under the same methodology from previous model year fleets
                shows whether the observed curve shape has changed significantly over
                time as standards have become more stringent. Importantly, these curves
                will include any market forces which make technology application
                variable over the distribution of footprint. These market forces will
                not be present in the ``maximum technology'' level curves: By making
                technology levels homogenous, this variation is removed. The current
                technology level curves built using both regression types and both
                regression weight methodologies from the MY 2008, MY 2010, and MY 2016
                fleets, shown in more detail in Chapter 4.4.2.1 of the PRIA, support
                the curve slopes finalized in the 2012 final rule. The curves built
                from most methodologies using each fleet generally shift, but remain
                very similar in slope. This suggests that the relationship of footprint
                to fuel economy, including both technology and market limits, has not
                significantly changed.
                (2) Maximum Technology Level Curves
                 As in prior rulemakings, technology differences between vehicle
                models were considered to be a significant factor producing uncertainty
                regarding the relationship between fuel consumption and footprint.
                Noting that attribute-based standards are intended to encourage the
                application of additional technology to improve fuel efficiency and
                reduce CO2 emissions across the distribution of footprint in
                the fleet, approaches were considered in which technology application
                is simulated for purposes of the curve fitting analysis in order to
                produce fleets that are less varied in technology content. This
                approach helps reduce ``noise'' (i.e., dispersion) in the plot of
                vehicle footprints and fuel consumption levels and identify a more
                technology-neutral relationship between footprint and fuel consumption.
                The results of updated analysis for maximum technology level curves are
                also shown in Chapter 4.4.2.2 of the PRIA. Especially if vehicles
                progress over time toward more similar size-specific efficiency,
                further removing variation in technology application both better
                isolates the relationship between fuel consumption and footprint and
                further supports the curve slopes finalized in the 2012 final rule.
                c) What other methodologies were considered?
                 The methods discussed above are descriptive in nature, using
                statistical analysis to relate observed fuel economy levels to observed
                footprints for known vehicles. As such, these methods are clearly based
                on actual data, answering the question ``how does fuel economy appear
                to be related to footprint?'' However, being independent of explicit
                engineering theory, they do not answer the question ``how might one
                expect fuel economy to be related to footprint?'' Therefore, as an
                alternative to the above methods, an alternative methodology was also
                developed and applied that, using full-vehicle simulation, comes closer
                to answering the second question, providing a basis either to
                corroborate answers to the first, or suggest that further investigation
                could be important.
                 As discussed in the 2012 final rule, several manufacturers have
                confidentially shared with the agencies what they described as
                ``physics-based'' curves, with each OEM showing significantly different
                shapes for the footprint-fuel economy relationships. This variation
                suggests that manufacturers face different curves given the other
                attributes of the vehicles in their fleets (i.e., performance
                [[Page 24255]]
                characteristics) and/or that their curves reflected different levels of
                technology application. In reconsidering the shapes of the proposed MYs
                2021-2026 standards, a similar estimation of physics-based curves
                leveraging third-party simulation work form Argonne National
                Laboratories (Argonne) was developed. Estimating physics-based curves
                better ensures that technology and performance are held constant for
                all footprints; augmenting a largely statistical analysis with an
                analysis that more explicitly incorporates engineering theory helps to
                corroborate that the relationship between fuel economy and footprint is
                in fact being characterized.
                 Tractive energy is the amount of energy it will take to move a
                vehicle.\259\ Here, tractive energy effectiveness is defined as the
                share of the energy content of fuel consumed which is converted into
                mechanical energy and used to move a vehicle--for internal combustion
                engine (ICE) vehicles, this will vary with the relative efficiency of
                specific engines. Data from Argonne simulations suggest that the limits
                of tractive energy effectiveness are approximately 25 percent for
                vehicles with internal combustion engines which do not possess
                integrated starter generator, other hybrid, plug-in, pure electric, or
                fuel cell technology.
                ---------------------------------------------------------------------------
                 \259\ Thomas, J. ``Drive Cycle Powertrain Efficiencies and
                Trends Derived from EPA Vehicle Dynamometer Results,'' SAE Int. J.
                Passeng. Cars--Mech. Syst. 7(4):2014, doi:10.4271/2014-01-2562.
                Available at https://www.sae.org/publications/technical-papers/content/2014-01-2562/ (last accessed June 15, 2018).
                ---------------------------------------------------------------------------
                 A tractive energy prediction model was also developed to support
                today's proposal. Given a vehicle's mass, frontal area, aerodynamic
                drag coefficient, and rolling resistance as inputs, the model will
                predict the amount of tractive energy required for the vehicle to
                complete the Federal test cycle. This model was used to predict the
                tractive energy required for the average vehicle of a given footprint
                \260\ and ``body technology package'' to complete the cycle. The body
                technology packages considered are defined in Table V-2, below. Using
                the absolute tractive energy predicted and tractive energy
                effectiveness values spanning possible ICE engines, fuel economy values
                were then estimated for different body technology packages and engine
                tractive energy effectiveness values.
                ---------------------------------------------------------------------------
                 \260\ The mass reduction curves used elsewhere in this analysis
                were used to predict the mass of a vehicle with a given footprint,
                body style box, and mass reduction level. The `Body style Box' is 1
                for hatchbacks and minivans, 2 for pickups, and 3 for sedans, and is
                an important predictor of aerodynamic drag. Mass is an essential
                input in the tractive energy calculation.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.059
                 Chapter 6 of the PRIA show the resultant CAFE levels estimated for
                the vehicle classes Argonne simulated for this analysis, at different
                footprint values and by vehicle ``box.'' Pickups are considered 1-box,
                hatchbacks and minivans are 2-box, and sedans are 3-box. These
                estimates are compared with the MY 2021 standards finalized in 2012.
                The general trend of the simulated data points follows the pattern of
                the previous MY 2021 standards for all technology packages and tractive
                energy effectiveness values presented in the PRIA. The tractive energy
                curves are intended to validate the curve shapes against a physics-
                based alternative, and the analysis suggests that the curve shapes
                track the physical relationship between fuel economy and tractive
                energy for different footprint values.
                 Physical limitations are not the only forces manufacturers face;
                their success is dependent upon producing vehicles that consumers
                desire and will purchase. For this reason, in setting future standards,
                the analysis will continue to consider information from statistical
                analyses that do not homogenize technology applications in addition to
                statistical analyses which do, as well as a tractive energy analysis
                similar to the one presented above.
                 The relationship between fuel economy and footprint remains
                directionally discernable but quantitatively uncertain. Nevertheless,
                each standard must commit to only one function. Approaching the
                question ``how is fuel economy related to footprint'' from different
                directions and applying different approaches has given EPA and NHTSA
                confidence that the function applied here appropriately and reasonably
                reflects the relationship between fuel economy and footprint.
                 The agencies invited comments on this conclusion and the supporting
                analysis. IPI raised concerns that ``. . . several dozen models (mostly
                subcompacts and sports cars) fall in the 30-40 square feet range, which
                are all subject to the same standards'' and that ``manufacturers of
                these models may have an incentive to decrease footprints as a
                compliance strategy, since doing so would not trigger more stringent
                standards.'' \261\ NHTSA and EPA agree that, all else equal, downsizing
                the smallest cars (e.g., Chevrolet Spark, Ford Fiesta, Mini Cooper,
                Mazda MX-5, Porsche 911, Toyota Yaris) would most likely tend to
                degrade overall highway safety. At the same time, as discussed above,
                the agencies recognize that small vehicles do appear attractive to some
                market segments (although obviously the Ford Fiesta and Porsche 911
                compete in different segments).
                [[Page 24256]]
                Therefore, there is a tension between on one hand, avoiding standards
                that unduly encourage safety-eroding downsizing and, on the other,
                avoiding standards that unduly penalize the market for small vehicles.
                The agencies examined this issue, and note that the market for the
                smallest vehicles has not evolved at all as estimated in the analysis
                supporting the 2012 final rule, and attribute this more to fuel prices
                and consumer demand for larger vehicles than to attribute-based CAFE
                and CO2 standards. For example, the market for vehicles with
                footprints less than 40 square foot was about 45 percent smaller in MY
                2017 than in MY 2010. The agencies also found that among the smallest
                vehicle models produced throughout MYs 2010-2017, most have become
                larger, not smaller. For example, while the Mazda MX-5's footprint
                decreased by 0.1 square foot (0.3 percent) during that time, the MY
                2017 versions of the Mini Cooper, Smart fortwo, Porsche 911, and Toyota
                Yaris had larger footprints than in MY 2010. With the market for very
                small vehicles shrinking, and with manufacturers not evidencing a
                tendency to make the smallest vehicles even smaller, the agencies are
                satisfied that it would be unwise to change the target functions such
                that targets never stop becoming more stringent as vehicle footprint
                becomes ever smaller, because doing so could further impede an already-
                shrinking market.
                ---------------------------------------------------------------------------
                 \261\ IPI, NHTSA-2018-0067-12362, p. 14.
                ---------------------------------------------------------------------------
                B. No-Action Alternative
                 As in the proposal, the No-Action Alternative applies the augural
                CAFE and final CO2 targets announced in 2012 for MYs 2021-
                2025.\262\ For MY 2026, this alternative applies the same targets as
                for MY 2025. The carbon dioxide equivalent of air conditioning
                refrigerant leakage credits, nitrous oxide, and methane emissions are
                included for compliance with the EPA standards for all model years
                under the no-action alternative.\263\
                ---------------------------------------------------------------------------
                 \262\ https://www.govinfo.gov/content/pkg/CFR-2014-title40-vol19/pdf/CFR-2014-title40-vol19-sec86-1818-12.pdf
                 \263\ EPA regulations use a different but mathematically
                equivalent approach to specify targets. Rather than using a function
                with nested minima and maxima functions, EPA regulations specify
                requirements separately for different ranges of vehicle footprint.
                Because these ranges reflect the combined application of the listed
                minima, maxima, and linear functions, it is mathematically
                equivalent and more efficient to present the targets as in this
                Section.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.060
                [[Page 24257]]
                 In comments on the DEIS, CBD et al. indicated that it was
                appropriate for NHTSA to use the augural CAFE standards as the baseline
                No Action regulatory alternative.\264\ However, CARB commented that the
                baseline regulatory alternative should include CARB's ZEV mandate, in
                part because EPA must consider ``other regulations promulgated by EPA
                or other government entities,'' and, according to CARB, there will be
                much more vehicle electrification in the future as manufacturers
                respond to market demand and also work to comply with the ZEV
                mandate.\265\ Similarly, EPA's Science Advisory Board recommended--
                despite the action taken in the One National Program Action--that the
                baseline include state ZEV mandates ``to be consistent with policies
                that would prevail in the absence of the rule change.'' \266\ EPA's
                Science Advisory Board further recommended including sensitivity
                analyses with different penetration rates of ZEVs.
                ---------------------------------------------------------------------------
                 \264\ CBD et al., NHTSA-2018-0067-12123, Attachment 1, at 13.
                 \265\ CARB, NHTSA-2018-0067-11873, at 124-125.
                 \266\ SAB at 12 and 29-30.
                ---------------------------------------------------------------------------
                 On the other hand, arguing for consideration of standards less
                stringent than those proposed in the NPRM, Walter Kreucher commented
                that rather than using the augural standards as the baseline, ``a
                better approach would be to assume a clean sheet of paper and start
                from the existing 2016MY fleet and its associated standards as the
                baseline using 0%/year increases for both passenger cars and light
                trucks for MYs 2017-2026.'' \267\ Similarly, AVE argued that because
                previously-promulgated standards for MYs 2018-2021 already present a
                significant challenge that ``will likely require almost every automaker
                to continue using credits for compliance, . . . AVE believes this
                rulemaking should reset . . . the current compliance baseline for cars
                and light trucks at MY 2018 . . .'' \268\ BorgWarner commented
                similarly that ``Beginning in MY 2018, standards should be reset to the
                levels the industry actually achieved. For MY 2018 and beyond,
                succeeding model year targets should be set with an annual rate of
                improvement defined by the slope of improvement the industry has
                achieved over the last six years. . . . Based on these data, our
                analysis suggests the most reasonable and logical rate of improvement
                falls between 2.0% to 2.6% for cars and trucks. Additionally, a single
                rate of improvement for the combined fleet should be considered.''
                \269\
                ---------------------------------------------------------------------------
                 \267\ Kreucher, W., NHTSA-2018-0067-0444, at 8.
                 \268\ AVE, NHTSA-2018-0067-11696, at 8-9.
                 \269\ BorgWarner, NHTSA-2018-0067-11895, at 3, 6.
                ---------------------------------------------------------------------------
                 The No-Action Alternative represents expectations regarding the
                world in the absence of a proposal, accounting for applicable laws
                already in place. Although manufacturers are already making significant
                use of compliance credits toward compliance with even MY 2017
                standards, the agencies are obligated to evaluate regulatory
                alternatives against the standards already in place through MY 2025.
                Similarly, even though manufacturers are already producing electric
                vehicles, EPA and NHTSA appropriately excluded California's ZEV mandate
                from the No-Action alternative for the NPRM, for several reasons.
                First, the ZEV mandate is not Federal law; second, as described in the
                proposal and subsequently finalized in regulatory text, the ZEV mandate
                is expressly and impliedly preempted by EPCA; third, EPA proposed to
                withdraw the waiver of CAA preemption in the NPRM and subsequently
                finalized this withdrawal. Accordingly, the agencies have, therefore,
                appropriately excluded the ZEV mandate from the No-Action alternative.
                However, as discussed below, the agencies' analysis does account for
                the potential that under every regulatory alternative, including the
                No-Action Alternative, vehicle electrification could increase in the
                future, especially if batteries become less expensive as gasoline
                becomes more expensive.
                C. Action Alternatives
                1. Alternatives in Final Rule
                 Table V-5 below shows the different alternatives evaluated in
                today's notice.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.061
                [[Page 24258]]
                 With one exception, the alternatives considered in the NPRM
                included the changes in stringency for the above alternatives.
                Alternative 3, the preferred alternative, is newly included for today's
                notice.\270\
                ---------------------------------------------------------------------------
                 \270\ As the agencies indicated in the NPRM, they were
                considering and taking comment ``on a wide range of alternatives and
                have specifically modeled eight alternatives.'' 83 FR at 42990 (Aug.
                24, 2018). The preferred alternative in this final rule was within
                the range of alternatives considered in the proposal, although it
                was not specifically modeled at that time. This issue is discussed
                in further detail below.
                ---------------------------------------------------------------------------
                 Regulations regarding implementation of NEPA requires agencies to
                ``rigorously explore and objectively evaluate all reasonable
                alternatives, and for alternatives which were eliminated from detailed
                study, briefly discuss the reasons for their having been eliminated.''
                \271\ This does not amount to a requirement that agencies evaluate the
                widest conceivable spectrum of alternatives. For example, a State
                considering adding a single travel lane to a preexisting section of
                highway would not be required to consider adding three lanes, or to
                consider dismantling the highway altogether.
                ---------------------------------------------------------------------------
                 \271\ 40 CFR 1502.14.
                ---------------------------------------------------------------------------
                 Among thousands of individual comments that mentioned the proposed
                standards very generally, some comments addressed the range and
                definition of these regulatory alternatives in specific terms, and
                these specific comments include comments on the stringency, structure,
                and particular provisions defining the set of regulatory alternatives
                under consideration.
                 As discussed throughout today's notice, the agencies have updated
                and otherwise revised many aspects of the analysis. The agencies have
                also reconsidered whether the set of alternatives studied in detail
                should be expanded to include standards less stringent than the
                proposal's preferred alternative, or to include standards more
                stringent than the proposal's no-action alternative. On one hand,
                comments from Walter Kreucher and AVE cited above indicate the agencies
                should consider relaxing standards below MY 2020 levels, and CEI
                challenged the agencies' failure to include less-stringent alternatives
                in the following comments on this question:
                 DOT failed to consider the possibility of freezing CAFE at an
                even more lenient standard than currently exists, nor did it
                consider making its proposed freeze take effect sooner than MY 2020.
                However, as DOT's own analysis strongly indicates, doing so would
                lead to even greater benefits and an even greater reduction in CAFE-
                related deaths and injuries. In short, DOT's failure to consider
                this possibility is arbitrary and capricious. It has an opportunity
                to remedy this in its final rule, and it should do so by selecting a
                standard that is even more lenient than the one it proposed. . . .
                It should have gone beyond its original set of alternatives and
                examined less stringent ones as well--until it found one that, for
                some reason or another, failed to produce greater safety benefits or
                failed to meet the statutory factors.\272\
                ---------------------------------------------------------------------------
                 \272\ CEI, NHTSA-2018-0067-12015, at 1.
                 On the other hand, a coalition of ten environmental advocacy
                organizations stated that the agencies should consider alternatives
                more stringent than those defining the baseline no action alternative,
                arguing that in light of CEQ guidance and the 2018 IPCC report on
                climate change, ``the increasing danger, increasing urgency, and
                increasing importance of vehicle emissions all rationally counsel for
                strengthening emission standards.'' \273\ CBD et al. observe that
                ``none of these alternatives [considered in the NPRM] increases fuel
                economy in comparison with the No Action Alternative, none conserves
                energy . . .'' and go on to assert that ``none represents maximum
                feasible CAFE standards.'' \274\ Similarly, EDF commented that ``. . .
                given its clear statutory directive to maximize fuel savings, NHTSA
                should have considered a range of alternatives that would be more
                protective than the existing standards,'' \275\ and three State
                agencies in Minnesota commented that ``more stringent standards are
                consistent with EPCA's purpose of energy conservation and the CAA's
                purpose of reducing harmful air pollutants.'' \276\ The North Carolina
                Department of Environmental Quality acknowledged the agencies'
                determination in the proposal that alternatives beyond the augural
                standards might be economically impracticable, but nevertheless argued
                that ``alternatives that exceed the stringency of the current standards
                are consistent with EPCA's purpose'' \277\ In oral testimony before the
                agencies, the New York State Attorney General also indicated that the
                agencies should consider alternatives more stringent than the augural
                standards.\278\ A coalition of States and cities commented that ``at a
                minimum, the existing standards should be left in place, but EPA should
                also consider whether to make the standards more stringent, not less,
                just as it has done in prior proposals.'' \279\ More specifically,
                through International Mosaic, some individuals commented that the
                agencies must ``fully and publicly consider a few options that require
                at least a seven annual percent [sic] improvement in vehicle fleet
                mileage.'' \280\ In comments on the DEIS, CBD, et al. went further,
                commenting that ``NHTSA's most stringent alternative must be set at no
                lower than a 9 percent improvement per year.'' \281\ Most manufacturers
                who commented on stringency did not identify specific regulatory
                alternatives that the agencies should consider, although Honda
                suggested that standards be set to increase in stringency at 5 percent
                annually for both passenger cars and light trucks throughout model
                years 2021-2026.282 283
                ---------------------------------------------------------------------------
                 \273\ CBD, et al., NHTSA-2018-0067-12057 p. 10. Also, see
                comments from Senator Tom Carper, NHTSA-2018-0067-11910, at 8-9, and
                from UCS, NHTSA-2018-0067-12039, at 3.
                 \274\ CBD, et al., NHTSA-2018-0067-12123, at 12-13.
                 \275\ EDF, NHTSA-2018-0067-11996, at 20.
                 \276\ Minnesota Pollution Control Agency, Department of
                Transportation, and Department of Health, NHTSA-2018-0067-11706, at
                5.
                 \277\ North Carolina Department of Environmental Quality, NHTSA-
                2018-0067-12025, at 37-38.
                 \278\ New York State Attorney General, Testimony of Austin
                Thompson, NHTSA-2018-0067-12305, at 13.
                 \279\ NHTSA-2018-0067-11735, at 49.
                 \280\ International Mosaic NHTSA-2018-0067-11154, at 1
                 \281\ CBD, et al., NHTSA-2018-0067-12123, at 17.
                 \282\ Honda, NHTSA-2018-0067-12019, EPA-HQ-OAR-2018-0283, at 54.
                 \283\ In model year 2021, the baseline standards for passenger
                cars and light trucks increase by about 4% and 6.5%, respectively,
                relative to standards for model year 2020. Depending on the
                composition of the future new vehicle fleet (i.e., the footprints
                and relative market shares of passenger cars and light trucks), this
                amounts to an overall average stringency increase of about 5.5%
                relative to model year 2020.
                ---------------------------------------------------------------------------
                 The agencies carefully considered these comments to expand the
                range of stringencies to be evaluated as possible candidates for
                promulgation. To inform this consideration, the agencies used the CAFE
                model to examine a progression of stringencies extending outside the
                range presented in the proposal and draft EIS, and as a point of
                reference, using a case that reverts to MY 2018 standards starting in
                MY 2021. Scenarios included in this initial screening exercise ranged
                as high as increasing annually at 9.5 percent during MYs 2021-2026,
                reaching average CAFE and CO2 requirements of 66 mpg and 120
                g/mi, respectively. Results of this analysis are presented in the
                following tables and charts. Focusing on MY 2029, the tables show
                average required and achieved CAFE (as mpg) and CO2 (as g/
                mi) levels for each scenario, along with average per-vehicle costs (in
                2018 dollars, relative to retaining MY 2017 technologies). The proposed
                (0%/0%), final (1.5%/1.5%), and baseline augural standards are shown in
                bold type. The charts present
                [[Page 24259]]
                the same results on a percentage basis, relative to values shown below
                for the scenario that reverts to MY 2018 standards starting in MY 2021.
                 For example, reverting to the MY 2018 CAFE standards starting in MY
                2021 yields an average CAFE requirement of 35 mpg by MY 2029, with the
                industry exceeding that standard by 5 mpg at an average cost of $1,255
                relative to MY 2017 technology. Under the augural standards, the MY
                2029 requirement increases to 47 mpg, the average compliance margin
                falls to 1 mpg, and the average cost increases to $2,770. In other
                words, compared to the scenario that reverts to MY 2018 stringency
                starting in MY 2021, the augural standards increase stringency by 34
                percent (from 35 to 47 mpg), increase average fuel economy by 20
                percent (from 40 to 48 mpg), and increase costs by 121 percent (from
                $1,255 to $2,770).
                 As indicated in the following two charts, the reality of
                diminishing returns clearly applies in both directions. On one hand,
                relaxing stringency below the proposed standards by reverting to MY
                2018 or MY 2019 standards reduces average MY 2029 costs by only modest
                amounts ($54-$121). As discussed in Section VIII, the agencies' updated
                analysis indicates that the proposed standards would not be maximum
                feasible considering the EPCA/EISA statutory factors, and would not be
                appropriate under the CAA after considering the appropriate factors. If
                further relaxation of standards appeared likely to yield more
                significant cost reductions, it is conceivable that such savings could
                outweigh further foregoing of energy and climate benefits. However,
                this screening analysis does not show dramatic cost reductions.
                Therefore, the agencies did not include these two less stringent
                alternatives in the detailed analysis presented in Section VII.
                 On the other hand, increases in stringency beyond the baseline
                augural standards show relative costs continuing to accrue much more
                rapidly than relative CAFE and CO2 improvements. As
                discussed below in Section VIII, even the no action alternative is
                already well beyond levels that can be supported under the CAA and
                EPCA. If further stringency increases appeared likely to yield more
                significant additional energy and environmental benefits, it is
                conceivable that these could outweigh these significant additional cost
                increases. However, this screening analysis shows no dramatic relative
                acceleration of energy and environmental benefits. Therefore, the
                agencies did not include stringencies beyond the augural standards in
                the detailed analysis presented in Section VII.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.062
                [[Page 24260]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.063
                [GRAPHIC] [TIFF OMITTED] TR30AP20.064
                [[Page 24261]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.065
                BILLING CODE 4910-59-C
                 Specific to model year 2021, some commenters argued that EPCA's
                lead time requirement prohibits NHTSA from revising CAFE standards for
                model year 2021.\284\ Regarding the revision of standards for model
                year 2021, NHTSA did consider EPCA's lead time requirement, and
                determined that while the agency would need to finalize a stringency
                increase at least 18 months before the beginning of the first affected
                model year, the agency can finalize a stringency decrease closer (or
                even after) the beginning of the first affected model year. The
                agency's reasoning is explained further in Section VIII. Therefore,
                NHTSA did not change regulatory alternatives to avoid any relaxation of
                stringency in model year 2021.
                ---------------------------------------------------------------------------
                 \284\ State of California, et al., NHTSA-2018-0067-11735, at
                78.; CBD, et al., NHTSA-2018-0067-12000, Appendix A, at 66.;
                National Coalition for Advanced Transportation, NHTSA-2018-0067-
                11969, at 46.
                ---------------------------------------------------------------------------
                 The Auto Alliance stated that ``the truck increase rate should be
                no greater than the car rate of increase and should be the `equivalent
                task' per fleet.'' \285\ Supporting these Alliance comments, FCA
                elaborated by commenting that ``(1) in MY2017, the latest data we have
                available, most trucks have a larger gap to standards than cars, and
                (2) all of the truck segments are challenged because consumers are
                placing a greater emphasis on capability than fuel economy.'' \286\
                Similarly, Ford commented that ``. . . the rates of increase in the
                stringency of the standards should remain equivalent between passenger
                cars and light duty trucks.'' \287\ Other commenters expressed general
                support for equalizing the rates at which the stringencies of passenger
                car and light truck standards increase.\288\
                ---------------------------------------------------------------------------
                 \285\ Alliance, NHTSA-2018-0067-12073, at 7-8
                 \286\ FCA, NHTSA-2018-0067-11943, at 46-47.
                 \287\ Ford, NHTSA-2018-0067-11928, at 3.
                 \288\ See, e.g., Global, NHTSA-2018-0067-12032, at 4; NADA,
                NHTSA-2018-0067-12064, at 13; BorgWarner, NHTSA-2018-0067-11895, at
                6.
                ---------------------------------------------------------------------------
                 For the final rule, the agencies have added an alternative in which
                stringency for both cars and trucks increases at 1.5 percent. This is
                consistent with comments received requesting that both fleets'
                standards increase in stringency by the same amount, and 1.5 percent
                represents a rate of increase within the range of rates of increase
                considered in the NPRM.
                 Throughout the NPRM, the agencies described their consideration as
                covering a range of alternatives.\289\ The preferred alternative for
                this final rule, an increase in stringency of 1.5 percent for both cars
                and trucks, falls squarely
                [[Page 24262]]
                within the range of alternatives proposed by the agencies.
                ---------------------------------------------------------------------------
                 \289\ 83 FR at 42986 (Aug. 24, 2018) (explaining, in ``Summary''
                section of NPRM, that ``comment is sought on a range of alternatives
                discussed throughout this document''); id. at 42988 (stating that
                the agencies are ``taking comment on a wide range of alternatives,
                including different stringencies and retaining existing
                CO2 standards and the augural CAFE standards''); 42990
                (``As explained above, the agencies are taking comment on a wide
                range of alternatives and have specifically modeled eight
                alternatives (including the proposed alternative) and the current
                requirements (i.e., baseline/no action).''); 43197 (``[T]oday's
                notice also presents the results of analysis estimating impacts
                under a range of other regulatory alternatives the agencies are
                considering.''); 43229 (explaining that ``technology availability,
                development and application, if it were considered in isolation, is
                not necessarily a limiting factor in the Administrator's selection
                of which standards are appropriate within the range of the
                Alternatives presented in this proposal.''); 43369 (``As discussed
                above, a range of regulatory alternatives are being considered.'').
                ---------------------------------------------------------------------------
                 The NPRM alternatives were bounded on the upper end by the
                baseline/no action alternative, and the proposed alternative on the
                lower end (0 percent per year increase in stringency for both cars and
                trucks). For passenger cars, the agencies considered a range of
                stringency increases between 0 percent and 2 percent per year for
                passenger cars, in addition to the baseline/no action alternative. For
                light trucks, the agencies considered a range of stringency increases
                between 0 percent and 3 percent per year, in addition to the baseline/
                no action alternative.
                 The agencies considered the same range of alternatives for this
                final rule. As with the proposal, the alternatives for stringency are
                bounded on the upper end by the baseline/no action alternative and on
                the lower end by 0 percent per year increases for both passenger cars
                and light trucks. Consistent with the proposal, for this final rule,
                the agencies considered stringency increases of between 0 and 2 percent
                per year for passenger cars and between 0 and 3 percent per year for
                light trucks, in addition to the baseline/no action alternative.
                 While it was not specifically modeled in the NPRM, the new
                preferred alternative of an increase in stringency of 1.5 percent for
                both cars and trucks was well within the range of alternatives
                considered. The proposal described the alternatives specifically
                modeled as options for the agencies, but also gave notice that they did
                not limit the agencies in selecting from among the range of
                alternatives under consideration.\290\
                ---------------------------------------------------------------------------
                 \290\ See, e.g., 83 FR at 43003 (Aug. 24, 2018) (``These
                alternatives were examined because they will be considered as
                options for the final rule. The agencies seek comment on these
                alternatives, seek any relevant data and information, and will
                review responses. That review could lead to the selection of one of
                the other regulatory alternatives for the final rule or some
                combination of the other regulatory alternatives (e.g., combining
                passenger cars standards from one alternative with light truck
                standards from a different alternative).''); id. at 43229
                (describing a factor relevant to ``the Administrator's selection of
                which standards are appropriate within the range of the Alternatives
                presented in this proposal'').
                ---------------------------------------------------------------------------
                 The agencies explained in the proposal that they were ``taking
                comment on a wide range of alternatives and have specifically modeled
                eight alternatives.'' \291\ As with the proposal, for the final rule,
                the agencies specifically modeled the upper and lower bounds of the
                baseline/no action alternative and 0 percent per year stringency
                increases for both passenger cars and light trucks. In both the
                proposal and the final rule, the agencies also modeled a stringency
                increase of 2 percent per year for passenger cars and 3 percent per
                year for light trucks, as well as a variety of other specific increases
                between 0 and 2 percent for passenger cars and 0 and 3 percent for
                light trucks.
                ---------------------------------------------------------------------------
                 \291\ 83 FR at 42990 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 The specific alternatives the agencies modeled for the final rule
                reflect their consideration of public comments. As discussed above,
                multiple commenters expressed support for equalizing the rates at which
                the stringencies of passenger car and light truck standards increase.
                To help the agencies evaluate alternatives that include the same
                stringency increase for passenger cars and light trucks, three of the
                seven alternatives (in addition to the baseline/no action alternative)
                that the agencies specifically modeled for the final rule included the
                same stringency increase for passenger cars and light trucks. This
                includes the new preferred alternative of an increase in stringency of
                1.5 percent for both cars and trucks. This alternative, and all others
                specifically modeled for the final rule, falls within the range of
                alternatives for stringency considered by the agencies in the proposal.
                 Beyond these stringency provisions discussed in the NPRM, the
                agencies also sought comment on a number of additional compliance
                flexibilities for the programs, as discussed in Section IX.
                2. Additional Alternatives Suggested by Commenters
                 Beyond the comments discussed above regarding the shapes of the
                functions defining fuel economy and CO2 targets, regarding
                the inclusion of non-CO2 emissions, and regarding the
                stringencies to be considered, the agencies also received a range of
                other comments regarding regulatory alternatives.
                 Some of these additional comments involved how CAFE and
                CO2 standards compare to one another for any given
                regulatory alternative. With a view toward maximizing harmonization of
                the standards, the Alliance, supported by some of its members'
                individual comments, indicated that ``to the degree flexibilities and
                incentives are not completely aligned between the CAFE and
                [CO2] programs, there must be an offset in the associated
                footprint-based targets to account for those differences. Some areas of
                particular concerns are air conditioning refrigerant credits, and
                incentives for advanced technology vehicles. The Alliance urges the
                Agencies to seek harmonization of the standards and flexibilities to
                the greatest extent possible. . . .'' \292\
                ---------------------------------------------------------------------------
                 \292\ Alliance, NHTSA-2018-0067-12073, at 40. See also FCA,
                NHTSA-2018-0067-11943, at 6-7.
                ---------------------------------------------------------------------------
                 On the other hand, discussing consideration of compliance credits
                but making a more general argument, the NYU Institute for Policy
                Integrity commented that ``. . . EPA is not allowed to set lower
                standards just for the sake of harmonization; to the contrary, full
                harmonization may be inconsistent with EPA's statutory
                responsibilities.'' \293\ Similarly, ACEEE argued that ``any
                consideration of an extension or expansion of credit provisions under
                the [carbon dioxide] or CAFE standards program should take as a
                starting point the assumption that the additional credits will allow
                the stringency of the standards to be increased.'' \294\
                ---------------------------------------------------------------------------
                 \293\ IPI, NHTSA-2018-0067-12213, at 21.
                 \294\ ACEEE, NHTSA-2018-0067-12122, at 3.
                ---------------------------------------------------------------------------
                 EPCA's requirement that NHTSA set standards at the maximum feasible
                levels is separate and ``wholly independent'' from the CAA's
                requirement, per Massachusetts v. EPA, that EPA issue regulations
                addressing pollutants that EPA has determined endanger public health
                and welfare.\295\ Nonetheless, as recognized by the Supreme Court,
                ``there is no reason to think the two agencies cannot both administer
                their obligations and yet avoid inconsistency.'' \296\ This conclusion
                was reached despite the fact that EPCA has a range of very specific
                requirements about how CAFE standards are to be structured, how
                manufacturers are to comply, what happens when manufacturers are unable
                to comply, and how NHTSA is to approach setting standards, and despite
                the fact that the CAA has virtually no such requirements. This means
                that while nothing about either EPCA or the CAA, much less the
                combination of the two, guarantees ``harmonization'' defining ``One
                National Program,'' the agencies are expected to be able to work out
                the differences.
                ---------------------------------------------------------------------------
                 \295\ Massachusetts v. EPA, 549 U.S. 497, 532 (2007).
                 \296\ Id.
                ---------------------------------------------------------------------------
                 Since tailpipe CO2 standards are de facto fuel economy
                standards, the more differences there are between CO2 and
                CAFE standards and compliance provisions, the more challenging it is
                for manufacturers to plan year-by-year production that responses to
                both, and the more difficult it is for affected stakeholders and the
                general public to understand regulation in this space. Therefore, even
                if the two statutes, taken together, do not guarantee ``full
                harmonization,'' steps toward greater
                [[Page 24263]]
                harmonization help with compliance planning and transparency--and meet
                the expectations set forth by the Supreme Court that the agencies avoid
                inconsistencies.
                 The agencies have taken important steps toward doing so. For
                example, EPA has adopted separate footprint-based CO2
                standards for passenger cars and light trucks, and has redefined CAFE
                calculation procedures to introduce recognition for the application of
                real-world fuel-saving technology that is not captured with traditional
                EPA two-cycle compliance testing. Detailed aspects of both sets of
                standards and corresponding compliance provisions are discussed at
                length in Section IX. The agencies never set out with the primary goal
                of achieving ``full harmonization,'' such that both sets of standards
                would lead each manufacturer to respond in exactly the same way in
                every model year.\297\ For example, EPA did not adopt the EPCA
                requirement that domestic passenger car fleets each meet a minimum
                standard, or the EPCA cap on compliance credit transfers between
                passenger car fleets. On the other hand, EPA also did not adopt the
                EPCA civil penalty provisions that have allowed some manufacturers to
                pay civil penalties as an alternative method of meeting EPCA
                obligations. These and other differences provide that even if CAFE and
                CO2 standards are ``mathematically'' harmonized, for any
                given manufacturer, the two sets of standards will not be identically
                burdensome in each model year. Inevitably, one standard will be more
                challenging than the other, varying over time, between manufacturers,
                and between fleets. This means manufacturers need to have compliance
                plans for both sets of standards.
                ---------------------------------------------------------------------------
                 \297\ Full harmonization would mean that, for example, if Ford
                would do some set of things over time in response to CAFE standards
                in isolation, it would do exactly the same things on exactly the
                same schedule in response to CO2 standards in isolation.
                ---------------------------------------------------------------------------
                 In 2012, recognizing that EPCA provides no clear basis to address
                HFC, CH4, or N2O emissions directly, the agencies
                ``offset'' CO2 targets from fuel economy targets (after
                converting the latter to a CO2 basis) by the amounts of
                credit EPA anticipated manufacturers would, on average, earn in each
                model years by reducing A/C leakage and adopting refrigerants with
                reduced GWPs. In 2012, EPA assumed that by 2021, all manufacturers
                would be earning the maximum available credit, and EPA's analysis
                assumed that all manufacturers would make progress at the same rate.
                However, as discussed above, data highlighted in comments by Chemours,
                Inc., demonstrate that actual manufacturers' adoption of lower-GWP
                refrigerants thus far ranges widely, with some manufacturers (e.g.,
                Nissan) having taken no such steps to move toward lower-GWP
                refrigerants, while others (e.g., JLR) have already applied lower-GWP
                refrigerants to all vehicles produced for sale in the U.S. Therefore,
                at least in practice, HFC provisions thus far continue to leave a gap
                (in terms of harmonization) between the two sets of standards. The
                proposal would have taken the additional step of decoupling provisions
                regarding HFC (i.e., A/C leakage credits), CH4, and
                N2O emissions from CO2 standards, addressing
                these in separate regulations to be issued in a new proposal. As
                discussed above, EPA did not finalize this proposal. Accordingly, for
                the regulatory alternatives considered today, EPA has reinstated
                offsets of CO2 targets from fuel economy targets, reflecting
                the assumption that all manufacturers will be earning the maximum
                available A/C leakage credit by MY 2021.
                 In addition to general comments on harmonization, the agencies
                received a range of comments on specific provisions--especially
                involving ``flexibilities''--that may or may not impact harmonization.
                With a view toward encouraging further electrification, NCAT proposed
                that EPA extend indefinitely the exclusion of upstream emissions from
                electricity generation, and also extend and potentially restructure
                production multipliers for PHEVs, EVs, and FCVs.\298\ On the other
                hand, connecting its comments back to the stringency of standards, NCAT
                also commented that ``. . . expansion of compliance flexibilities in
                the absence of any requirement to improve [CO2] reduction or
                fuel economy (as under the agencies' preferred option) could result in
                an effective deterioration of existing [CO2] and fuel
                economy performance, as well as little or no effective support for
                advanced vehicle technology development or deployment.'' \299\ Global
                Automakers indicated that the final rule ``should include a package of
                programmatic elements that provide automakers with flexible compliance
                options that promote the full breadth of vehicle technologies,'' such
                options to include the extension of ``advanced technology'' production
                multipliers through MY 2026, the indefinite exclusion of emissions from
                electricity generation, the extension to passenger cars of credits
                currently granted for the application of ``game changing'' technologies
                (e.g., HEVs) only to full-size pickup trucks, an increase (to 15 g/mi)
                of the cap on credits for off-cycle technologies, an updated credit
                ``menu'' of off-cycle technologies, and easier process for handling
                applications for off-cycle credits.\300\ The Alliance also called for
                expanded sales multipliers and a permanent exclusion of emissions from
                electricity generation.\301\ Walter Kreucher recommended the agencies
                consider finalizing the proposed standards but also keeping the augural
                standards as ``voluntary targets'' to ``provide compliance with the
                statutes and an aspirational goal for manufacturers.'' \302\
                ---------------------------------------------------------------------------
                 \298\ NCAT, NHTSA-2018-0067-11969, at 3-5.
                 \299\ Id.
                 \300\ Global Automakers, NHTSA-2018-0067-12032, at 4 et seq.
                 \301\ Alliance, NHTSA-2018-0067-12073, at 8.
                 \302\ Kreucher, W., NHTSA-2018-0067-0444, at 9.
                ---------------------------------------------------------------------------
                 The agencies have carefully considered these comments, and have
                determined that the current suite of ``flexibilities'' generally
                provide ample incentive more rapidly to develop and apply advanced
                technologies and technologies that produce fuel savings and/or
                CO2 reductions that would otherwise not count toward
                compliance. The agencies also share some stakeholders' concern that
                expanding these flexibilities could increase the risk of ``gaming''
                that would make compliance less transparent and would unduly compromise
                energy and environmental benefits. Nevertheless, as discussed in
                Section IX, EPA is adopting new multiplier incentives for natural gas
                vehicles. EPA is also finalizing some changes to procedures for
                evaluating applications for off-cycle credits, and expects these
                changes to make this process more accurate and more efficient. Also,
                EPA is revising its regulations to not require manufacturers to account
                for upstream emissions associated with electricity use for electric
                vehicles and plug-in hybrid electric vehicles through model year 2026;
                compliance will instead be based on tailpipe emissions performance only
                and not include emissions from electricity generation until model year
                2027. As discussed below, even with this change, and even accounting
                for continued increases in fuel prices and reductions in battery
                prices, BEVs are projected in this final rule analysis to continue to
                account for less than 5 percent of new light vehicle sales in the U.S.
                through model year 2026. To the extent that this projection turns out
                to reflect reality, this means that the impact of upstream emissions
                from electricity use on the projected CO2
                [[Page 24264]]
                reductions associated with these standards would likely remain small.
                Regarding comments suggesting that the augural standards should be
                finalized as ``voluntary targets,'' the agencies have determined that
                having such targets exist alongside actual regulatory requirements
                would be, at best, unnecessary and confusing.
                 Beyond these additional proposals, some commenters' proposals
                clearly fell outside authority provided under EPCA or the CAA. Ron
                Lindsay recommended the agencies ``consider postponing the rule changes
                until the U.S. can establish a legally binding national and
                international carbon budget and a binding mechanism to adhere to it.''
                \303\ EPCA requires NHTSA to issue standards for MY 2022 by April 1,
                2020, and previously-issued EPA regulations commit EPA to revisiting MY
                2021-2025 standards on a similar schedule. These statutory and
                regulatory provisions do not include a basis to delay decisions pending
                an international negotiation for which prospects and schedules are both
                unknown.
                ---------------------------------------------------------------------------
                 \303\ Ron Lindsay, EPA-HQ-OAR-2018-0283-1414, at 6.
                ---------------------------------------------------------------------------
                 SCAQMD, supported by Shyam Shukla, indicated that the agencies
                should consider an alternative that keeps the waiver for California's
                CO2 standards in place.\304\ NCAT and the North Carolina DEQ
                offered similar comments and CBD, et al. commented that ``among the set
                of more stringent alternatives that NEPA requires the agency to
                consider, NHTSA must include action alternatives that retain the
                standards California and other states have lawfully adopted.'' \305\ As
                discussed above, the agencies recently issued a final rule addressing
                the issue of California's authority. NEPA does not require NHTSA to
                include action alternatives that cannot be lawfully realized.
                ---------------------------------------------------------------------------
                 \304\ SCAQMD, NHTSA-2018-0067-5666, at 1-2; Shyam Shukla, NHTSA-
                2018-0067-5793, at 1-2.
                 \305\ NCAT, NHTSA-2018-0067-11969, at 64; NCDEQ, NHTSA-2018-
                0067-12025, at 38; CBD et al., NHTSA-2018-0067-12123, Attachment 1,
                at 18.
                ---------------------------------------------------------------------------
                 International Mosiac commented that NHTSA's DEIS ``is fatally
                flawed . . . because it does not consider any market-based alternatives
                (e.g., a `cap and trade' type option).'' \306\ While EPCA/EISA does
                include very specific provisions regarding trading of CAFE compliance
                credits, the statute provides no authority for a broad-based cap-and-
                trade program involving other sectors. Similarly, Michalek, et al.
                wrote that ``a more economically efficient approach of, taxing
                emissions and fuel consumption at socially appropriate levels would
                allow households to determine whether to reduce fuel consumption and
                emissions by driving less, by buying a vehicle with more fuel saving
                technologies, or by buying a smaller vehicle--or, alternatively, not to
                reduce fuel consumption and emissions at all but rather pay a cost
                based on the damages they cause. Forcing improvements only through one
                mechanism (fuel-saving technologies) increases the cost of achieving
                these outcomes.'' \307\ While some economists would agree with these
                comments, Congress has provided no clear authority for NHTSA or EPA to
                implement either an emissions tax or a broad-based cap-and-trade
                program in which motor vehicles could participate.
                ---------------------------------------------------------------------------
                 \306\ International Mosaic, NHTSA-2018-0067-11154, at 1-2.
                 \307\ Michalek, et al., NHTSA-2018-0067-11903, at 13.
                ---------------------------------------------------------------------------
                3. Details of Alternatives Considered in Final Rule
                a) Alternative 1
                 Alternative 1 holds the stringency of targets constant and MY 2020
                levels through MY 2026.
                [[Page 24265]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.066
                b) Alternative 2
                 Alternative 2 increases the stringency of targets annually during
                MYs 2021-2026 (on a gallon per mile basis, starting from MY 2020) by
                0.5 percent for passenger cars and 0.5 percent for light trucks.
                [[Page 24266]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.067
                c) Alternative 3
                 Alternative 3; the final standards promulgated today, increases the
                stringency of targets annually during MYs 2021-2026 (on a gallon per
                mile basis, starting from MY 2020) by 1.5 percent for passenger cars
                and 1.5 percent for light trucks.
                [[Page 24267]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.068
                d) Alternative 4
                 Alternative 4 increases the stringency of targets annually during
                MYs 2021-2026 (on a gallon per mile basis, starting from MY 2020) by
                1.0 percent for passenger cars and 2.0 percent for light trucks.
                [[Page 24268]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.069
                e) Alternative 5
                 Alternative 5 increases the stringency of targets annually during
                MYs 2022-2026 (on a gallon per mile basis, starting from MY 2021) by
                1.0 percent for passenger cars and 2.0 percent for light trucks.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.070
                [[Page 24269]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.071
                f) Alternative 6
                 Alternative 6 increases the stringency of targets annually during
                MYs 2021-2026 (on a gallon per mile basis, starting from MY 2020) by
                2.0 percent for passenger cars and 3.0 percent for light trucks.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.072
                [GRAPHIC] [TIFF OMITTED] TR30AP20.073
                [[Page 24270]]
                g) Alternative 7
                 Alternative 7 increases the stringency of targets annually during
                MYs 2022-2026 (on a gallon per mile basis, starting from MY 2021) by
                2.0 percent for passenger cars and 3.0 percent for light trucks.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.074
                [GRAPHIC] [TIFF OMITTED] TR30AP20.075
                 EPCA, as amended by EISA, requires that any manufacturer's
                domestically-manufactured passenger car fleet must meet the greater of
                either 27.5 mpg on average, or 92 percent of the average fuel economy
                projected by the Secretary for the combined domestic and non-domestic
                passenger automobile fleets manufactured for sale in the U.S. by all
                manufacturers in the model year, which projection shall be published in
                the Federal Register when the standard for that model year is
                promulgated in accordance with 49 U.S.C. 32902(b).\308\ Any time NHTSA
                establishes or changes a passenger car standard for a model year, the
                MDPCS for that model year must also be evaluated or re-evaluated and
                established accordingly. Thus, this final rule establishes the
                applicable MDPCS for MYs 2021-2026. Table V-22 lists the minimum
                domestic passenger car standards.
                ---------------------------------------------------------------------------
                 \308\ 49 U.S.C. 32902(b)(4).
                 [GRAPHIC] [TIFF OMITTED] TR30AP20.076
                
                [[Page 24271]]
                VI. Analytical Approach as Applied to Regulatory Alternatives
                A. Overview of Methods
                 Like analyses accompanying the NPRM and past CAFE and CAFE/
                CO2 rulemakings, the analysis supporting today's notice
                spans a range of technical topics, uses a range of different types of
                data and estimates, and applies several different types of computer
                models. The purpose of the analysis is not to determine the standards,
                but rather to provide information for consideration in doing so. The
                analysis aims to answer the question ``what impacts might each of these
                regulatory alternatives have?''
                 Over time, NHTSA's and, more recently, NHTSA's and EPA's analyses
                have expanded to address an increasingly wide range of types of
                impacts. Today's analysis involves, among other things, estimating how
                the application of various combinations of technologies could impact
                vehicles' costs and fuel economy levels (and CO2 emission
                rates), estimating how vehicle manufacturers might respond to standards
                by adding fuel-saving technologies to new vehicles, estimating how
                changes in new vehicles might impact vehicle sales and operation, and
                estimating how the combination of these changes might impact national-
                scale energy consumption, emissions, highway safety, and public health.
                In addition, the EIS accompanying today's notice addresses impacts on
                air quality and climate. The analysis of these factors informs and
                supports both NHTSA's application of the statutory requirements
                governing the setting of ``maximum feasible'' fuel-economy standards
                under EPCA, including, among others, technological feasibility and
                economic practicability, and EPA's application of the CAA requirements
                for tailpipe emissions.
                 Supporting today's analysis, the agencies have brought to bear a
                variety of different types of data, a few examples of which include
                fuel economy compliance reports, historical sales and average
                characteristics of light-duty vehicles, historical economic and
                demographic measures, historical travel demand and energy prices and
                consumption, and historical measures of highway safety. Also supporting
                today's analysis, the agencies have applied several different types of
                estimates, a few examples of which include projections of the future
                cost of different fuel-saving technologies, projections of future GDP
                and the number of households, estimates of the ``gap'' between
                ``laboratory'' and on-road fuel economy, and estimates of the social
                cost of CO2 emissions and petroleum ``price shocks.''
                 With a view toward transparency, repeatability, and efficiency, the
                agencies have used a variety of computer models to conduct the majority
                of today's analysis. For example, the agencies have applied DOE/EIA's
                National Energy Modeling System (NEMS) to estimate future energy
                prices, EPA's MOVES model to estimate tailpipe emission rates for ozone
                precursors and other criteria pollutants, DOE/Argonne's GREET model to
                estimate emission rates for ``upstream'' processes (e.g., petroleum
                refining), and DOE/Argonne's Autonomie simulation tool to estimate the
                fuel consumption impacts of different potential combinations of fuel-
                saving technology. In addition, the EIS accompanying today's notice
                applies photochemical models to estimate air quality impacts, and
                applies climate models to estimate climate impacts of overall emissions
                changes.
                 Use of these different types of data, estimates, and models is
                discussed further below in the most closely relevant sections. For
                example, the agencies' use of NEMS is discussed below in the portion of
                Section VI that addresses the macroeconomic context, which includes
                fuel prices, and the agencies use of Autonomie is discussed in the
                portion of Section VI.B.3 that addresses the agencies' approach to
                estimating the effectiveness of various technologies (in reducing fuel
                consumption and CO2 emissions).
                 Providing an integrated means to estimate both vehicle
                manufacturers' potential responses to CAFE or CO2 standards
                and, in turn, many of the different potential direct results (e.g.,
                changes in new vehicle costs) and indirect impacts (e.g., changes in
                rates of fleet turnover) of those responses, the CAFE Model plays a
                central role in the agencies' analysis supporting today's notice. The
                agencies used the specific models mentioned above to develop inputs to
                the CAFE model, such as fuel prices and emission factors. Outputs from
                the CAFE Model are discussed in Sections VII and VIII of today's
                notice, and in the accompanying RIA. The EIS accompanying today's
                notice makes use of the CAFE Model's estimates of changes in total
                emissions from light-duty vehicles, as well as corresponding changes in
                upstream emissions. These changes in emissions are included in the set
                of inputs to the models used to estimate air quality and climate
                impacts.
                 The remainder of this overview focuses on the CAFE Model. The
                purpose of this overview is not to provide a comprehensive technical
                description of the model,\309\ but rather to give an overview of the
                model's functions, to explain some specific aspects not addressed
                elsewhere in today's notice, and to discuss some model aspects that
                were the subject of significant public comment. Some model functions
                and related comments are addressed in other parts of today's notice.
                For example, the model's handling of Autonomie-based fuel consumption
                estimates is addressed in the portion of Section VI.B.3 that discusses
                the agencies' application of Autonomie. The model documentation
                accompanying today's notice provides a comprehensive and detailed
                description of the model's functions, design, inputs, and outputs.
                ---------------------------------------------------------------------------
                 \309\ The CAFE Model is available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system with documentation and all inputs and outputs supporting
                today's notice.
                ---------------------------------------------------------------------------
                1. Overview of CAFE Model
                 The basic design of the CAFE Model is as follows: The system first
                estimates how vehicle manufacturers might respond to a given regulatory
                scenario, and from that potential compliance solution, the system
                estimates what impact that response will have on fuel consumption,
                emissions, and economic externalities. A regulatory scenario involves
                specification of the form, or shape, of the standards (e.g., flat
                standards, or linear or logistic attribute-based standards), scope of
                passenger car and truck regulatory classes, and stringency of the CAFE
                and CO2 standards for each model year to be analyzed.
                 Manufacturer compliance simulation and the ensuing effects
                estimation, collectively referred to as compliance modeling, encompass
                numerous subsidiary elements. Compliance simulation begins with a
                detailed user-provided initial forecast of the vehicle models offered
                for sale during the simulation period. The compliance simulation then
                attempts to bring each manufacturer into compliance with the standards
                defined by the regulatory scenario contained within an input file
                developed by the user. For example, a regulatory scenario may define
                CAFE or CO2 standards that increase in stringency by 4
                percent per year for 5 consecutive years.
                 The model applies various technologies to different vehicle models
                in each manufacturer's product line to simulate how each manufacturer
                might make progress toward compliance with the specified standard.
                Subject to a variety of user-controlled constraints, the model applies
                technologies based on
                [[Page 24272]]
                their relative cost-effectiveness, as determined by several input
                assumptions regarding the cost and effectiveness of each technology,
                the cost of compliance (determined by the change in CAFE or
                CO2 credits, CAFE-related civil penalties, or value of
                CO2 credits, depending on the compliance program being
                evaluated and the effective-cost mode in use), and the value of avoided
                fuel expenses. For a given manufacturer, the compliance simulation
                algorithm applies technologies either until the manufacturer runs out
                of cost-effective technologies, until the manufacturer exhausts all
                available technologies, or, if the manufacturer is assumed to be
                willing to pay civil penalties, until paying civil penalties becomes
                more cost-effective than increasing vehicle fuel economy. At this
                stage, the system assigns an incurred technology cost and updated fuel
                economy to each vehicle model, as well as any civil penalties incurred
                by each manufacturer. This compliance simulation process is repeated
                for each model year available during the study period.
                 This point marks the system's transition between compliance
                simulation and effects calculations. At the conclusion of the
                compliance simulation for a given regulatory scenario, the system
                contains multiple copies of the updated fleet of vehicles corresponding
                to each model year analyzed. For each model year, the vehicles'
                attributes, such as fuel types (e.g., diesel, electricity), fuel
                economy values, and curb weights have all been updated to reflect the
                application of technologies in response to standards throughout the
                study period. For each vehicle model in each of the model year specific
                fleets, the system then estimates the following: Lifetime travel, fuel
                consumption, carbon dioxide and criteria pollutant emissions, the
                magnitude of various economic externalities related to vehicular travel
                (e.g., noise), and energy consumption (e.g., the economic costs of
                short-term increases in petroleum prices). The system then aggregates
                model-specific results to produce an overall representation of modeling
                effects for the entire industry.
                 Different categorization schemes are relevant to different types of
                effects. For example, while a fully disaggregated fleet is retained for
                purposes of compliance simulation, vehicles are grouped by type of fuel
                and regulatory class for the energy, carbon dioxide, criteria
                pollutant, and safety calculations. Therefore, the system uses model-
                by-model categorization and accounting when calculating most effects,
                and aggregates results only as required for efficient reporting.
                2. Representation of the Market
                 As a starting point, the model needs enough information to
                represent each manufacturer covered by the program. As discussed below
                in Section VI.B.1, the MY 2017 analysis fleet contains information
                about each manufacturer's:
                 Vehicle models offered for sale--their current (i.e., MY
                2017) production volumes, manufacturer suggested retail prices (MSRPs),
                fuel saving technology content and other attributes (curb weight, drive
                type, assignment to technology class and regulatory class);
                 Production considerations--product cadence of vehicle
                models (i.e., schedule of model redesigns and ``freshenings''), vehicle
                platform membership, degree of engine and/or transmission sharing (for
                each model variant) with other vehicles in the fleet; and
                 Compliance constraints and flexibilities--preference for
                full compliance or penalty payment/credit application, willingness to
                apply additional cost-effective fuel saving technology in excess of
                regulatory requirements, projected applicable flexible fuel credits,
                and current credit balance (by model year and regulatory class) in
                first model year of simulation.
                Representation of Fuel-Saving Technologies
                 The modeling system defines technology pathways for grouping and
                establishing a logical progression of technologies that can be applied
                to a vehicle. Technologies that share similar characteristics form
                cohorts that can be represented and interpreted within the CAFE Model
                as discrete entities. The following Table VI-1 shows the technologies
                available within the modeling system used for this final rule. Each
                technology is discussed in detail below. However, an understanding of
                the technologies considered and how they are defined in the model
                (e.g., a 6-speed manual transmission is defined as ``MT6'') is helpful
                for the following explanation of the compliance simulation and the
                inputs required for that simulation.
                BILLING CODE 4910-59-P
                [[Page 24273]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.077
                [[Page 24274]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.078
                BILLING CODE 4910-59-C
                 These entities are then laid out into pathways (or paths), which
                the system uses to define relations of mutual exclusivity between
                conflicting sets of technologies. For example, as presented in the next
                section, technologies on the Turbo Engine path are incompatible with
                those on the HCR Engine or the Diesel Engine paths. As such, whenever a
                vehicle uses a technology from one pathway (e.g., turbo), the modeling
                system immediately disables the incompatible technologies from one or
                more of the other pathways (e.g., HCR and diesel).
                 In addition, each path designates the direction in which vehicles
                are allowed to advance as the modeling system evaluates specific
                technologies for application. Enforcing this directionality within the
                model ensures that a vehicle that uses a more advanced or more
                efficient technology (e.g., AT8) is not allowed to ``downgrade'' to a
                less efficient option (e.g., AT5). Visually, as portrayed in the charts
                in the sections that follow, this is represented by an arrow leading
                from a preceding technology to a succeeding one, where vehicles begin
                at the root of each path, and traverse to each successor technology in
                the direction of the arrows.
                 The modeling system incorporates twenty technology pathways for
                evaluation as shown below. Similar to individual technologies, each
                path carries an intrinsic application level that denotes the scope of
                applicability of all technologies present within that path, and whether
                the pathway is evaluated on one vehicle at a time, or on a collection
                of vehicles that share a common platform, engine, or transmission.
                [[Page 24275]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.079
                 Even though technology pathways outline a logical progression
                between related technologies, all technologies available to the system
                are evaluated concurrently and independently of each other. Once all
                technologies have been examined, the model selects a solution deemed to
                be most cost-effective for application on a vehicle. If the modeling
                system applies a technology that resides later in the pathway, it will
                subsequently disable all preceding technologies from further
                consideration to prevent a vehicle from potentially downgrading to a
                less advanced option. Consequently, the system skips any technology
                that is already present on a vehicle (either those that were available
                on a vehicle from the input fleet or those that were previously applied
                by the model). This ``parallel technology'' approach, unlike the
                ``parallel path'' methodology utilized in the preceding versions of the
                model, allows the system always to consider the entire set of available
                technologies instead of foregoing the application of potentially more
                cost-effective options that happen to reside further down the
                pathway.\310\ This revised approach addresses comments summarized
                below, and allows the system to analyze all available technology
                options concurrently and independently of one other without having to
                first apply one or more ``predecessor'' technologies. For example, if
                model inputs are such that a 7-speed transmission is cost-effective,
                but not as cost-effective as an 8-speed transmission, the revised
                approach enables the model to skip over the 7-speed transmission
                entirely, whereas the NPRM version of the model might first apply the
                7-speed transmission and then consider whether to proceed immediately
                to the 8-speed transmission. As such, the model's choices for
                evaluation of new technology solutions becomes slightly less
                restrictive, allowing it immediately to consider and apply more
                advanced options, and increasing the likelihood that the a globally
                optimum solution is selected.
                ---------------------------------------------------------------------------
                 \310\ Previous versions of the CAFE Model followed a ``low-
                cost'' first approach where the system would stop evaluating
                technologies residing within a given pathway as soon as the first
                cost-effective option within that path was reached.
                ---------------------------------------------------------------------------
                 Some commenters supported the agencies' use of such pathways in the
                simulation of manufacturers' potential application of technologies. As
                one of a dozen examples of CAFE model design elements that lead to the
                transparent representation of real-world factors, the Alliance
                highlighted ``recognition of the need for manufacturers to follow
                `technology' pathways that retain capital and implementation expertise,
                such as specializing in one type of engine or transmission instead of
                following an unconstrained optimization that would cause manufacturers
                to leap to unrelated technologies and show overly optimistic costs and
                benefits.'' \311\ Similarly, Toyota commented that ``the inertia of
                capital investments and engineering expertise dedicated to one
                compliance technology or set of technologies makes it unreasonable for
                manufacturers to immediately switch to another technology path.'' \312\
                ---------------------------------------------------------------------------
                 \311\ Alliance, NHTSA-2018-0067-12073, at 9.
                 \312\ Toyota, NHTSA-2018-0067-12098, at 7.
                ---------------------------------------------------------------------------
                 Other commenters cited the use of technology pathways as inherently
                overly restrictive. For example, as an example of ``arbitrary model
                constraints,'' a coalition of commenters cited the fact the model
                ``prohibit[s] manufacturers from switching vehicle technology
                pathways.'' \313\ Also, EDF, UCS, and CARB cited the combination of
                technology pathways, decision making criteria, and model inputs as
                producing unrealistic results.\314\ Regarding the technology pathways,
                specifically, EDF's consultant argued that the technology paths are not
                [[Page 24276]]
                transparent, and cited the potential that specific paths may not
                necessarily be arranged in progression from least to most cost-
                effective--that ``NHTSA ignores the cost of the technology when
                developing this list.'' \315\ Relatedly, as EDF's consultant commented:
                ---------------------------------------------------------------------------
                 \313\ CBD, et al., NHTSA-2018-0067-12057, at 3.
                 \314\ EDF, NHTSA-2018-0067-12108, Appendix A, at 57 et seq.;
                UCS, NHTSA-2018-0067-12039, Appendix, at 25 et seq.; Roush
                Industries, NHTSA-2018-0067-11984, at 5.
                 \315\ EDF, NHTSA-2018-0067-12108, Appendix B, at 69.
                 [T]he Volpe Model is not designed to look backwards along its
                technology paths. Thus, the opportunity to recover the expenditure
                of inefficient technology is missed. NHTSA might argue that a
                manufacturer will not invest in 10-speed transmissions, for example,
                and then return to an older design. Whether or not this is true in
                real life, such a view would put too much stake in the Volpe Model
                projections. The model simply projects what could be done, not what
                will be. Anyone examining the progression of technology and noting
                the reversion of transmission technology could easily modify the
                model inputs to avoid this. Also, if NHTSA evaluated combinations of
                technologies prior to entering them in the model piecemeal, it would
                automatically avoid such apparent problems.\316\
                ---------------------------------------------------------------------------
                 \316\ Ibid., at 70.
                 The agencies also received additional public comments on specific
                paths and specific interactions between paths (e.g., involving engines
                and hybridization). These comments are addressed below.
                 The agencies have carefully considered these comments and the
                approach summarized below reflects some corresponding revision. As
                mentioned above, the CAFE model now approaches the technology paths in
                a such way that, faced with two cost-effective technologies on the same
                path, the model can proceed directly to the more advanced technology if
                that technology is the more cost effective of the two.
                 However, the agencies reject assertions that the model's use of
                technology paths is not transparent. The agencies provided extensive
                explanatory text, figures, model documentation, and model source code
                specifically addressing these paths (and other model features). This
                transparency appears evident in that commenters (sometimes while
                claiming that a specific feature of the model is not transparent)
                presented analytical results involving changes to corresponding inputs
                that required a detailed understanding of that feature's operation.
                 Regarding comments that the technology paths should be arranged in
                order of cost-effectiveness, the agencies note that such comments
                presume, without merit, that costs, fuel consumption impacts, and other
                inputs (e.g., fuel prices) that logically impact manufacturers'
                decision-making are not subject to uncertainty. These inputs are all
                subject to uncertainty, and the CAFE Model's arrangement of
                technologies into several paths is responsive to these uncertainties.
                Nevertheless, the agencies maintain that some technologies do reflect a
                higher level of advancement than others (e.g., 10-speed transmissions
                vs. 5-speed transmissions), and while manufacturers may, in practice,
                occasionally revert to less advanced technologies, it is appropriate
                and reasonable to conduct the agencies' analysis in a manner that
                assumes manufacturers will continue to make forward progress. As
                observed by EDF's consultant's remarks, the CAFE Model ``simply
                projects what could be done, not what will be.'' While no model, much
                less any model relying on information that can be made publicly
                available, can hope to represent precisely each manufacturers' actual
                detailed constrains related to product development and planning, such
                constraints are real and important. The agencies agree that the CAFE
                Model's representation of such constraints--including the Model's use
                of technology paths--provides a reasonable means of accounting for
                them.
                4. Compliance Simulation
                 The CAFE model provides a way of estimating how vehicle
                manufacturers could attempt to comply with a given CAFE standard by
                adding technology to fleets that the agencies anticipate they will
                produce in future model years. This exercise constitutes a simulation
                of manufacturers' decisions regarding compliance with CAFE or
                CO2 standards.
                 This compliance simulation begins with the following inputs: (a)
                The analysis fleet of vehicles from model year 2017 discussed below in
                Section VI.B.1, (b) fuel economy improving technology estimates
                discussed below in Section VI.C, (c) economic inputs discussed below in
                Section VI.D, and (d) inputs defining baseline and potential new CAFE
                or CO2 standards discussed above in Section V. For each
                manufacturer, the model applies technologies in both a logical sequence
                and a cost-optimizing strategy in order to identify a set of
                technologies the manufacturer could apply in response to new CAFE or
                CO2 standards. The model applies technologies to each of the
                projected individual vehicles in a manufacturer's fleet, considering
                the combined effect of regulatory and market incentives while
                attempting to account for manufacturers' production constraints.
                Depending on how the model is exercised, it will apply technology until
                one of the following occurs:
                 (1) The manufacturer's fleet achieves compliance \317\ with the
                applicable standard and adding additional technology in the current
                model year would be attractive neither in terms of stand-alone (i.e.,
                absent regulatory need) cost-effectiveness nor in terms of facilitating
                compliance in future model years;
                ---------------------------------------------------------------------------
                 \317\ When determining whether compliance has been achieved in
                the CAFE program, existing CAFE credits that may be carried over
                from prior model years or transferred between fleets are also used
                to determine compliance status. For purposes of determining the
                effect of maximum feasible CAFE standards, however, EPCA prohibits
                NHTSA from considering these mechanisms for years being considered
                (though it does so for model years that are already final) and the
                agency runs the CAFE model without enabling these options. 49 U.S.C.
                32902(h)(3).
                ---------------------------------------------------------------------------
                 (2) The manufacturer ``exhausts'' available technologies; \318\ or
                ---------------------------------------------------------------------------
                 \318\ In a given model year, it is possible that production
                constraints cause a manufacturer to ``run out'' of available
                technology before achieving compliance with standards. This can
                occur when: (a) An insufficient volume of vehicles are expected to
                be redesigned, (b) vehicles have moved to the ends of each
                (relevant) technology pathway, after which no additional options
                exist, or (c) engineering aspects of available vehicles make
                available technology inapplicable (e.g., secondary axle disconnect
                cannot be applied to two-wheel drive vehicles).
                ---------------------------------------------------------------------------
                 (3) For manufacturers assumed to be willing to pay civil penalties
                (in the CAFE program), the manufacturer reaches the point at which
                doing so would be more cost-effective (from the manufacturer's
                perspective) than adding further technology.
                 The model accounts explicitly for each model year, applying
                technologies when vehicles are scheduled to be redesigned or freshened
                and carrying forward technologies between model years once they are
                applied (until, if applicable, they are superseded by other
                technologies). The model then uses these simulated manufacturer fleets
                to generate both a representation of the U.S. auto industry and to
                modify a representation of the entire light-duty registered vehicle
                population. From these fleets, the model estimates changes in physical
                quantities (gallons of fuel, pollutant emissions, traffic fatalities,
                etc.) and calculates the relative costs and benefits of regulatory
                alternatives under consideration.
                 The CAFE model accounts explicitly for each model year, in turn,
                because manufacturers actually ``carry forward'' most technologies
                between model years, tending to concentrate the application of new
                technology to vehicle redesigns or mid-cycle ``freshenings,'' and
                design cycles vary widely among manufacturers and specific products.
                [[Page 24277]]
                Comments by manufacturers and model peer reviewers strongly support
                explicit year-by-year simulation. Year-by-year accounting also enables
                accounting for credit banking (i.e., carry-forward), as discussed
                above, and at least four environmental organizations recently submitted
                comments urging the agencies to consider such credits, citing NHTSA's
                2016 results showing impacts of carried-forward credits.\319\ Moreover,
                EPCA/EISA requires that NHTSA make a year-by-year determination of the
                appropriate level of stringency and then set the standard at that
                level, while ensuring ratable increases in average fuel economy through
                MY 2020. The multi-year planning capability, simulation of ``market-
                driven overcompliance,'' and EPCA credit mechanisms (again, for
                purposes of modeling the CAFE program) increase the model's ability to
                simulate manufacturers' real-world behavior, accounting for the fact
                that manufacturers will seek out compliance paths for several model
                years at a time, while accommodating the year-by-year requirement. This
                same multi-year planning structure is used to simulate responses to
                standards defined in grams CO2/mile, and utilizing the set
                of specific credit provisions defined under EPA's program.
                ---------------------------------------------------------------------------
                 \319\ Comment by Environmental Law & Policy Center, Natural
                Resources Defense Council (NRDC), Public Citizen, and Sierra Club,
                Docket ID EPA-HQ-OAR-2015-0827-9826, at 28-29.
                ---------------------------------------------------------------------------
                 After the light-duty rulemaking analysis accompanying the 2012
                final rule that finalized NHTSA's standards through MY 2021, NHTSA
                began work on changes to the CAFE model with the intention of better
                reflecting constraints of product planning and cadence for which
                previous analyses did not account. This involves accounting for
                expected future schedules for redesigning and ``freshening'' vehicle
                models, and accounting for the fact that a given engine or transmission
                is often shared among more than one vehicle model, and a given vehicle
                production platform often includes more than one vehicle model. These
                real product planning considerations are explained below.
                 Like earlier versions, the current CAFE model provides the
                capability for integrated analysis spanning different regulatory
                classes, accounting both for standards that apply separately to
                different classes and for interactions between regulatory classes.
                Light vehicle CAFE and CO2 standards are specified
                separately for passenger cars and light trucks. However, there is
                considerable sharing between these two regulatory classes, where a
                single engine, transmission, or platform can appear in both the
                passenger car and light truck regulatory class. For example, some
                sport-utility vehicles are offered in 2WD versions (classified as
                passenger cars for compliance purposes) and 4WD versions (classified as
                light trucks for compliance purposes). Integrated analysis of
                manufacturers' passenger car and light truck fleets provides the
                ability to account for such sharing and reduces the likelihood of
                finding solutions that could involve introducing impractical and
                unrealistic levels of complexity in manufacturers' product lines. In
                addition, integrated fleet analysis provides the ability to simulate
                the potential that manufacturers could earn CAFE and CO2
                credits by over complying with the standard in one fleet and use those
                credits toward compliance with the standard in another fleet (i.e., to
                simulate credit transfers between regulatory classes).\320\
                ---------------------------------------------------------------------------
                 \320\ Note, however, that EPCA prohibits NHTSA from considering
                the availability of such credit trading when setting maximum
                feasible fuel economy standards. 49 U.S.C. 32902(h)(3).
                ---------------------------------------------------------------------------
                 The CAFE model also accounts for EPCA's requirement that compliance
                be determined separately for fleets of domestic passenger cars and
                fleets of imported passenger cars. The model accounts for all three
                CAFE regulatory classes simultaneously (i.e., in an integrated way) yet
                separately: Domestic passenger cars, imported passenger cars, and light
                trucks. The model further accounts for two related specific statutory
                requirements specifically involving this distinction between domestic
                and imported passenger cars. First, EPCA/EISA requires that any given
                fleet of domestic passenger cars meet a minimum standard, irrespective
                of any available compliance credits. Second, EPCA/EISA requires
                compliance with the standards applicable to the domestic passenger car
                fleet without regard to traded or transferred credits.\321\
                ---------------------------------------------------------------------------
                 \321\ 49 U.S.C. 32903(f)(2) and (g)(4).
                ---------------------------------------------------------------------------
                 However, the CAA has no such limitation regarding compliance by
                domestic and imported vehicles; EPA did not adopt provisions similar to
                the aforementioned EPCA/EISA requirements and is not doing so today.
                Therefore, the CAFE model determines compliance for manufacturers'
                overall passenger car and light truck fleets for EPA's program.
                 Each manufacturer's regulatory requirement represents the
                production-weighted harmonic mean of their vehicle's targets in each
                regulated fleet. This means that no individual vehicle has a
                ``standard,'' merely a target, and each manufacturer is free to
                identify a compliance strategy that makes the most sense given its
                unique combination of vehicle models, consumers, and competitive
                position in the various market segments. As the CAFE model provides
                flexibility when defining a set of regulatory standards, each
                manufacturer's requirement is dynamically defined based on the
                specification of the standards for any simulation and the distribution
                of footprints within each fleet.
                 Given this information, the model attempts to apply technology to
                each manufacturer's fleet in a manner that, given product planning and
                engineering-related considerations, optimizes the selected cost-related
                metric. The metric supported by the NPRM version of the model is termed
                ``effective cost.'' The effective cost captures more than the
                incremental cost of a given technology; it represents the difference
                between their incremental cost and the value of fuel savings to a
                potential buyer over the first 30 months of ownership.\322\ In addition
                to the technology cost and fuel savings, the effective cost also
                includes the change in CAFE civil penalties from applying a given
                technology and any estimated welfare losses associated with the
                technology (e.g., earlier versions of the CAFE model simulated low-
                range electric vehicles that produced a welfare loss to buyers who
                valued standard operating ranges between re-fueling events). Comments
                on this metric are discussed below, as are model changes responding to
                these comments.
                ---------------------------------------------------------------------------
                 \322\ The length of time over which to value fuel savings in the
                effective cost calculation is a model input that can be modified by
                the user. This analysis uses 30 months' worth of fuel savings in the
                effective cost calculation, using the price of fuel at the time of
                vehicle purchase.
                ---------------------------------------------------------------------------
                 This construction allows the model to choose technologies that both
                improve a manufacturer's regulatory compliance position and are most
                likely to be attractive to its consumers. This also means that
                different assumptions about future fuel prices will produce different
                rankings of technologies when the model evaluates available
                technologies for application. For example, in a high fuel price regime,
                an expensive but very efficient technology may look attractive to
                manufacturers because the value of the fuel savings is sufficiently
                high both to counteract the higher cost of the technology and,
                implicitly, to satisfy consumer demand to balance price increases with
                reductions in operating cost.
                [[Page 24278]]
                 In general, the model adds technology for several reasons but
                checks these sequentially. The model then applies any ``forced''
                technologies. Currently, only variable valve timing (VVT) is forced to
                be applied to vehicles at redesign since it is the root of the engine
                path and the reference point for all future engine technology
                applications.\323\ The model next applies any inherited technologies
                that were applied to a leader vehicle on the same vehicle platform and
                carried forward into future model years where follower vehicles (on the
                shared system) are freshened or redesigned (and thus eligible to
                receive the updated version of the shared component). In practice, very
                few vehicle models enter without VVT, so inheritance is typically the
                first step in the compliance loop. Next, the model evaluates the
                manufacturer's compliance status, applying all cost-effective
                technologies regardless of compliance status.\324\ Then the model
                applies expiring overcompliance credits (if allowed to do so under the
                perspective of either the ``unconstrained'' or ``standard setting''
                analysis, for CAFE purposes).\325\ At this point, the model checks the
                manufacturer's compliance status again. If the manufacturer is still
                not compliant (and is unwilling to pay civil penalties, again for CAFE
                modeling), the model will add technologies that are not cost-effective
                until the manufacturer reaches compliance. If the manufacturer exhausts
                opportunities to comply with the standard by improving fuel economy/
                reducing emissions (typically due to a limited percentage of its fleet
                being redesigned in that year), the model will apply banked CAFE or
                CO2 credits to offset the remaining deficit. If no credits
                exist to offset the remaining deficit, the model will reach back in
                time to alter technology solutions in earlier model years.
                ---------------------------------------------------------------------------
                 \323\ As a practical matter, this affects very few vehicles.
                More than 95 percent of vehicles in the market file either already
                have VVT present or have surpassed the basic engine path through the
                application of hybrids or electric vehicles.
                 \324\ For further explanation of how the CAFE model considers
                the effective cost of applying different technologies see the CAFE
                Model Documentation for the final rule, at S5.3 Compliance
                Simulation Algorithm.
                 \325\ As mentioned above, EPCA prohibits consideration of
                available credits when setting maximum feasible fuel economy
                standards. 49 U.S.C. 32902(h)(3).
                ---------------------------------------------------------------------------
                 The CAFE model implements multi-year planning by looking back,
                rather than forward. When a manufacturer is unable to comply through
                cost-effective (i.e., producing effective cost values less than zero)
                technology improvements or credit application in a given year, the
                model will ``reach back'' to earlier years and apply the most cost-
                effective technologies that were not applied at that time and then
                carry those technologies forward into the future and re-evaluate the
                manufacturer's compliance position. The model repeats this process
                until compliance in the current year is achieved, dynamically
                rebuilding previous model year fleets and carrying them forward into
                the future, and accumulating CAFE or CO2 credits from over-
                compliance with the standard wherever appropriate.
                 In a given model year, the model determines applicability of each
                technology to each vehicle platform, model, engine, and transmission.
                The compliance simulation algorithm begins the process of applying
                technologies based on the CAFE or CO2 standards specified
                during the current model year. This involves repeatedly evaluating the
                degree of noncompliance, identifying the next ``best'' technology
                (ranked by the effective cost discussed earlier) available on each of
                the parallel technology paths described above and applying the best of
                these. The algorithm combines some of the pathways, evaluating them
                sequentially instead of in parallel, to ensure appropriate incremental
                progression of technologies.
                 The algorithm first finds the best next applicable technology in
                each of the technology pathways and then selects the best among these.
                For CAFE purposes, the model applies the technology to the affected
                vehicles if a manufacturer is either unwilling to pay penalties or if
                applying the technology is more cost-effective than paying penalties.
                Afterwards, the algorithm reevaluates the manufacturer's degree of
                noncompliance and continues application of technology. Once a
                manufacturer reaches compliance (i.e., the manufacturer would no longer
                need to pay penalties), the algorithm proceeds to apply any additional
                technology determined to be cost-effective (as discussed above).
                Conversely, if a manufacturer is assumed to prefer to pay penalties,
                the algorithm only applies technology up to the point where doing so is
                less costly than paying penalties. The algorithm stops applying
                additional technology to this manufacturer's products once no more
                cost-effective solutions are encountered. This process is repeated for
                each manufacturer present in the input fleet. It is then repeated for
                each model year. Once all model years have been processed, the
                compliance simulation algorithm concludes. The process for
                CO2 standard compliance simulation is similar, but without
                the option of penalty payment, such that technologies are applied until
                compliance (accounting for any modeled application of credits) is
                achieved. For both CAFE and CO2 standards, the model also
                applies any additional (i.e., beyond required for compliance)
                technology that ``pays back'' within a specified period (for the NPRM
                and today's analysis, 30 months).
                 Some commenters argued that the CAFE model applies constraints that
                excessively limit options manufacturers have to add technology, causing
                the model to overestimate costs to achieve a given level of
                improvement.\326\ Some of these commenters further argued that the
                agencies should assume greater potential to apply technologies that
                contribute to compliance by improving air conditioner efficiency or
                otherwise reducing ``off cycle'' fuel consumption and CO2
                emissions.\327\ Other commenters argued that such constraints, while
                warranting some refinements, help the model to simulate manufacturers'
                decision making realistically and to estimate technology effectiveness
                and costs reasonably.328 329
                ---------------------------------------------------------------------------
                 \326\ NHTSA-2018-0067-12057, CBD, et. al, p. 3.
                 \327\ NHTSA-2018-0067-11741, ICCT, Attachment 2, p. 4.
                 \328\ NHTSA-2018-0067-12073, Alliance of Automobile
                Manufacturers, pp. 134-36.
                 \329\ American Honda Motor Co., ``Honda Comments on the NPRM and
                various proposals contained therein--Prepared for NHTSA, EPA and
                ARB,'' October 17, 2018, pp. 12-16.
                ---------------------------------------------------------------------------
                 Some commenters questioned the ``effective cost'' metric the model
                uses to decide among available options, claiming that the metric also
                causes the model to avoid selection of pathways that are not always
                economically optimal.\330\ One of these commenters recommended the
                agencies modify the effective cost metric for CO2 compliance
                by removing the term placing a monetary value on progress toward
                compliance, and instead dividing the remaining net cost (i.e., the
                increase in technology costs minus a portion of the fuel outlays
                expected to be avoided) by the additional CO2 credits
                earned.\331\ Another of these commenters claimed on one hand, that the
                effective cost metric ``does not include a measurement of the
                technology's reduction in fuel consumption or CO2
                emissions'' and, on the other, that the metric inappropriately places a
                value on avoided fuel consumption.\332\
                ---------------------------------------------------------------------------
                 \330\ NHTSA-2018-0067-11741, ICCT, Attachment 3, p. I-62.
                 \331\ NHTSA-2018-0067-12039, UCS, Technical Appendix, pp. 28-32.
                 \332\ NHTSA-2018-0067-12108, EDF, Appendix B, p. 67.
                ---------------------------------------------------------------------------
                 One commenter claimed that the model inappropriately allows earned
                [[Page 24279]]
                credits (including CO2 program credits for which EPA has
                granted a one-time exemption from carry-forward limits) to expire while
                also showing undue degrees overcompliance with standards, and further
                proposed that the model be modified to simulate both credit ``carry
                back'' (aka ``borrowing'') and credit trading between
                manufacturers.\333\
                ---------------------------------------------------------------------------
                 \333\ NHTSA-2018-0067-12039, UCS, Technical Appendix, pp. 36-40.
                ---------------------------------------------------------------------------
                 In addition, some commenters indicated that the agencies' analysis
                (impliedly, its modeling) should account for some States' mandates that
                manufacturers sell minimum quantities of ``Zero Emission Vehicles''
                (ZEVs).334 335
                ---------------------------------------------------------------------------
                 \334\ NHTSA-2018-0067-12036, Volvo, p. 5.
                 \335\ NHTSA-2018-0067-11813, South Coast AQMD, Attachment 1, p.
                4 and EIS comments, p. 9.
                ---------------------------------------------------------------------------
                 Regarding the model's representation of engineering and product
                planning constraints, the agencies maintain that having such
                constraints produces more realistic potential (as mentioned above, not
                ``predicted'') pathways forward from manufacturers' current fleets than
                would be the case were these constraints removed. For example, while
                manufacturers' product plans are protected as confidential business
                information (CBI), some manufacturers' public comments demonstrate
                year-by-year balancing such as the CAFE model emulates.\336\ Also, even
                manufacturers that have invested in technologies such as hybrid
                electric powertrains and Atkinson cycle engines have commented that a
                manufacturers' past investments will constrain the pathways it can
                practicably take.\337\ Therefore, the agencies have retained the
                model's basic structural constraints, have updated and expanded the
                model's technology paths (and, as discussed, the model's logic for
                approaching these paths), and have updated inputs defining the range of
                manufacturer-, technology-, and product-specific constraints. These
                updates are discussed below at greater length.
                ---------------------------------------------------------------------------
                 \336\ See, e.g., FCA, pp. 5-6.
                 \337\ Toyota, Attachment 1, p. 10.
                ---------------------------------------------------------------------------
                 The agencies have also reconsidered opportunities manufacturers may
                have to expand the application of technologies that contribute to
                compliance by improving air conditioner efficiency or otherwise
                reducing ``off cycle'' fuel consumption and CO2 emissions,
                or to earn credit toward CO2 compliance by using
                refrigerants with lower global warming potential (GWP) or reducing the
                potential for refrigerant leaks. The version of the model used for the
                proposal accommodates inputs that, for each of these adjustments or
                credits, applies the same value to every model year. The agencies have
                revised the model to accommodate inputs that specify the degree of
                adjustment or credit separately for each model year, and have applied
                inputs that assume manufacturers will increase application of these
                improvements to the highest levels reported within the industry.
                 Regarding comments on the effective cost metric the model uses to
                compare and select among available options to add technology, the
                agencies have considered changes such as those mentioned above. Given
                the myriad of factors that manufacturers can consider, any weighing to
                be conducted using publicly-available information will constitute a
                simplified representation. Nevertheless, within the model's context, it
                is obvious that any weighing of options should, at a minimum, consider
                some measure of each option's costs and benefits. Since this aspect of
                the model involves simulating manufacturers' decisions, it is also
                clearly appropriate that these costs and benefits be considered from a
                manufacturer perspective rather than a social perspective.
                 The effective cost metric used for the NPRM version of the model
                represents the cost of a given option as the cost to apply a given
                technology to a given set of vehicles, and represents the benefit of
                the same option as the extent to which the manufacturer might expect
                buyers would be willing to pay for fuel economy (as represented by a
                portion of the projected fuel savings), combined with any reduction in
                CAFE civil penalties that the manufacturer might ultimately need to
                pass along to buyers. The reduction in CAFE civil penalties places a
                value on progress made toward compliance with CAFE standards. The CAA
                provides no direction regarding CO2 standards, so the model
                accepts inputs specifying an analogous basis for valuing changes in the
                quantity of CO2 credits earned from (or required by) a
                manufacturer's fleet. Because each of these three components
                (technology cost, fuel benefit, and compliance benefit) is expressed in
                dollars, subtracting benefits from costs produces a net cost, and after
                dividing net costs by the number of affected vehicles, it is logical
                to, at each step, select the option that produces the most negative net
                unit cost. This approach can be interpreted as maximizing net benefits
                (to the manufacturer).
                 As an alternative, the agencies considered a simpler metric that
                considers only the cost of the option and the extent to which the
                option increases the quantity of earned credits, and does not require
                input assumptions regarding how to value progress toward compliance.
                Such a metric is expressed in dollars per ton or dollars per gallon
                such that seeking options that produce the smallest (positive) values
                can be interpreted as maximizing cost effectiveness (of progress toward
                compliance). However, simply comparing technology costs to
                corresponding compliance improvements would implicitly assume that
                manufacturers do not respond at all to fuel prices. This assumption is
                clearly unrealistic. For example, if diesel fuel costs $5 per gallon
                and gasoline costs $2 per gallon, manufacturers will be reluctant to
                respond to stringent CAFE or CO2 standards by replacing
                gasoline engines with diesel engines. Manufacturers' comments credibly
                assert that fuel prices matter, and in the agencies' judgment,
                simulations of decisions between available options should continue to
                account for avoided fuel outlays.
                 On the other hand, while any metric should incorporate some measure
                of progress toward compliance, it is not obvious that this progress
                must be expressed in monetary terms. While the CAFE civil penalty
                provisions provide a logical basis for doing so with respect to CAFE,
                the recently-introduced (through EISA) option to trade credit between
                manufacturers adds an alternative basis that is undefined and
                uncertain, in part because terms of past trades are not known to the
                agencies. Also, as mentioned above, EPCA/EISA's civil penalty
                provisions are not applicable to noncompliance with CO2
                standards.
                 Therefore, for the purpose of selecting among available options to
                add technology, the agencies consider it reasonable to use the degree
                of compliance improvement in ``raw'' (i.e., not monetized) form, and to
                divide net costs (i.e., technology costs minus a portion of expected
                avoided fuel outlays) by this improvement. Under a range of side-by-
                side tests, this change to the effective cost metric most frequently
                produced lower overall estimates of compliance costs. However,
                differences vary among manufacturers, model years, and regulatory
                alternatives, and also depend on other model inputs. For example, at
                high fuel prices, the new metric tends to select more expensive
                pathways than the NPRM's metric, and with the new metric, a case
                simulating ``perfect trading'' of CO2 compliance credits
                tends to show such trading increasing compliance costs rather than, as
                expected, decreasing such costs.
                 The version of the model used for the proposal simulates the
                potential that, for
                [[Page 24280]]
                a given fleet in a given model year, a manufacturer might be able to
                use credits from an earlier model year or a different fleet. This
                version of the model did not explicitly simulate the potential that,
                for a given fleet in a given model year, a manufacturer might be to use
                credits from a future model year or a different manufacturer. However,
                the agencies did apply model inputs that reflected assumptions
                regarding possible trading of credits actually earned prior to model
                year 2016 (the earliest represented in detail in the agencies'
                analysis), and the agencies did examine a case (included in the
                sensitivity analysis) involving hypothetical ``perfect'' trading of
                CO2 credits among manufacturers by treating the industry as
                a single ``manufacturer.'' Although past versions of the CAFE Model had
                included code under development with a view toward eventually
                simulating one or both of these provisions, this code had never
                proceeded beyond preliminary experimentation, and had never been the
                focus of peer reviews or application in published analyses.
                 Nevertheless, the agencies considered expanding the model to
                simulate credit ``carry back'' (or ``borrowing'') and trading
                (explicitly, rather than in an idealized hypothetical way). The
                agencies closely examined the corresponding model revisions proposed by
                UCS and determined that such methods would not produce repeatable
                results. This is because the approach proposed by UCS ``randomly swaps
                items in list to minimize trading bias.'' \338\
                ---------------------------------------------------------------------------
                 \338\ UCS, NHTSA-2018-0067-12039, Technical Appendix, at 84-87.
                ---------------------------------------------------------------------------
                 Even if such revisions could be modified to produce non-random
                results, including credit banking and trading would introduce highly
                speculative elements into the agencies' analysis. While manufacturers
                have occasionally indicated plans to carry back credits from future
                model years, those plans have sometimes backfired when projected
                credits have failed to materialize, e.g., by misjudging consumer demand
                for more efficient vehicles. In the agencies' judgment, it would be
                inappropriate to set standards based on an analysis that relies on the
                type of borrowing that has been known to fail. To rely also on credit
                trading during the model years included in the analysis would compound
                this undue speculation. For example, including credit borrowing and
                trading throughout the analysis, as some commenters proposed, would
                lead to an analysis that depends on the potential that, in order to
                comply with the MY 2022 standard for light trucks, FCA could use
                credits it expects to be able to buy from another manufacturer in MY
                2025. Even if the agencies' analysis had knowledge of and made use of
                manufacturers' actual product plans, expectations about the ability to
                borrow others' unearned credits would necessarily be considered risky
                and unreliable. Within an analysis that, to provide for public
                disclosure, extrapolates forward many years from the most recent
                observed fleet, such transactions would add an unreasonable level of
                speculation. Therefore, the agencies have declined to introduce credit
                borrowing and trading into the model's logic.
                 The analysis presented in the proposal applied inputs reflecting
                potential application of credits earned earlier than the first year
                modeled explicitly. However, as observed by some commenters, those
                inputs did not fully account for the one-time exemption from the 5-year
                limit on the extent to which manufacturers may carry forward
                CO2 credits. The agencies have updated the analysis fleet to
                MY 2017 and, in doing so, have updated inputs specifying how credits
                earned to MY 2017 might be applied. These updates implement a
                reasonably full accounting of these ``legacy'' credits, including of
                the one-time exemption from the credit life limit.
                 As mentioned above, some commenters also indicated that the model
                is unrealistically ``reluctant'' to apply credits carried forward from
                early model years. As explained in the proposal and in the model
                documentation, the model's application of carried-forward credits is
                partially controlled by model inputs, which, for the proposal, were set
                to assume that manufacturers would tend to retain credits as long as
                possible. This assumption is entirely consistent with manufacturers'
                past practice and logical in a context wherein the stringency of
                standards is generally increasing over time. Even though using credits
                in some model years might seem initially advantageous, doing so means
                foregoing actual improvements likely to be needed in later model years.
                 Regarding the model's treatment of mandates and credits for the
                sale of ZEVs, as indicated in the model documentation accompanying the
                proposal, these capabilities were experimental in that version of the
                model. The reference case analysis for today's notice, like that for
                the proposal, does not simulate compliance with ZEV mandates.\339\
                ---------------------------------------------------------------------------
                 \339\ The agencies note their finalization of the One National
                Program Final Action, in which EPA partially withdrew a waiver of
                CAA preemption previously granted to the State of California
                relating to its ZEV mandate, and NHTSA finalized regulations
                providing that State ZEV mandates are impliedly and expressly
                preempted by EPCA. This joint action is available at 84 FR 51310.
                ---------------------------------------------------------------------------
                 For the NPRM, the CAFE model was exercised with inputs extending
                this explicit simulation of technology application through MY 2032, as
                the agencies anticipated this was sufficiently beyond MY 2026 that
                nearly all multiyear planning attributable to MY 2026 standards should
                be accounted for, and any compliance credits carried forward from MY
                2026 would have expired. The analysis met this expectation, and the
                agencies presented analysis of the resultant estimated impacts over the
                useful lives of vehicles produced prior to MY 2030. The agencies
                invited comment on all aspects of the analysis, and relevant to this
                aspect of the analysis--i.e., its perspective and temporal span--EDF
                stated that that these led the agencies to overstate the proposal's
                positive impacts on safety, in part because by explicitly representing
                vehicle model years only through 2032, the agencies had failed to
                account for the impact of distant model years prices and fuel economy
                levels on the retention and scrappage of vehicles produced through MY
                2029.\340\ For example, some vehicles produced in MY 2026 will likely
                still be on the road during calendar years (CY) 2033-2050 and the rates
                at which these MY 2026 vehicles will be scrapped during CYs 2033-2050
                will be impacted by the prices and fuel economy levels of vehicles
                produced during MYs 2033-2050.
                ---------------------------------------------------------------------------
                 \340\ EDF, NHTSA-2018-0067-12108, Attachment A at 11 and
                Attachment B at 11-28.
                ---------------------------------------------------------------------------
                 The agencies have addressed this comment by expanding model inputs
                to extend the explicit simulation of technology application through MY
                2050. Most of these expanded model inputs involve the analysis fleet
                and inputs defining the cost and availability of various fuel-saving
                technologies. These inputs are discussed below. The agencies also made
                minor modifications to the model in order to extend model outputs to
                cover this wider span and to carry forward each regulatory
                alternative's standards automatically through the last year to be
                modeled (e.g., extending standards without change from MY 2032 through
                MY 2050). The model documentation discusses these
                [[Page 24281]]
                minor changes.\341\ In addition, although the agencies published
                detailed model output files documenting all estimated annual impacts
                through calendar year 2089, the notice and PRIA both emphasized the
                above-mentioned ``model year'' perspective, as in past regulatory
                analyses supporting CAFE and CO2 standards. Recognizing that
                an alternative ``calendar year'' perspective is of interest to EDF and,
                perhaps other stakeholders, the agencies have expanded the presentation
                of results in today's notice and FRIA by presenting some physical
                impacts (e.g., fuel consumption and CO2 emissions) as well
                as monetized benefits, costs, and net benefits for each of CYs 2017-
                2050. All of these results appear in the model output files published
                with today's notice, as do corresponding results for more specific
                impacts (e.g., year-by-year components of monetized social costs).\342\
                ---------------------------------------------------------------------------
                 \341\ The model and documentation are available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
                 \342\ Detailed model inputs and outputs are available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
                ---------------------------------------------------------------------------
                5. Calculation of Physical Impacts
                 Once it has completed the simulation of manufacturers' potential
                application of technology in response to CAFE/CO2 standards
                and fuel prices, the CAFE Model calculates impacts of the resultant
                changes in new vehicle fuel economy levels and prices. This involves
                several steps.
                 The model calculates changes in the total quantity of new vehicles
                sold in each model year as well as the relative shares passenger cars
                and light trucks comprise of the overall new vehicle market. The
                agencies received many comments on the estimation of sales impacts, and
                as discussed below, today's analysis applies methods and corresponding
                estimates that reflect careful consideration of these comments. Related
                to these calculations, the model now operates in an iterated fashion
                with a view toward obtaining sales impacts that are balanced with
                changes in vehicle prices and fuel economy levels. This involves
                solving for compliance, calculating sales impacts, re-solving for
                compliance, and repeating these steps as many times as specified in
                model inputs. For today's analysis, the agencies operated the model
                with four iterations, as early testing suggested three iterations
                should be sufficient for fleetwide results to converge between
                iterations. The model documentation describes the procedures for
                iteration in detail.
                 The impacts on outlays for new vehicles occur coincident with the
                sale of these vehicles so the model can simply calculate and record
                these for each model year included in the analysis. However, virtually
                all other impacts result from vehicle operation that extends long after
                a vehicle is produced. Like other models (including, e.g., NEMS), the
                CAFE Model includes procedures (sometimes referred to as ``stock
                models'' or as models of fleet turnover) to estimate annual rates at
                which new vehicles are used and subsequently scrapped. The agencies
                received many comments on procedures for estimating vehicle scrappage
                and on procedures for estimating annual quantities of highway travel,
                accounting for the elasticity of travel demand with respect to per-mile
                costs for fuel. Below, Section VI.D.1 discusses these comments and
                reviews procedures and corresponding estimates that also reflect
                careful consideration of these comments.
                 For each vehicle model in each model year, these procedures result
                in estimates of the number of vehicles remaining in service in each
                calendar year, as well as the annual mileage accumulation (i.e.,
                vehicle miles traveled, or VMT) in each calendar year. As mentioned
                above, most of the physical impacts of interest derive from this
                vehicle operation. Also discussed above, the simulated application of
                technology results in ``initial'' and ``final'' estimates of the cost,
                fuel type, fuel economy, and fuel share (for, in particular, PHEVs that
                can run on gasoline or electricity) applicable to each vehicle model in
                each model year. Together with quantities of travel, and with estimates
                of the ``gap'' between ``laboratory'' and ``on-road'' fuel economy,
                these enable calculation of quantities of fuel consumed in each year
                during the useful life of each vehicle model produced in each model
                year.\343\ The model documentation provides specific procedures and
                formulas implementing these calculations.
                ---------------------------------------------------------------------------
                 \343\ The agencies have applied the same estimates of the ``on
                road gap'' as applied for the analysis supporting the NPRM. For
                operation on gasoline, diesel, E85, and CNG, this gap is 20 percent;
                for electricity and hydrogen, 30 percent.
                ---------------------------------------------------------------------------
                 As for the NPRM, the model calculates emissions of CO2
                and other air pollutants, reporting emissions both from vehicle
                tailpipes and from upstream processes (e.g., petroleum refining)
                involved in producing and supplying fuels. Section VI.D.3 below reviews
                methods, models, and estimates used in performing these calculations.
                The model also calculates impacts on highway safety, accounting for
                changes in travel demand, changes in vehicle mass, and continued past
                and expected progress in vehicle safety (through, e.g., the application
                of new crash avoidance systems). Section VI.D.2 discusses methods, data
                sources, and estimates involved in estimating safety impacts, comments
                on the same, and changes included in today's analysis. In response to
                the NPRM, some comments urged the agencies also to quantify different
                types of health impacts from changes in air pollution rather than only
                accounting for such impacts in aggregate estimates of the social costs
                of air pollution. Considering these comments, the agencies added such
                calculations to the model, as discussed in Section VI.D.3.
                6. Calculation of Benefits and Costs
                 Having estimated how technologies might be applied going forward,
                and having estimated the range of resultant physical impacts, the CAFE
                Model calculates a variety of private and social benefits and costs,
                reporting these from the consumer, manufacturer, and social
                perspectives, both in undiscounted and discounted present value form
                (given inputs specifying the corresponding discount rate and present
                year). Estimates of regulatory costs are among the direct outputs of
                the simulation of manufacturers' potential responses to new standards.
                Other benefits and costs are calculated based on the above-mentioned
                estimates of travel demand, fuel consumption, emissions, and safety
                impacts. The agencies received many comments on the NPRM's calculation
                of benefits and costs, and Section VI.D.1 discusses these comments and
                presents the methods, data sources, and estimates used in calculating
                benefits and costs reported here.
                7. Structure of Model Inputs and Outputs
                 All CAFE Model inputs and outputs described above are specified in
                Microsoft Excel format, and the user can define and edit all inputs to
                the system. Table VI-3 describes (non-exhaustively) which inputs are
                contained within each input file and Table VI-4 describes which outputs
                are contained in each output file. This is important for three reasons:
                (1) Each file is discussed throughout the following sections; (2)
                several commenters conflated aspects of the model with its inputs; and
                (3) several commenters seemed confused about where to find specific
                information in the output files. This information was described in
                detail in the NPRM CAFE Model Documentation, but is reproduced here for
                quick reference. When specifically referencing the input
                [[Page 24282]]
                or output file used for the NPRM or final rule in the following
                discussion, NPRM or FRM, respectively, will precede the file name.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.080
                [GRAPHIC] [TIFF OMITTED] TR30AP20.081
                 A catalog of the Argonne National Laboratory Autonomie fuel economy
                technology effectiveness value output files are reproduced in the
                following Table VI-5 as well. The left column shows the terminology
                used in this text to refer to the file, while the right column
                describes each file. NPRM or FRM, respectively, may precede the
                terminology in the text as appropriate.
                [[Page 24283]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.082
                 Finally, Table VI-6 lists the terminologies used to refer to other
                model-related documents which are referred to frequently throughout the
                text. NPRM or FRM, respectively, may precede the terminology in the
                text as appropriate.
                [[Page 24284]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.083
                B. What inputs does the compliance analysis require?
                1. Analysis Fleet
                 The starting point for the evaluation of the potential feasibility
                of different stringency levels for future CAFE and CO2
                standards is the analysis fleet, which is a snapshot of the recent
                vehicle market. The analysis fleet provides a baseline from which to
                project what and how additional technologies could feasibly be applied
                to vehicles in a cost-effective manner to raise those vehicles' fuel
                economy and lower their CO2 emission levels.\344\ The fleet
                characterization also provides a reference point with data for other
                factors considered in the analysis, including environmental effects and
                effects estimated by the economic modules (i.e., sales, scrappage, and
                labor utilization). When the scope of the analysis widens, another
                piece of data must be included for each vehicle in the analysis fleet
                to map a given element of the fleet appropriately onto an analysis
                module.
                ---------------------------------------------------------------------------
                 \344\ The CAFE model does not generate compliance paths a
                manufacturer should, must, or will deploy. It is intended as a tool
                to demonstrate a compliance pathway a manufacturer could choose. It
                is almost certain all manufacturers will make compliance choices
                differing from those projected by the CAFE model.
                ---------------------------------------------------------------------------
                 For the analysis presented in this final rule, the analysis fleet
                includes information about vehicles that is essential for each analysis
                module. The first part of projecting how additional technologies could
                be applied to vehicles is knowing which vehicles are produced by which
                manufacturers, the fuel economies of those vehicles, how many of each
                are sold, whether they are passenger cars or light trucks, and their
                footprints. This is important because it improves understanding of the
                overall impacts of different levels of CAFE and CO2
                standards; overall impacts that result from industry's response to
                standards, and industry's response, is made up of individual
                manufacturer responses to the standards in light of the overall market
                and their individual assessment of consumer acceptance. Establishing an
                accurate representation of manufacturers' existing fleets (and the
                vehicle models in them) that will be subject to future standards helps
                in predicting potential individual manufacturer responses to those
                future standards in addition to potential changes in those standards.
                 Another part of projecting how additional fuel economy improving
                technologies could be applied to vehicles is knowing which fuel saving
                technologies manufacturers have equipped on which vehicles. In many
                cases, the agencies also collect and reference additional information
                on other vehicle attributes to help with this process.\345\ Accounting
                for technologies already applied to vehicles helps avoid ``double-
                counting'' the value of those technologies, by assuming they are still
                available to be applied to improve fuel economy and reduce
                CO2 emissions. It also promotes more realistic
                determinations of what additional technologies can feasibly be applied
                to those vehicles: If a manufacturer has already started down a
                technological path to fuel economy or performance improvements, the
                agencies do not assume it will completely abandon that path because
                doing so would be unrealistic and fails to represent accurately
                manufacturer responses to standards. Each vehicle model (and
                configurations of each model) in the analysis fleet, therefore, has a
                comprehensive list of its technologies, which is important because
                different configurations may have different technologies applied to
                them.\346\ In addition, to properly account for technology costs, the
                agencies assign each vehicle to a technology class and an engine class.
                Technology classes reference each vehicle to a set of full vehicle
                simulations, so that the agencies may project fuel efficiency with
                combinations of additional fuel saving equipment and hybrid and
                electric vehicle battery costs.
                ---------------------------------------------------------------------------
                 \345\ For instance, curb weight, horsepower, drive
                configuration, pickup bed length, oil type, body style, aerodynamic
                drag coefficients, and rolling resistance coefficients, and (if
                applicable) battery sizes are all required to assign technology
                content properly.
                 \346\ Considering each vehicle model/configuration also improves
                the ability to consider the differential impacts of different levels
                of potential standards on different manufacturers, since all vehicle
                model/configurations ``start'' at different places, in terms of
                technologies already used and how those technologies are used.
                ---------------------------------------------------------------------------
                 Yet another part of projecting which vehicles might exist in future
                model years is developing reasonable real-world assumptions about when
                and how manufacturers might apply certain technologies to vehicles. The
                analysis fleet accounts for links between vehicles, recognizing vehicle
                platforms will share technologies, and the vehicles that make up that
                platform should receive (or not receive) additional technological
                improvements together. Shared engines, shared transmissions, and shared
                vehicle platforms for mass reduction technology are considered. In
                addition, each vehicle model/configuration in the analysis fleet also
                has information about its redesign
                [[Page 24285]]
                schedule, i.e., the last year it was redesigned and when the agencies
                expect it to be redesigned again. Redesign schedules are a key part of
                manufacturers' business plans, as each new product can cost more than
                $1B, and involve a significant portion of a manufacturer's scarce
                research, development, and manufacturing and equipment budgets and
                resources.\347\ Manufacturers have repeatedly told the agencies that
                sustainable business plans require careful management of resources and
                capital spending, and that the length of time each product remains in
                production is crucial to recouping the upfront product development and
                plant/equipment costs, as well as the capital needed to fund the
                development and manufacturing equipment needed for future products.
                Because the production volume of any given vehicle model varies within
                a manufacturer's product line, and varies among different
                manufacturers, redesign schedules typically vary for each model and
                manufacturer. Some (relatively few) technological improvements are
                small enough that they can be applied in any model year; a few other
                technological improvements may be applied during a refreshening (when a
                few additional changes are made, but well short of a full redesign),
                but others are major enough that they can only be cost-effectively
                applied at a vehicle redesign, when many other things about the vehicle
                are already changing. Ensuring the CAFE model makes technological
                improvements to vehicles only when it is feasible to do so also helps
                the analysis better represent manufacturer responses to different
                levels of standards.
                ---------------------------------------------------------------------------
                 \347\ Shea, T., Why Does It Cost So Much For Automakers To
                Develop New Models? Autoblog (Jul. 27, 2010), https://www.autoblog.com/2010/07/27/why-does-it-cost-so-much-for-automakers-to-develop-new-models/.
                ---------------------------------------------------------------------------
                 Finally, the agencies restrict the applications of some
                technologies on some vehicles upon determining the technology is not
                compatible with the functional and performance requirements of the
                vehicle, or if the manufacturers are unlikely to apply a specific
                technology to a specific vehicle for reasons articulated with
                confidential business information that the agencies found credible.
                 Other data important for the analysis that are referenced to the
                analysis fleet include baseline economic, environmental, and safety
                information. Vehicle fuel tank size is required to estimate range and
                refueling benefit while curb weights and safety class assignments help
                the agencies consider how changes in vehicle mass may affect safety.
                The agencies identify the final assembly location for each vehicle,
                engine, and transmission, as well as the percent of U.S. content to
                support the labor impact analysis. In addition, the aforementioned
                accounting for first-year vehicle production volumes (i.e., the number
                of vehicles of each new model sold in MY 2017, for this analysis) is
                the foundation for estimating how future vehicle sales might change in
                response to different potential standards.
                 The input file for the CAFE model characterizing the analysis
                fleet, referred to as the ``market inputs'' file or ``market data''
                file, accordingly includes a large amount of data about vehicles, their
                technological characteristics, the manufacturers and fleets to which
                they belong, and initial prices and production volumes, which provide
                the starting points for projection (by the sales model) to ensuing
                model years. In the Draft TAR (which utilized a MY 2015 analysis fleet)
                and NPRM (which utilized a MY 2016 analysis fleet), the agencies needed
                to populate about 230,000 cells in the market data file to characterize
                the fleet. For this final rule (which utilized a MY 2017 analysis
                fleet), the agencies populated more than 400,000 cells to characterize
                the fleet. While the fleet is not actually much more heterogeneous in
                reality,\348\ the agencies have provided and collected more data to
                justify the characterization of the analysis fleet, and to support the
                functionality of modules in the CAFE model.
                ---------------------------------------------------------------------------
                 \348\ The expansion of cells is primarily due to (1) considering
                more technologies, and (2) listing trim levels separately, which
                often yields more precise curb weights and more accurate
                manufacturer suggested retail prices.
                ---------------------------------------------------------------------------
                 A solid characterization of a recent model year as an analytical
                starting point helps realistically estimate ways manufacturers could
                potentially respond to different levels of standards, and the modeling
                strives to simulate realistically how manufacturers could progress from
                that starting point. While manufacturers can respond in many ways
                beyond those represented in the analysis (e.g., applying other
                technologies, shifting production volumes, changing vehicle footprint),
                such that it is impossible to predict with any certainty exactly how
                each manufacturer will respond, it is still important to establish a
                solid foundation from which to estimate potential costs and benefits of
                potential future standards. The following sections discuss aspects of
                how the analysis fleet was built for this analysis, and includes
                discussion of the comments on fleet that the agencies received on the
                proposed rule.
                a) Principles on Data Sources Used To Populate the Analysis Fleet
                 The source data for vehicles in the analysis fleet and their
                technologies is a central input for the analysis. The sections below
                discuss pros and cons of different potential sources and what the
                agencies used for this analysis, and responds to comments the agencies
                received on data sources in the proposal.
                (1) Use of Confidential Business Information Versus Publicly-Releasable
                Sources
                 Since 2001, CAFE analysis has used either confidential, forward-
                estimating product plans from manufacturers, or publicly available data
                on vehicles already sold as a starting point for determining what
                technologies can be applied to what vehicles in response to potential
                different levels of standards. The use of either data source requires
                certain tradeoffs. Confidential product plans comprehensively represent
                what vehicles a manufacturer expects to produce in coming years,
                accounting for plans to introduce new vehicles and fuel-saving
                technologies and, for example, plans to discontinue other vehicles and
                even brands. This information can be very thorough and can improve the
                accuracy of the analysis, but cannot be publicly released. This makes
                it difficult for public commenters to reproduce the analysis for
                themselves as they develop their comments. Some non-industry commenters
                have also expressed concern about manufacturers having an incentive in
                the submitted plans to underestimate (deliberately or not) their future
                fuel economy capabilities and overstate their expectations about, for
                example, the levels of performance of future vehicle models in order to
                affect the analysis. Accordingly, since 2010, EPA and NHTSA have based
                analysis fleets almost exclusively on information from commercial and
                public sources, starting with CAFE compliance data and adding
                information from other sources.
                 An analysis fleet based primarily on public sources can be released
                to the public, solving the issue of commenters being unable to
                reproduce the overall analysis. However, industry commenters have
                argued such an analysis fleet cannot accurately reflect manufacturers'
                actual plans to apply fuel-saving technologies (e.g., manufacturers may
                apply turbocharging to improve not just fuel economy, but also to
                improve vehicle performance) or manufacturers' plans to change product
                offerings by introducing some vehicles and brands and discontinuing
                other
                [[Page 24286]]
                vehicles and brands, precisely because that information is typically
                confidential business information (CBI). A fully-publicly-releasable
                analysis fleet holds vehicle characteristics unchanged over time and
                lacks some level of accuracy when projected into the future. For
                example, over time, manufacturers introduce new products and even
                entire brands. On the other hand, plans announced in press releases do
                not always ultimately bear out, nor do commercially available third-
                party forecasts. Assumptions could be made about these issues to
                improve the accuracy of a publicly releasable analysis fleet, but
                concerns include that this information would either be largely
                incorrect, or, if the assumptions were correct, information would be
                released that manufacturers would consider CBI.
                 Furthermore, some technologies considered in the rulemaking are
                difficult to observe in the analysis fleet without expensive teardown
                study and time-consuming benchmarking. Not giving credit for these
                technologies puts the analysis at significant risk of double-counting
                the effectiveness of these technologies, as manufacturers cannot equip
                technologies twice to the same vehicle for double the fuel economy
                benefit. As discussed in the Draft TAR, the agencies assigned little
                (if any) technology application in the baseline fleet for some of these
                technologies.\349\ For the NPRM MY 2016 fleet development process, the
                agencies again offered the manufacturers the opportunity to volunteer
                CBI to the agencies to help inform the technology content of the
                analysis fleet, and many manufacturers did. The agencies were able to
                confirm that many manufacturers had already included many hard-to-
                observe technologies in the MY 2016 fleet (which they were not properly
                given credit for in the characterization of the MY 2014 and MY 2015
                fleets presented in Draft TAR) so the agencies reflected this new
                information in the NPRM analysis and in the analysis presented today.
                ---------------------------------------------------------------------------
                 \349\ These technologies include low rolling resistance
                technology (incorrectly applied to zero baseline vehicles in Draft
                TAR), low-drag brakes (incorrectly applied to zero baseline vehicles
                in Draft TAR), electric power steering (incorrectly applied to too
                few vehicles in Draft TAR), accessory drive improvements
                (incorrectly applied to zero baseline vehicles in Draft TAR), engine
                friction reduction (previously named LUBEFR1, LUBEFR2, and LUBEFR3),
                secondary axle disconnect and transmission improvements.
                ---------------------------------------------------------------------------
                 In addition, many manufacturers provided confidential comment on
                the potential applicability of fuel-saving technologies to their fleet.
                In particular, many manufacturers confidentially identified specific
                engine technologies that they will not use in the near term, either on
                specific vehicles, or at all. Reasons varied: Some manufacturers cited
                intellectual property concerns, and others stated functional
                performance concerns for some engine types on some vehicles. Other
                manufacturers shared forward-looking product plans, and explained that
                it would be cost prohibitive to scrap significant investments in one
                technology in favor of another. This topic is discussed in more detail
                in Section VI.B.1.b)(6), below.
                 The agencies sought comment on how to address this issue going
                forward, recognizing both the competing interests involved and the
                typical timeframes for CAFE and CO2 standards rulemakings.
                 Many commenters expressed concern with the agencies using any CBI
                as part of the rulemaking process. Some commenters expressed concern
                that use of CBI would make the CAFE model subject to inaccuracies
                because manufacturers would only provide additional information in
                situations in which a correction to the agencies' baseline assumptions
                would favor the manufacturers.\350\ The agencies recognize this as a
                reasonable concern, but the analysis presented in the Draft TAR
                consistently assumed very little (if any) technology had been applied
                in the baseline. In addition, many manufacturers shared information on
                advanced technologies that were not yet in production in MY 2017, but
                could be used in the future; manufacturer contributions helped the
                agencies better model many advanced engine technologies and to include
                them in today's analysis, and inclusion of these technologies (and
                costs) in the analysis sometimes lowered the projected cost of
                compliance for stringent alternatives. Other commenters expressed
                concern that automakers would supply false or incomplete information
                that would unduly restrict what technologies can be deployed.\351\ When
                possible, the agencies sought independently to verify manufacturer CBI
                (or claims made by other stakeholders) through lab testing and
                benchmarking.\352\ The agencies found no evidence of misrepresentation
                of engineering specifications in the MY 2017 fleet in manufacturer CBI;
                instead, the agencies were able to verify independently many CBI
                submissions, and confirm the credibility of information provided from
                those sources.
                ---------------------------------------------------------------------------
                 \350\ NHTSA-2018-0067-12039, Union of Concerned Scientists.
                 \351\ NHTSA-2018-0067-11741, ICCT.
                 \352\ For instance, the agencies continue to evaluate tire
                rolling resistance on production vehicles via independent lab
                testing, and the agencies bench-marked the operating behavior and
                calibration of many engines and transmissions.
                ---------------------------------------------------------------------------
                 Some commenters requested that more CBI be used in the analysis.
                For instance, some commenters suggested that the agencies should return
                to the use of product plans and announcements regarding future fleets
                because manufacturers had already committed investments to bring
                announced products to market.\353\ However, if the agencies were to
                assume that these commitments were already in the baseline, the
                agencies would underestimate the cost of compliance for stringent
                alternatives. Moreover, while upfront investments to bring technologies
                to market are significant, the total marginal costs of components are
                typically large in comparison over the entire product life-cycle, and
                these costs have not yet been realized in vehicles not yet produced.
                ---------------------------------------------------------------------------
                 \353\ NHTSA-2018-0067-11956, PA Department of Environmental
                Protection.
                ---------------------------------------------------------------------------
                 The agencies did make use of some forward-looking CBI in the
                analysis. The agencies received many comments from manufacturers on the
                technological feasibility, or functional applicability of some fuel
                saving technologies to certain vehicles, or certain vehicle
                applications, and the agencies took this information into consideration
                when projecting compliance pathways. These cases are discussed
                generally in Section VI.B.1.b)(6), below, and specifically for each
                technology in those technology sections. Some commenters expressed that
                the use of CBI for future product plans would be acceptable, but only
                if the agencies disclosed the CBI affecting all vehicles through MY
                2025 at the time of publication.\354\ Functionally, this is not
                possible. Manufacturer's confidential product plans cannot be made
                public, as prohibited under NHTSA's regulations at 49 CFR part 512, and
                if the information meets the requirements of section 208(c) of the
                Clean Air Act. If the agencies disclosed confidential information, it
                would not only violate the terms on which the agencies obtained the
                CBI, but it is unlikely that manufacturers would continue to offer CBI,
                which in turn would likely degrade the quality of the analysis. The
                agencies believe that the use of CBI in the NPRM and final rule
                analysis--to confirm, reference, or to otherwise modify aspects of the
                analysis that can be made public--threads the needle between a more
                accurate but less transparent analysis (using more CBI) and a less
                accurate but more transparent analysis (using less CBI).
                ---------------------------------------------------------------------------
                 \354\ NHTSA-2018-0067-11741.
                ---------------------------------------------------------------------------
                [[Page 24287]]
                (2) Source Data and Vintage Used in the Analysis
                 Based on the assumption that a publicly-available analysis fleet
                continued to be desirable, manufacturer compliance submissions to EPA
                and NHTSA were used as a starting point for the NPRM and final rule
                analysis fleets. Generally, manufacturer compliance submissions break
                down vehicle fuel economy and production volume by regulatory class,
                and include some very basic product information (typically including
                vehicle nameplate, engine displacement, basic transmission information,
                and drive configuration). Many different trim levels of a product are
                typically rolled up and reported in an aggregated fashion, and these
                groupings can make decomposition of different fuel-saving, road load
                reducing technologies extremely difficult. For instance, vehicles in
                different test weight classes, with different tires or aerodynamic
                profiles may be aggregated and reported together.\355\ A second portion
                of the compliance submission summarizes production volume by vehicle
                footprints (a key compliance measure for standard setting) by
                nameplate, and includes some basic information about engine
                displacement, transmission, and drive configuration. Often these
                production volumes by footprint do not fit seamlessly together with the
                production volumes for fuel economy, so the agencies must reconcile
                this information.
                ---------------------------------------------------------------------------
                 \355\ Some fuel-economy compliance information for pickup trucks
                span multiple cab and box configurations, but manufacturers reported
                these disparate vehicles together.
                ---------------------------------------------------------------------------
                 Information from the MY 2016 fleet was chosen as the foundation for
                the NPRM analysis fleet because, at the time the rulemaking analysis
                was initiated, the 2016 fleet represented the most up-to-date
                information available in terms of individual vehicle models and
                configurations, production technology levels, and production volumes.
                If MY 2017 data had been used while this analysis was being developed,
                the agencies would have needed to use product planning information that
                could not be made available to the public until a later date.
                 The NPRM analysis fleet was initially developed with 2016 mid-model
                year compliance data because final compliance data was not available at
                that time, and the timing provided manufacturers the opportunity to
                review and comment on the characterization of their vehicles in the
                fleet. With a view toward developing an accurate characterization of
                the 2016 fleet to serve as an analytical starting point, corrections
                and updates to mid-year data (e.g., to production estimates) were
                sought, in addition to corroboration or correction of technical
                information obtained from commercial and other sources (to the extent
                that information was not included in compliance data), although future
                product planning information from manufacturers (e.g., future product
                offerings, products to be discontinued) was not requested, as most
                manufacturers view such information as CBI. Manufacturers offered a
                range of corrections to indicate engineering characteristics (e.g.,
                footprint, curb weight, transmission type) of specific vehicle model/
                configurations, as well as updates to fuel economy and production
                volume estimates in mid-year reporting. After following up on a case-
                by-case basis to investigate significant differences, the analysis
                fleet was updated.
                 Sales, footprint, and fuel economy values with final compliance
                data were also updated if that data was available. In a few cases,
                final production and fuel economy values were slightly different for
                specific MY 2016 vehicle models and configurations than were indicated
                in the NPRM analysis; however, other vehicle characteristics (e.g.,
                footprint, curb weight, technology content) important to the analysis
                were reasonably accurate. While some commenters have, in the past,
                raised concerns that non-final CAFE compliance data is subject to
                change, the potential for change is likely not significant enough to
                merit using final data from an earlier model year reflecting a more
                outdated fleet. Moreover, even ostensibly final CAFE compliance data is
                frequently subject to later revision (e.g., if errors in fuel economy
                tests are discovered), and the purpose of the analysis was not to
                support enforcement actions but rather to provide a realistic
                assessment of manufacturers' potential responses to future standards.
                 Manufacturers integrated a significant amount of new technology in
                the MY 2016 fleet, and this was especially true for newly-designed
                vehicles launched in MY 2016. While subsequent fleets will involve even
                further application of technology, using available data for MY 2016
                provided the most realistic detailed foundation for analysis that could
                be made available publicly in full detail, allowing stakeholders to
                reproduce the analysis presented in the proposal independently. Insofar
                as future product offerings are likely to be more similar to vehicles
                produced in 2016 than to vehicles produced in earlier model years,
                using available data regarding the 2016 model year provided the most
                realistic, publicly releasable foundation for constructing a forecast
                of the future vehicle market for this proposal. Many comments
                responding to the Draft TAR, EPA's Proposed Determination, EPA's 2017
                Request for Comment, and the NPRM preceding today's notice stated that
                the most up-to-date analysis fleet possible should be used, because a
                more up-to-date analysis fleet will better capture how manufacturers
                apply technology and will account better for vehicle model/
                configuration introductions and deletions.356 357
                ---------------------------------------------------------------------------
                 \356\ 82 FR 39551 (Aug. 21, 2017).
                 \357\ For example, in 2016 comments to dockets EPA-HQ-OAR-2015-
                0827 and NHTSA-2016-0068, the Alliance of Automobile Manufacturers
                commented that ``the Alliance supports the use of the most recent
                data available in establishing the baseline fleet, and therefore
                believes that NHTSA's selection [of, at the time, model year 2015]
                was more appropriate for the Draft TAR.'' Alliance at 82, Docket ID.
                EPA-HQ-OAR-2015-0827-4089. Global Automakers commented that ``a one-
                year difference constitutes a technology change-over for up to 20%
                of a manufacturer's fleet. It was also generally understood by
                industry and the agencies that several new, and potentially
                significant, technologies would be implemented in MY 2015. The use
                of an older, outdated baseline can have significant impacts on the
                modeling of subsequent Reference Case and Control Case
                technologies.'' Global Automakers at A-10, Docket ID EPA-HQ-OAR-
                2015-0827-4009.
                ---------------------------------------------------------------------------
                 On the other hand, some commenters suggested that because
                manufacturers continue improving vehicle performance and utility over
                time, an older analysis fleet should be used to estimate how the fleet
                could have evolved had manufacturers applied all technological
                potential to fuel economy rather than continuing to improve vehicle
                performance and utility.\358\ Because manufacturers change and improve
                product offerings over time, conducting analysis with an older analysis
                fleet (or with a fleet using fuel economy levels and CO2
                emissions rates that have been adjusted to reflect an assumed return to
                levels of performance and utility typical of some past model year)
                would miss this real-world trend. While such an analysis could project
                what industry could do if, for example, manufacturers devoted all
                technological improvements toward raising fuel economy and reducing
                CO2 emissions (and if consumers decided to purchase these
                vehicles), the agencies do not believe it would be consistent with a
                transparent examination of what effects different levels of standards
                would have
                [[Page 24288]]
                on individual manufacturers and the fleet as a whole.
                ---------------------------------------------------------------------------
                 \358\ For example, in 2016 comments to dockets EPA-HQ-OAR-2015-
                0827 and NHTSA-2016-0068, UCS stated ``in modeling technology
                effectiveness and use, the agencies should use 2010 levels of
                performance as the baseline.'' UCS at 4, Docket ID. EPA-HQ-OAR-2015-
                0827-4016.
                ---------------------------------------------------------------------------
                 All else being equal, using a newer analysis fleet will produce
                more realistic estimates of impacts of potential new standards than
                using an outdated analysis fleet. However, among relatively current
                options, a balance must be struck between input freshness, and input
                completeness and accuracy.\359\ During assembly of the inputs for the
                NPRM analysis, final compliance data was available for the MY 2015
                model year but not, in a few cases, for MY 2016. However, between mid-
                year compliance information and manufacturers' specific updates
                discussed above, a robust and detailed characterization of the MY 2016
                fleet was developed. While information continued to develop regarding
                the MY 2017 and, to a lesser extent MY 2018 and even MY 2019 fleets,
                this information was--even in mid-2017--too incomplete and inconsistent
                to be assembled with confidence into an analysis fleet for modeling
                supporting deliberations regarding the NPRM analysis.
                ---------------------------------------------------------------------------
                 \359\ Comments provided through a recent peer review of the CAFE
                model recognize the competing interests behind this balance. For
                example, referring to NHTSA's 2016 Draft TAR analysis, one of the
                peer reviewers commented as follows: ``The NHTSA decision to use MY
                2015 data is wise. In the TAR they point out that a MY 2016
                foundation would require the use of confidential data, which is less
                desirable. Clearly they would also have a qualitative vision of the
                MY 2016 landscape while employing MY 2015 as a foundation. Although
                MY 2015 data may still be subject to minor revision, this is
                unlikely to impact the predictive ability of the model . . . A more
                complex alternative approach might be to employ some 2016 changes in
                technology, and attempt a blend of MY 2015 and MY 2016, while
                relying of estimation gained from only MY 2015 for sales. This
                approach may add some relevancy in terms of technology, but might
                introduce substantial error in terms of sales.''
                ---------------------------------------------------------------------------
                 Manufacturers requested that the baseline fleet supporting the
                final rule incorporate the MY 2018 or most recent information
                available.\360\ Other commenters expressed desire for multiple fleets
                of various vintages to compare the updated model outputs with those of
                previous rule-makings. Specifically, some commenters requested that
                older fleet vintages (MY 2010, for instance) be developed in parallel
                with the MY 2017 fleet so that those too may be used as inputs for the
                model.\361\
                ---------------------------------------------------------------------------
                 \360\ NHTSA-2018-0067-12150, Toyota North America.
                 \361\ NHTSA-2018-0067-11741, ICCT.
                ---------------------------------------------------------------------------
                 Between the NPRM and this final rule, manufacturers submitted final
                compliance data for the MY 2017 fleet. When the agencies pulled
                together information for the fleet for the final rule, the agencies
                decided to use the highest-quality, most up-to-date information
                available. Given that pulling this information together takes some
                time, and given that ``final'' compliance submissions often lag
                production by a few years, the agencies decided to use 2017 model year
                as the base year for the analysis fleet, as the agencies stated in the
                NPRM.\362\ While the agencies could have used preliminary 2018 data or
                even very early 2019 data, this information was not available in time
                to support the final rulemaking. Likewise, the agencies chose not to
                revert to a previous model year (for instance 2016 or 2012) because
                many manufacturers have incorporated fuel savings technologies over the
                last few years, realized some benefits for fuel economy, and adjusted
                the performance or sales mix of vehicles to remain competitive in the
                market. Also, using an earlier model year would provide less accurate
                projections because the analysis would be based on what manufacturers
                could have done in past model years and would have estimated the fuel
                economy improvements instead of using known information on the
                technologies that were employed and the actual fuel economy that
                resulted from applying those technologies.
                ---------------------------------------------------------------------------
                 \362\ 83 FR 43006 (``If newer compliance data (i.e., MY 2017)
                becomes available and can be analyzed during the pendency of this
                rulemaking, and if all other necessary steps can be performed, the
                analysis fleet will be updated, as feasible, and made publicly
                available.'').
                ---------------------------------------------------------------------------
                 Some additional information (about off-cycle technologies, for
                instance) was often not reported by manufacturers in MY 2017 formal
                compliance submissions in a way that provided clear information on
                which technologies were included on which products. As part of the
                formal compliance submission, some manufacturers voluntarily submitted
                additional information (about engine technologies, for instance). While
                this data was generally of very high quality, there were some mistakes
                or inconsistencies with publicly available information, causing the
                agencies to contact the manufacturers to understand and correct
                identified issues. In most cases, however, the formal compliance data
                was very limited in nature, and the agencies collected additional
                information necessary to characterize fully the fleet from other
                sources, and scrutinized additional information submitted by
                manufacturers carefully, independently verifying when possible.
                 Specifically, the agencies downloaded and reviewed numerous
                marketing brochures and product launch press releases to confirm
                information submitted by manufacturers and to fill in information
                necessary for the analysis fleet that was not provided in the
                compliance data. Product brochures often served as the basis for the
                curb weights used in the analysis. This publicly available manufacturer
                information sometimes also included aerodynamic drag coefficients,
                information about steering architecture, start-stop systems, pickup bed
                lengths, fuel tank capacities, and high-voltage battery capacities. The
                agencies recorded vehicle horsepower, compression ratio, fuel-type, and
                recommended oil weight rating from a combination of manufacturer
                product brochures and owner's manuals. The product brochures, as well
                as online references such as Autobytel, informed which combinations of
                fuel saving technologies were available on which trim levels, and what
                the manufacturer suggested retail price was for many products. Overall
                this information proved helpful for assigning technologies to vehicles,
                and for getting data (such as fuel tank size \363\) necessary for the
                analysis. These reference materials have been included in the
                rulemaking documentation.\364\
                ---------------------------------------------------------------------------
                 \363\ The quality of data for today's analysis fleet is notably
                improved for fuel tank capacity, which factors into the calculation
                of refueling time benefits. In many previous analyses, fuel tank
                sizes were often stated as estimates or proxies, and not sourced so
                carefully.
                 \364\ Publicly available data used to supplement analysis fleet
                information is available in the docket.
                ---------------------------------------------------------------------------
                 The agencies elected not to develop fleets of previous model year
                vintages that could be used in parallel as an input to the CAFE model.
                Developing a detailed characterization of the fleet of any vintage
                would be a huge undertaking with few benefits. As the scope has
                increased, and as additional modules are added, going back in time to
                re-characterize a previous fleet in a format that works with CAFE model
                updates can be time- and resource-prohibitive for the agencies, even if
                that work is adapting a fleet that was used in previous rule-making
                analysis. Doing so also offers little value in determining what
                potential fuel saving technology can be added to a more recent fleet
                during the rulemaking timeframe.
                 The MY 2017 manufacturer-submitted data, verified and supplemented
                by the agencies with publicly-available information, therefore
                presented the fullest, most up-to-date data set that the agencies could
                have used to support this analysis.
                [[Page 24289]]
                b) Characterizing Vehicles and Their Technology Content
                 The starting point for projecting what additional fuel economy
                improving technologies could feasibly be applied to vehicles is knowing
                what vehicles are produced by which manufacturers and what technologies
                exist on those vehicles. Rows in the market data file are the smallest
                portion of the fleet to which technology may be applied as part of a
                projected compliance pathway. For the analysis presented in this final
                rule, the agencies, when possible, attempted to include vehicle trim
                level information in discrete rows. A manufacturer, for example GM, may
                produce one or more vehicle makes (or brands), for example Chevrolet,
                Buick and others. Each vehicle make may offer one or more vehicle
                models, for example Malibu, Traverse and others. And each vehicle model
                may be available in one or more trim levels (or standard option
                levels), for example ``RS,'' ``Premier'' and others, which have
                different levels of standard options, and in some cases, different
                engines and transmissions.
                 Manufacturer compliance submissions, discussed above, were used as
                a starting point to define working rows in the market data file;
                however, often the rows needed to be further disaggregated to correctly
                characterize vehicle information covered in the scope of the analysis,
                and analysis fleet. Manufacturers often grouped vehicles with multiple
                trim levels together because they often included the same fuel-saving
                technologies and may be aggregated to simplify reporting. However, the
                manufacturer suggested retail prices of different trim levels are
                certainly different, and other features relevant to the analysis are
                occasionally different.
                 As a result of further disaggregating compliance information, the
                number of rows in the market data file increased from 1,667 rows used
                in the NPRM to 2,952 rows for this final rule analysis. The agencies do
                not have data on sales volumes for each nameplate by trim level, and
                used an approach that evenly distributed volume across offered trim
                levels, within the defined constraints of the compliance data.\365\
                Evenly distributing the volume across trim levels is a simplification,
                but this action should (1) highlight some difficulties that could be
                encountered when acquiring data for a full-vehicle consumer choice
                model should the agencies pursue developing one in the future
                (discussed further, below), and (2) lower the average sales volume per
                row in the market data file, thereby allowing the application of very
                advanced electrification technologies in smaller lumps. The latter
                effect is responsive to comments (discussed below) that suggested
                electrification technologies could be more cost-effectively deployed in
                lower volumes, and that the CAFE model artificially constrains cost
                effective technologies that may be deployed, resulting in higher costs
                and large over-compliance.
                ---------------------------------------------------------------------------
                 \365\ The sum of volumes by nameplate configuration, for fuel
                economy value, and for footprint value remains the same.
                ---------------------------------------------------------------------------
                (1) Assigning Vehicle Technology Classes
                 While each vehicle in the analysis fleet has its list of observed
                technologies and equipment, the ways in which manufacturers apply
                technologies and equipment do not always coincide perfectly with how
                the analysis characterizes the various technologies that improve fuel
                economy and reduce CO2 emissions. To improve how the
                observed vehicle fleet ``fits into'' the analysis, each vehicle model/
                configuration is ``mapped'' to the full-vehicle simulation modeling by
                Argonne National Laboratory that is used to estimate the effectiveness
                of the fuel economy-improving/CO2 emissions-reducing
                technologies considered. Argonne produces full-vehicle simulation
                modeling for many combinations of technologies, on many types of
                vehicles, but it did not simulate literally every single manufacturer's
                vehicle model/configuration in the analysis fleet because it would be
                impractical to assemble the requisite detailed information--much of
                which would likely only be provided on a confidential basis--specific
                to each vehicle model/configuration and because the scale of the
                simulation effort would correspondingly increase by at least two orders
                of magnitude. Instead, Argonne simulated 10 different vehicle types
                corresponding to the ``technology classes'' generally used in CAFE
                analysis over the past several rulemakings (e.g., small car, small
                performance car, pickup truck, etc.). Each of those 10 different
                vehicle types was assigned a set of ``baseline characteristics'' to
                which Argonne added combinations of fuel-saving technologies and then
                ran simulations to determine the fuel economy achieved when applying
                each combination of technologies to that vehicle type given its
                baseline characteristics.
                BILLING CODE 4910-59-P
                [[Page 24290]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.089
                BILLING CODE 4910-59-C
                 In the analysis fleet, inputs assign each specific vehicle model/
                configuration to a technology class, and once there, map to the
                simulation within that technology class most closely matching the
                combination of observed technologies and equipment on that vehicle.
                This mapping to a specific simulation result most closely representing
                a given vehicle model/configuration's initial technology ``state''
                enables the CAFE model to estimate the same vehicle model/
                configuration's fuel economy after application of some other
                combination of technologies, leading to an alternative technology
                state.
                (2) Assigning Vehicle Technology Content
                 As explained above, the analysis fleet is defined not only by the
                vehicles it contains, but also by the technologies on those vehicles.
                Each vehicle in the analysis fleet has an associated list of observed
                technologies and equipment that can improve fuel economy and reduce
                CO2 emissions.\366\ With a portfolio of descriptive
                technologies arranged by manufacturer and model, the analysis fleet can
                be summarized and project how vehicles in that fleet may increase fuel
                economy over time via the application of additional technology.
                ---------------------------------------------------------------------------
                 \366\ These technologies are generally grouped into the
                following categories: Vehicle technologies include mass reduction,
                aerodynamic drag reduction, low rolling resistance tires, and
                others. Engine technologies include engine attributes describing
                fuel type, engine aspiration, valvetrain configuration, compression
                ratio, number of cylinders, size of displacement, and others.
                Transmission technologies include different transmission
                arrangements like manual, 6-speed automatic, 10-speed automatic,
                continuously variable transmission, and dual-clutch transmissions.
                Hybrid and electric powertrains may complement traditional engine
                and transmission designs or replace them entirely.
                ---------------------------------------------------------------------------
                 In many cases, vehicle technology is clearly observable from the
                2017 compliance data (e.g., compliance data indicates clearly which
                vehicles have turbochargers and which have continuously variable
                transmissions), but in some cases technology levels are less
                observable. For the latter, like levels of mass reduction, the analysis
                categorized levels of technology already used in a given vehicle.
                Similarly, engineering judgment was used to determine if higher mass
                reduction levels may be used practicably and safely for a given
                vehicle.
                 Either in mid-year compliance data for MY 2016, final compliance
                data for MY 2017, or separately and at the agencies' invitation prior
                to the NPRM or in comments in responses to the NPRM, most manufacturers
                provided guidance on the technology already present in each of their
                vehicle model/configurations. This information was not as complete for
                all manufacturers' products as needed for the analysis, so, in some
                cases, information was supplemented with publicly available data,
                typically from manufacturer media sites. In limited cases,
                manufacturers did not supply information, and
                [[Page 24291]]
                information from commercial and publicly available sources was used.
                 The agencies continued to evaluate emerging technologies in the
                analysis. In response to comments,\367\ and given recent product
                launches for MY 2020, and some very recently announced future product
                offerings, the agencies elevated some technologies that were discussed
                in the NPRM to the compliance simulation. As a result, several
                additional engine technologies, expanded levels of mass reduction
                technology, and some additional combinations of engines with plug-in
                hybrid, or strong hybrid technology are available in the compliance
                pathways for the final rule analysis.
                ---------------------------------------------------------------------------
                 \367\ NHTSA-2018-0067-11741.
                ---------------------------------------------------------------------------
                 In addition, some redundant technologies, or technologies that were
                inadvertently represented on the technology tree as being available to
                be applied twice, have been consolidated. For instance, previous basic
                versions of engine friction reduction were layered on top of basic
                engine maps, but the efficiency in many modern engine maps already
                include the benefits of that engine friction reduction technology. The
                following Table VI-8 lists the technologies considered in the final
                rule analysis, with the data sources used to map those technologies to
                vehicles in the analysis fleet.
                BILLING CODE 4910-59-P
                [[Page 24292]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.090
                [[Page 24293]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.091
                [[Page 24294]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.092
                BILLING CODE 4910-59-C
                 Industry commenters generally stated the MY 2016 baseline
                technology content presented in the NPRM as an improvement over
                previous analyses because it more accurately accounted for technology
                already used in the fleet.368 369 In contrast, some
                commenters expressed preference for EPA's baseline technology
                assignment assumptions presented in the Draft TAR for mass reduction,
                tire rolling resistance, and aerodynamic drag because those assumptions
                projected very few technology improvements were present in the baseline
                fleet. In assessing the comments, the agencies found that
                [[Page 24295]]
                using the EPA Draft TAR approach would lead to projected compliance
                pathways with overestimated fuel economy improvements and
                underestimated costs.\370\
                ---------------------------------------------------------------------------
                 \368\ NHTSA-2018-0067-12073, Alliance of Automobile
                Manufacturers.
                 \369\ NHTSA-2018-0067-12150, Toyota North America.
                 \370\ NHTSA-2018-0067-11741, ICCT.
                ---------------------------------------------------------------------------
                 Many of those assumptions were neither scientifically meritorious,
                nor isolated examples. For instance, for the EPA Draft TAR and Proposed
                Determination analyses, the BMW i3, a vehicle with full carbon fiber
                bodysides and downsized, mass-reduced wheels and tires (some of the
                most advanced mass reducing technologies commercialized in the
                automotive industry), was assumed to have 1.0 percent mass reduction (a
                very minor level of mass reduction). Similarly, previous analyses
                assigned the Chevrolet Corvette, a performance vehicle that has long
                been a platform for commercializing advanced weight saving
                technologies,\371\ with zero mass reduction. For aerodynamic drag,
                previous EPA analysis assumed that pickup trucks could achieve the
                aerodynamic drag profile typical of a sedan, with little regard for
                form drag constraints or frontal area (and headroom, or ground
                clearance) considerations. These assumptions commonly led to
                projections of a 20 percent improvement in mass, aerodynamic drag, and
                tire rolling resistance, even when a large portion of those
                improvements had either already been implemented, or were not
                technologically feasible. On the other hand, in the Draft TAR, NHTSA
                presented methodologies to evaluate content for mass reduction
                technology, aerodynamic drag improvements, and rolling resistance
                technologies that better accounted for the actual level of technologies
                in the analysis fleet. Throughout the rulemaking process, the agencies
                reconciled these differences, jointly presented improved approaches in
                the NPRM similar to what NHTSA presented in the Draft TAR, and again
                used those reconciled approaches in today's analysis.\372\
                ---------------------------------------------------------------------------
                 \371\ See, e.g., Fiberglass to Carbon Fiber: Corvette's
                Lightweight Legacy, GM (August 2012), https://media.gm.com/media/us/en/gm/news.detail.html/content/Pages/news/us/en/2012/Aug/0816_corvette.html.
                 \372\ Because these road load technologies are no longer double
                counted, the projected compliance pathway in the NPRM, and in
                today's analysis for stringent alternatives, often requires more
                advanced fuel saving technologies than previously projected,
                including higher projected penetration rates of hybrid and electric
                vehicle technologies.
                ---------------------------------------------------------------------------
                 Many commenters correctly observed that the analysis fleet in the
                NPRM recognized more technology content in the baseline than in the
                Draft TAR (with higher penetration rates of tire rolling resistance and
                aerodynamic drag improvements, for instance), but also that the fuel
                economy values of the fleet had not improved all that much from the
                previous year. Some commenters concluded that the NPRM baseline
                technology assignment process was arbitrary and overstated the
                technology content already present in the baseline
                fleet.373 374 The agencies agree that there was a large
                increase in the amount of road load technology credited in the baseline
                fleet between EPA's Draft TAR and the jointly produced NPRM, and
                clarify that this change was largely due to a recognition of
                technologies that were actually present in the fleet, but not properly
                accounted for in previous analyses. The change in penetration rates of
                road load technologies (after accounting for glider share updates,
                which is discussed in more detail in the mass reduction technology
                section) between the NPRM and today's analysis is relatively small.
                ---------------------------------------------------------------------------
                 \373\ NHTSA-2018-0067-11741, ICCT.
                 \374\ NHTSA-2018-0067-12039, Union of Concerned Scientists.
                ---------------------------------------------------------------------------
                 Many commenters noted that the different baseline road load
                assumptions (and other technology modeling) materially affect
                compliance pathways, and projected costs.\375\ ICCT commented that the
                agencies should conduct sensitivity analyses assuming every vehicle in
                the analysis fleet is set to zero percent road load technology
                improvement, to demonstrate how the technology content of the analysis
                fleet affected the compliance scenarios.\376\
                ---------------------------------------------------------------------------
                 \375\ NHTSA-2018-0067-11928, Ford Motor Company.
                 \376\ NHTSA-2018-0067-11741, ICCT.
                ---------------------------------------------------------------------------
                 While the agencies have clearly described the methods by which
                initial road load technologies are assigned in Section VI.C.4 Mass
                Reduction, Section VI.C.5 Aerodynamics, and Section VI.C.6 Tire Rolling
                Resistance below, the agencies considered a sensitivity case that
                assumed no mass reduction, rolling resistance, or aerodynamic
                improvements had been made to the MY 2017 fleet (i.e., setting all
                vehicle road levels to zero--MRO, AERO and ROLL0). While this is an
                unrealistic characterization of the initial fleet, the agencies
                conducted a sensitivity analysis to understand any affect it may have
                on technology penetration along other paths (e.g. engine and hybrid
                technology). Under the CAFE program, the sensitivity analysis shows a
                slight decrease in reliance on engine technologies (HCR engines,
                turbocharge engines, and engines utilizing cylinder deactivation) and
                hybridization (strong hybrids and plug-in hybrids) in the baseline
                (relative to the central analysis). The consequence of this shift to
                reliance on lower-level road load technologies is a reduction in
                compliance cost in the baseline of about $300 per vehicle (in MY 2026).
                As a result, cost savings in the preferred alternative are reduced by
                about $200 per vehicle. Under the CO2 program, the general
                trend in technology shift is less dramatic (though the change in BEVs
                is larger) than the CAFE results. The cost change is also comparable,
                but slightly smaller ($200 per vehicle in the baseline) than the CAFE
                program results. Cost savings under the preferred alternative are
                further reduced by about $100. With the lower technology costs in all
                cases, the consumer payback periods decreased as well. These results
                are consistent with the approach taken by manufacturers who have
                already deployed many of the low-level road load reduction
                opportunities to improve fuel economy.
                 Some commenters preferred that the agencies develop a different
                methodology based on reported road load coefficients (``A,'' ``B'' and
                ``C'' coastdown coefficients) to estimate levels of aerodynamic drag
                improvement and rolling resistance in the baseline fleet that did not
                rely on CBI.\377\ The agencies considered this, but determined that
                using CBI to assign baseline aerodynamic drag levels and rolling
                resistance values was more accurate and appropriate. Estimating
                aerodynamic drag levels and rolling resistance levels from coastdown
                coefficients is not straightforward, and to do it well would require
                information the agencies do not have (much of which is also CBI). For
                instance, rotational inertias of wheel, tire, and brake packages can
                affect coastdown, so mass of the vehicle is not sufficient. The frontal
                area of the vehicles, a key component for calculating aerodynamic drag,
                is rarely known, and often requires manufacturer input to get an
                accurate value. Other important vehicle features like all-wheel-drive
                should also be accounted for, and the agencies would struggle to
                correctly identify improvements in rolling resistance, low-drag brakes,
                and secondary axle disconnect, because all of these technologies would
                present similar signature on a coast down test. All of these
                technologies are represented as technology pathways in today's
                analysis. Manufacturers acknowledged the possibility of using road load
                coefficients to estimate rolling resistance and aerodynamic features,
                but warned that the process ``required
                [[Page 24296]]
                various assumptions and is not very accurate,'' and stated that the use
                of CBI to assess aerodynamic and rolling resistance technologies is an
                ``accurate and practical solution'' to assign these difficult to
                observe technologies.\378\
                ---------------------------------------------------------------------------
                 \377\ NHTSA-2018-0067-11741, ICCT.
                 \378\ NHTSA-2018-0067-12073, Alliance of Automobile
                Manufacturers.
                ---------------------------------------------------------------------------
                (3) Assigning Engine Configurations
                 Engine technology costs can vary significantly by the configuration
                of the engine. For instance, adding variable valve lift to each
                cylinder on an engine would cost more for an engine with eight
                cylinders than an engine with four cylinders. Similarly, the cost of
                adding a turbocharger to an engine and downsizing the engine would be
                different going from a naturally aspirated V8 to a turbocharged V6 than
                going from a naturally aspirated V6 to a turbocharged I4. As discussed
                in detail in the engine technology section of this document, the cost
                files for the CAFE model account for instances such as these examples.
                 Information in the analysis fleet enables the CAFE model to
                reference the intended engine costs. The ``Engine Technology Class
                (Observed)'' lists the architecture of the observed engine. Notably,
                the analysis assumes that nearly all turbo charged engines take
                advantage of downsizing to optimize fuel efficiency, minimize the cost
                of turbo charging, and to maintain performance (to the extent
                practicable) with the naturally aspirated counterpart engine.
                Therefore, engines observed in the fleet that have already been down-
                sized must reference costs for a larger basic engine, which assumes
                down-sizing with the application of turbo technology. In these cases,
                the ``Engine Technology Class'' which is used to reference costs will
                be larger than the ``Engine Technology Class (Observed).''
                 This is the same process agencies used in the NPRM, and it corrects
                a previous error in the Draft TAR analysis, which incorrectly
                underestimated turbocharged engine costs.\379\ Some commenters
                expressed confusion and disagreement with this correction, with some
                even commenting that the analysis baselessly inflated costs of
                turbocharging technologies between the Draft TAR and the NPRM.\380\ To
                be clear, this was a correction so that the costs used to calculate
                turbocharged engine costs accurately reflected the total costs for a
                turbocharged engine.
                ---------------------------------------------------------------------------
                 \379\ For instance, the Draft TAR engine costs would map an
                observed V6 Turbo engine to I4 Turbo engine costs, by referencing a
                4C1B engine cost.
                 \380\ NHTSA-2018-0067-11741, ICCT.
                ---------------------------------------------------------------------------
                (4) Characterizing Shared Vehicle Platforms, Engines, and Transmissions
                 Another aspect of characterizing vehicle model/configurations in
                the analysis fleet is based on whether they share a ``platform'' with
                other vehicle model/configurations. A ``platform'' refers to engineered
                underpinnings shared on several differentiated products. Manufacturers
                share and standardize components, systems, tooling, and assembly
                processes within their products (and occasionally with the products of
                another manufacturer) to manage complexity and costs for development,
                manufacturing, and assembly.
                 The concept of platform sharing has evolved over time. Years ago,
                manufacturers rebadged vehicles and offered luxury options only on
                premium nameplates (and manufacturers shared some vehicle platforms in
                limited cases). Today, manufacturers share parts across highly
                differentiated vehicles with different body styles, sizes, and
                capabilities that may share the same platform. For instance, the Honda
                Civic and Honda CR-V share many parts and are built on the same
                platform. Engineers design chassis platforms with the ability to vary
                wheelbase, ride height, and even driveline configuration. Assembly
                lines can produce hatchbacks and sedans to cost-effectively utilize
                manufacturing capacity and respond to shifts in market demand. Engines
                made on the same line may power small cars or mid-size sport utility
                vehicles. In addition, although the agencies' analysis, like past CAFE
                analyses, considers vehicles produced for sale in the U.S., the agency
                notes these platforms are not constrained to vehicle models built for
                sale in the U.S.; many manufacturers have developed, and use, global
                platforms, and the total number of platforms is decreasing across the
                industry. Several automakers (for example, General Motors and Ford)
                either plan to, or already have, reduced their number of platforms to
                less than 10 and account for the overwhelming majority of their
                production volumes on that small number of platforms.
                 Vehicle model/configurations derived from the same platform are so
                identified in the analysis fleet. Many manufacturers' use of vehicle
                platforms is well documented in the public record and widely recognized
                among the vehicle engineering community. Engineering knowledge,
                information from trade publications, and feedback from manufacturers
                and suppliers was also used to assign vehicle platforms in the analysis
                fleet.
                 When the CAFE model is deciding where and how to add technology to
                vehicles, if one vehicle on the platform receives new technology, other
                vehicles on the platform also receive the technology as part of their
                next major redesign or refresh.\381\ Similar to vehicle platforms,
                manufacturers create engines that share parts. For instance,
                manufacturers may use different piston strokes on a common engine
                block, or bore out common engine block castings with different
                diameters to create engines with an array of displacements. Head
                assemblies for different displacement engines may share many components
                and manufacturing processes across the engine family. Manufacturers may
                finish crankshafts with the same tools to similar tolerances. Engines
                on the same architecture may share pistons, connecting rods, and the
                same engine architecture may include both six and eight cylinder
                engines. One engine family may appear on many vehicles on a platform,
                and changes to that engine may or may not carry through to all the
                vehicles. Some engines are shared across a range of different vehicle
                platforms. Vehicle model/configurations in the analysis fleet that
                share engines belonging to the same platform are also identified as
                such.
                ---------------------------------------------------------------------------
                 \381\ The CAFE model assigns mass reduction technology at a
                platform level, but many other technologies may be assigned and
                shared at a vehicle nameplate or vehicle model level.
                ---------------------------------------------------------------------------
                 It is important to note that manufacturers define common engines
                differently. Some manufacturers consider engines as ``common'' if the
                engines shared an architecture, components, or manufacturing processes.
                Other manufacturers take a narrower definition, and only assume
                ``common'' engines if the parts in the engine assembly are the same. In
                some cases, manufacturers designate each engine in each application as
                a unique powertrain. For example, a manufacturer may have listed two
                engines separately for a pair that share designs for the engine block,
                the crank shaft, and the head because the accessory drive components,
                oil pans, and engine calibrations differ between the two. In practice,
                many engines share parts, tooling, and assembly resources, and
                manufacturers often coordinate design updates between two similar
                engines. Engine families, designated in the analysis using ``engine
                codes,'' for each manufacturer were tabulated and assigned based on
                data-driven criteria. If engines shared a common cylinder count and
                configuration, displacement, valvetrain, and fuel type, those engines
                [[Page 24297]]
                may have been considered together. In addition, if the compression
                ratio, horsepower, and displacement of engines were only slightly
                different, those engines were considered the same for the purposes of
                redesign and sharing.
                 Vehicles in the analysis fleet with the same engine family will,
                therefore, adopt engine technology in a coordinated fashion.
                Specifically, if such vehicles have different design schedules (i.e.,
                refresh and redesign schedules), and a subset of vehicles using a given
                engine add engine technologies during of a redesign or refresh that
                occurs in an early model year (e.g., 2018), other vehicles using the
                same engine ``inherit'' these technologies at the soonest ensuing
                refresh or redesign. This is consistent with a view that, over time,
                most manufacturers are likely to find it more practicable to shift
                production to a new version of an engine than to continue production of
                both the new engine and a ``legacy'' engine indefinitely. By grouping
                engines together, the CAFE model controls future engine families to
                ensure reasonable powertrain complexity. This means, however, that for
                manufacturers that submitted highly atomized engine and transmission
                portfolios, there is a practical cap on powertrain complexity and the
                ability of the manufacturer to optimize the displacement of (i.e.,
                ``right size'') engines perfectly for each vehicle configuration. This
                concept is discussed further in Section VI.B.4.a), below.
                 Like with engines, manufacturers often use transmissions that are
                the same or similar on multiple vehicles. Manufacturers may produce
                transmissions that have nominally different machining to castings, or
                manufacturers may produce transmissions that are internally identical,
                except for the final gear ratio. In some cases, manufacturers sub-
                contract with suppliers that deliver whole transmissions. In other
                cases, manufacturers form joint ventures to develop shared
                transmissions, and these transmission platforms may be offered in many
                vehicles across manufacturers. Manufacturers use supplier and joint-
                venture transmissions to a greater extent than they do with engines. To
                reflect this reality, shared transmissions were considered for
                manufacturers as appropriate. Transmission configurations are referred
                to in the analysis as ``transmission codes.'' Like the inheritance
                approach outlined for engines, if one vehicle application of a shared
                transmission family upgraded the transmission, other vehicle
                applications also upgraded the transmission at the next refresh or
                redesign year. To define common transmissions, the agencies considered
                transmission type (manual, automatic, dual-clutch, continuously
                variable), number of gears, and vehicle architecture (front-wheel-
                drive, rear-wheel-drive, all-wheel-drive based on a front-wheel drive
                platform, or all-wheel-drive based on a rear-wheel-drive platform). If
                vehicles shared these attributes, these transmissions were grouped for
                the analysis. Vehicles in the analysis fleet with the same transmission
                configuration will adopt transmission technology together, as described
                above.
                 Having all vehicles that share a platform (or engines that are part
                of a family) adopt fuel economy-improving/CO2 emissions-
                reducing technologies together, subject to refresh/redesign
                constraints, reflects the real-world considerations described above,
                but also overlooks some decisions manufacturers might make in the real
                world in response to market pull. Accordingly, even though the analysis
                fleet is incredibly complex, it is also over-simplified in some
                respects compared to the real world. For example, the CAFE model does
                not currently attempt to simulate the potential for a manufacturer to
                shift the application of technologies to improve performance rather
                than fuel economy. Therefore, the model's representation of the
                ``inheritance'' of technology can lead to estimates a manufacturer
                might eventually exceed fuel economy standards as technology continues
                to propagate across shared platforms and engines. While the agencies
                have previously seen examples of extended periods during which some
                manufacturers exceeded one or both CAFE and/or CO2
                standards, in plenty of other examples, manufacturers chose to
                introduce (or even reintroduce) technological complexity into their
                vehicle lineups in response to buyer preferences. Going forward, and
                recognizing the recent trend for consolidating platforms, it seems
                likely manufacturers will be more likely to choose efficiency over
                complexity in this regard; therefore, the potential should be lower
                that today's analysis turns out to be oversimplified compared to the
                real world.
                 Manufacturers described shared engines, transmissions, and vehicle
                platforms as ``standard business practice'' and they were encouraged
                that the NHTSA analysis in the Draft TAR, and the jointly issued NPRM
                placed realistic limits on the number of unique engines and
                transmissions in a powertrain portfolio.\382\ In previous rulemakings,
                stakeholders pointed out that shared parts and portfolio complexity
                should be considered (but were not), and that the proliferation of
                unique technology combinations resulting from unconstrained compliance
                pathways would jeopardize economies of scale in the real world.\383\
                ---------------------------------------------------------------------------
                 \382\ NHTSA-2018-0067-12150, Toyota North America.
                 \383\ Alliance of Automobile Manufacturers, EPA-HQ-OAR-0827 and
                NHTSA-2016-0068.
                ---------------------------------------------------------------------------
                 HD Systems acknowledged that previous rulemakings did not
                appropriately consider part sharing, but contended that in today's
                global marketplace, manufacturers have flexibility to compete in new
                ways that break old part sharing rules.\384\ The agencies acknowledge
                that some transmissions are now sourced through suppliers, and that
                economies of scale could, in the future be achieved at an industry
                level instead of a manufacturer level; however, even when manufacturers
                outsource a transmission, recent history suggests they apply that
                transmission to multiple vehicles to control assembly plant and service
                parts complexity, as they would if they were making the transmission
                themselves. Similarly, even for global platforms, or global
                powertrains, there is little evidence that manufacturers fragment
                powertrain line-ups for a vehicle, or a set of vehicles that have
                typically used the same engine. The agencies will continue to consider
                how to capture more accurately the ways vehicles share engines,
                transmissions, and platforms in future rulemakings, but the part-
                sharing and modeling approach presented in the NPRM and this final rule
                represents a marked improvement over previous analysis.
                ---------------------------------------------------------------------------
                 \384\ NHTSA-2018-0067-11985, HD Systems.
                ---------------------------------------------------------------------------
                (5) Characterizing Production Design Cycles
                 Another aspect of characterizing vehicles in the analysis fleet is
                based on when they can next be refreshed or redesigned. Redesign
                schedules play an important role in determining when new technologies
                may be applied. Many technologies that improve fuel economy and reduce
                CO2 emissions may be difficult to incorporate without a
                major product redesign. Therefore, each vehicle model in the analysis
                fleet has an associated redesign schedule, and the CAFE model uses that
                schedule to implement significant advances in some technologies (like
                major mass reduction) to redesign years, while allowing manufacturers
                to include minor advances (such as improved tire rolling
                [[Page 24298]]
                resistance) during a vehicle ``refresh,'' or a smaller update made to a
                vehicle, which can happen between redesigns. In addition to refresh and
                redesign schedules associated with vehicle model/configurations,
                vehicles that share a platform subsequently have platform-wide refresh
                and redesign schedules for mass reduction technologies.
                 To develop the refresh/redesign cycles used for the NPRM vehicles
                in the analysis fleet, information from commercially available sources
                was used to project redesign cycles through MY 2022, as was done for
                NHTSA's analysis for the 2016 Draft TAR.\385\ Commercially available
                sources' estimates through MY 2022 are generally supported by detailed
                consideration of public announcements plus related intelligence from
                suppliers and other sources, and recognize that uncertainty increases
                considerably as the forecasting horizon is extended. For MYs 2023-2035,
                in recognition of that uncertainty, redesign schedules were extended
                considering past pacing for each product, estimated schedules through
                MY 2022, and schedules for other products in the same technology
                classes. As mentioned above, potentially confidential forward-looking
                information was not requested from manufacturers; nevertheless, all
                manufacturers had an opportunity to review the estimates of product-
                specific redesign schedules. A few manufacturers provided related
                forecasts and, for the most part, that information corroborated the
                estimates.
                ---------------------------------------------------------------------------
                 \385\ In some cases, data from commercially available sources
                was found to be incomplete or inconsistent with other available
                information. For instance, commercially available sources identified
                some newly imported vehicles as new platforms, but the international
                platform was midway through the product lifecycle. While new to the
                U.S. market, treating these vehicles as new entrants would have
                resulted in artificially short redesign cycles if carried forward,
                in some cases. Similarly, commercially available sources labeled
                some product refreshes as redesigns, and vice versa. In these
                limited cases, the data was revised to be consistent with other
                available information or typical redesign and refresh schedules for
                CAFE modeling. In these limited cases, the forecast time between
                redesigns and refreshes was updated to match the observed past
                product timing.
                ---------------------------------------------------------------------------
                 Some commenters suggested supplanting these estimated redesign
                schedules with estimates applying faster cycles (e.g., four to five
                years), and this approach was considered for the analysis. Some
                manufacturers tend to operate with faster redesign cycles and may
                continue to do so, and manufacturers tend to redesign some products
                more frequently than others. However, especially considering that
                information presented by manufacturers largely supports estimates
                discussed above, applying a ``one size fits all'' acceleration of
                redesign cycles would not improve the analysis; instead, assuming a
                fixed, shortened redesign schedule across the industry would likely
                reduce consistency with the real world, especially for light trucks,
                which are redesigned, on average, no less than every six years (see
                Table VI-9, below). Moreover, if some manufacturers accelerate
                redesigns in response to new standards, doing so would likely involve
                costs (greater levels of stranded capital, reduced opportunity to
                benefit from ``learning''-related cost reductions) greater than
                reflected in other inputs to the analysis.
                 As discussed in the NPRM, manufacturers use diverse strategies with
                respect to when, and how often they update vehicle designs. While most
                vehicles have been redesigned sometime in the last five years, many
                vehicles have not. In particular, vehicles with lower annual sales
                volumes tend to be redesigned less frequently, perhaps giving
                manufacturers more time to recoup the investment needed to bring the
                product to market. In some cases, manufacturers continue to produce and
                sell vehicles designed more than a decade ago.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.093
                [[Page 24299]]
                 Each manufacturer may use different strategies throughout their
                product portfolio, and a component of each strategy may include the
                timing of refresh and redesign cycles. Table VI-10 summarizes the
                average time between redesigns, by manufacturer, by vehicle technology
                class. Dashes mean the manufacturer has no volume in that vehicle
                technology class in the MY 2017 analysis fleet. Across the industry,
                manufacturers average 6.6 years between product redesigns.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.094
                 Trends on redesign schedules identified in the NPRM remain in place
                for today's analysis. Pick-up trucks have much longer redesign
                schedules than small cars. Some manufacturers redesign vehicles often,
                while other manufacturers redesign vehicles less often. Even if two
                manufacturers have similar redesign cadence, the model years in which
                the redesigns occur may still be different and dependent on where each
                of the manufacturer's products are in their life cycle.
                 Table VI-11 summarizes the average age of manufacturers' offering
                by vehicle technology class. A value of ``0.0'' means that every
                vehicle for a manufacturer in the vehicle technology class, represented
                by the MY 2017 analysis fleet was new in MY 2017. Across the industry
                manufacturers redesigned MY 2017 vehicles an average of 3.5 years
                earlier, meaning the average MY 2017 vehicle was last redesigned in
                approximately MY 2013, also on average near a midpoint in their product
                lifecycle.
                [[Page 24300]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.095
                BILLING CODE 4910-59-C
                 Some commenters cited examples of vehicles in the NPRM analysis
                fleet where the redesign years were off by a year here or there in the
                2017-2022 timeframe relative to the most recent public announcements,
                or that the extended forecasts were too rigid.\386\ The CAFE model
                structurally requires an input for the redesign years, and the agencies
                worked to make these generally representative without disclosing
                precise CBI product plans. Many of the redesign schedules were carried
                over from the NPRM, with a few minor updates.
                ---------------------------------------------------------------------------
                 \386\ NHTSA-2018-0067-11723, Natural Resources Defense Council.
                ---------------------------------------------------------------------------
                 Some commenters contended that the agencies should not look at the
                historical data to project the timing between redesigns (``business as
                usual''), but should instead adopt a ``policy case'' with an
                accelerated pace of redesigns and refreshes.\387\ Some commenters
                suggested that the agencies use a standard 5 or 6 year redesign
                schedule for all manufacturers and all products as a way to lower
                projected costs.\388\ Other stakeholders commented that the entire
                industry should be modeled with the ability to redesign everything at
                one time in the near term because that would not presuppose precisely
                how manufacturers may adjust their fleet.\389\
                ---------------------------------------------------------------------------
                 \387\ NHTSA-2018-0067-11723, Natural Resources Defense Council.
                 \388\ NHTSA-2018-0067-11985, HD Systems.
                 \389\ NHTSA-2018-0067-12039, Union of Concerned Scientists.
                ---------------------------------------------------------------------------
                 If the agencies were to implement any such approaches, the agencies
                would need to more precisely account for tooling costs, research and
                development costs, and product lifecycle marketing costs, or risk
                missing ``hidden costs'' of a shortened cadence. To account properly
                for these, the CAFE model would require major changes, and would
                require specific inputs that are currently covered generically under
                the retail price equivalency (RPE) factor.\390\ The agencies considered
                these comments, and decided the process for refresh and redesign
                outlined in the NPRM was a reasonable and realistic approach to
                characterize product changes. The agencies conducted sensitivity
                analysis with compressed redesign and refresh schedules, though these
                ignore the resulting compressed amortization schedules, missing
                important costs that are incorporated in the current RPE assumptions.
                ---------------------------------------------------------------------------
                 \390\ Shorter redesign schedules are likely to put upward
                pressure on RPE, as the manufacturers would have less time to recoup
                investments.
                ---------------------------------------------------------------------------
                 Some commenters claimed that the agency had extraordinarily
                extended redesign schedule of 17.7 years for FCA between 2021-2025, and
                an average redesign time of 25.8 years for Ford between 2022-2025.\391\
                The agencies found these claims inaccurate and without basis. Table VI-
                10, ``Summary of Sales Weighted Average Time
                [[Page 24301]]
                between Engineering Redesigns, by Manufacturer, by Vehicle Technology
                Class'' summarizes the data used in today's analysis (which is very
                similar to the information used in the NPRM, with some minor
                adjustments and updates to the fleet), and the detailed information
                vehicle-by-vehicle is reported in the ``market data'' file. The
                agencies recognize that the natural sequence of redesigns for some
                manufacturers and some products is not ideal to meet stringent
                alternatives, which is part of the consideration for economic
                practicability and technological feasibility. Manufacturers commented
                supportively on the idea of vehicle specific redesign schedules, and
                the redesign cadence used in the NPRM, as these contribute to realistic
                assessments of new technology penetration within the fleet, and
                acknowledge the heterogeneity in the product development approaches and
                business practices for each manufacturer.\392\ One commenter recognized
                that redesign and refresh schedules represented a vast improvement over
                phase-in caps to model the adoption of mature technologies.\393\
                ---------------------------------------------------------------------------
                 \391\ NHTSA-2018-0067-11723, Natural Resources Defense Council.
                 \392\ NHTSA-2018-0067-11928, Ford Motor Company.
                 \393\ NHTSA-2018-0067-0444, Walter Kreucher.
                ---------------------------------------------------------------------------
                 Other commenters argued that the structural construct of
                technologies only being available at redesign or at refresh (via
                inheritance) did not reflect real world actions and was not supported
                by any actual data.\394\ Other commenters acknowledged the inheritance
                of engine and transmission technologies at refresh as an important,
                positive feature of the CAFE model.\395\ HD Systems argued that an
                engine or transmission package available in other markets on a global
                platform could be imported to the U.S. market during refresh, and did
                not require a ``leader'' at redesign in the U.S. market to seed
                adoption. HDS cited a few examples where manufacturers have introduced
                strong hybrid powertrains on an existing vehicle a year or two after
                the product launch, not associated with any particular vehicle redesign
                or refresh.
                ---------------------------------------------------------------------------
                 \394\ NHTSA-2018-0067-11985, HD Systems.
                 \395\ NHTSA-2018-0067-11723, Natural Resources Defense Council.
                ---------------------------------------------------------------------------
                 The agencies carefully considered these comments, and observed that
                some relatively low volume hybrid options may appear after launch, or
                that some transmissions were quickly replaced shortly after a major
                redesign. In many of these cases, launch delays, warranty claims, or
                other external factors contributed to, at least in part, an atypically
                timed introduction of fuel saving technology to the fleet.\396\ At this
                point, this does not appear to be a mainstream, or preferred industry
                practice. However, the agencies will continue to evaluate this. For
                future rulemaking, the agencies may consider engine refresh and
                redesign cycles for engines and transmissions. These may be separate
                from vehicle redesign and refresh schedules because the powertrain
                product lifecycles may be longer on average than the typical vehicle
                redesign schedules. This approach, if researched and implemented in
                future analysis, could provide some opportunity for manufacturers to
                introduce new powertrain technologies independent of the vehicle
                redesign schedules, in addition to inheriting advanced powertrain
                technology as refresh as already modeled in the NPRM and today's
                analysis.
                ---------------------------------------------------------------------------
                 \396\ Such instances are observable in detailed CAFE and
                CO2 compliance data submitted to EPA and NHTSA.
                ---------------------------------------------------------------------------
                 For today's analysis, the agencies, with a few exceptions based on
                updated publicly available information, carried over redesign cadences
                for each vehicle nameplate as presented in the NPRM. The agencies do
                not claim that the projected redesign years will perfectly match what
                industry does--notably because refresh and redesign information is CBI
                and the agencies have applied more generalized schedules to protect the
                CBI. Also, what any individual manufacturer may choose to do today
                could be completely different than what it chooses to do tomorrow due
                to changing business circumstances and plans--but the agencies have
                worked to ensure the timing of redesigns will be roughly correct
                (especially in the near term), and that the time between redesigns will
                continue forward for each manufacturer as it has based on recent
                history. The agencies have also increased the frequency of refreshes in
                response to comments about the proliferation of some engine and
                transmission families through manufacturers' product portfolios.
                 Also for today's analysis, the agencies now explicitly model CAFE
                compliance pathways out through 2050. For the model to work as
                intended, the agencies must project refresh and redesign schedules out
                through 2050. The agencies recognize that the accuracy of predictions
                about the distant future, particularly about refresh and redesign
                cycles through the 2030-2050 timeframe, are likely to be poor. If
                historical evolution of the industry continues, many of the nameplates
                carried forward in the fleet are likely to be out of production, and
                new nameplates not considered in the analysis are sure to emerge.
                Still, carrying forward the MY 2017 fleet with the current refresh and
                redesign cadences is consistent with the current analysis, and imposing
                an alternative schedule on the fleet, or making up new nameplates and
                retiring older nameplates without a clear basis, would lack proper
                foundation.
                BILLING CODE 4910-59-P
                [[Page 24302]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.096
                BILLING CODE 4910-59-C
                (6) Defining Technology Adoption Features
                 In some circumstances, the agencies may reference full vehicle
                simulation effectiveness data for technology combinations that are not
                able to be, or are not likely to be applied to all vehicles. In some
                cases, a specific technology as modeled only exists on paper, and
                questions remain about the technological feasibility of the efficiency
                characterization.\397\ Or, a technology may perform admirably on the
                test cycle, but fail to meet all functional, or performance
                requirements for certain vehicles.\398\ In other cases, the
                intellectual property landscape may make commercialization of one
                technology risky for a manufacturer without the consent of the
                intellectual property owner.\399\ In such cases, the agencies may not
                allow a technology to be applied to a certain vehicle. The agencies
                designate this in the ``market data'' file with a ``SKIP'' for the
                technology and vehicle. The logic is explained technology by technology
                in this document, as the logic was explained in the PRIA for this rule.
                ---------------------------------------------------------------------------
                 \397\ High levels of aerodynamic drag reduction for some body
                styles, or EPA's previous, speculative characterization of ``HCR2''
                engines, for example.
                 \398\ Examples of applications that are unsuitable for certain
                technologies include low end torque requirements for HCR engines on
                high load vehicles, or towing and trailering applications,
                continuously variable transmissions in high torque applications, and
                low rolling resistance tires on vehicles built for precision
                cornering and high lateral forces, or instant acceleration from a
                stand still.
                 \399\ Variable compression ratio engines, for example.
                ---------------------------------------------------------------------------
                 Some commenters argued that the restrictions of technologies on a
                case-by-case basis required case-by-case explanation (and not objective
                specification defined cut-offs), and that the use of CBI for
                performance considerations was unacceptable unless fully
                disclosed.\400\ As discussed above, the agencies are not able to
                disclose CBI. Stakeholders have had plenty of opportunities to comment
                on the applicability of technologies, including the few that have used
                SKIP logic restrictions for a portion of the fleet.
                ---------------------------------------------------------------------------
                 \400\ NHTSA-2018-0067-11741, ICCT.
                ---------------------------------------------------------------------------
                 Other commenters suggested an optimistic and wholly unfounded
                approach to manufacturer innovation, arguing that costs would continue
                to come down (beyond what is currently modeled with cost learning), and
                the list of fuel-saving technologies would continually regenerate
                itself (even if the technological mechanism for fuel saving
                technologies was not yet identified).\401\ Therefore, the argument goes
                that people will figure out new ways to improve fuel saving
                technologies to increase their applicability, and the current
                technology characterization should be enabled for selection with no
                restriction--not because the commenter knows how the technology will be
                adapted, but that the commenter believes the technology could,
                eventually, within the timeline of the rulemaking, be adapted, brought
                to market, and be accepted by consumers. While the agencies recognize
                the improvements that many manufacturers
                [[Page 24303]]
                have achieved in fuel saving technologies, some of which were difficult
                to foresee, the agencies have an obligation under the law to be
                judicious and specific about technological feasibility, and to avoid
                speculative conclusions about technologies to justify the rulemaking.
                ---------------------------------------------------------------------------
                 \401\ NHTSA-208-0067-12122-33, American Council for an Energy-
                Efficient Economy.
                ---------------------------------------------------------------------------
                c) Other Analysis Fleet Data
                (1) Safety Classes
                 The agencies referenced the mass-size-safety analysis to project
                the effects changes in weight may have on crash fatalities. That
                analysis, discussed in more detail in Section VI.D.2, considers how
                weight changes may affect safety for cars, crossover utility vehicles
                and sport utility vehicles, and pick-up trucks. To consider these
                effects, the agencies mapped each vehicle in the analysis fleet to the
                appropriate ``Safety Class.''
                (2) Labor Utilization
                 The analysis fleet summarizes components of direct labor for each
                vehicle considered in the analysis. The labor is split into three
                components: (1) Dealership hours worked on sales functions per vehicle,
                (2) direct assembly labor for final assembly, engine, and transmission,
                and (3) percent U.S. content.
                 In the MY 2016 fleet for the NPRM, the agencies catalogued
                production locations and plant employment, reviewed annual reports from
                the North American Dealership Association to estimate dealership
                employment (27.8 hours per vehicle sold), and estimated the industry
                average labor hours for final assembly of vehicles (30 hours per
                vehicle produced), engine machining and assembly (4 hours per engine
                produced), and transmission production (5 hours per transmission
                produced).
                 Today's analysis fleet carries over the estimated labor
                coefficients for sales and production, but references the most recent
                Part 583 American Automobile Labeling Act Report for percent U.S.
                content and for the location of vehicle assembly, engine assembly, and
                transmission assembly.\402\
                ---------------------------------------------------------------------------
                 \402\ Part 583 American Automobile Labeling Act Report,
                available at https://www.nhtsa.gov/part-583-american-automobile-labeling-act-reports.
                ---------------------------------------------------------------------------
                (3) Production Volumes for Sales Analysis
                 A final important aspect of projecting what vehicles will exist in
                future model years and potential manufacturer responses to standards is
                estimating how future sales might change in response to different
                potential standards. If potential future standards appear likely to
                have major effects in terms of shifting production from cars to trucks
                (or vice versa), or in terms of shifting sales between manufacturers or
                groups of manufacturers, that is important for the agencies to
                consider. For previous analyses, the CAFE model used a static forecast
                contained in the analysis fleet input file, which specified changes in
                production volumes over time for each vehicle model/configuration. This
                approach yielded results that, in terms of production volumes, did not
                change between scenarios or with changes in important model inputs. For
                example, very stringent standards with very high technology costs would
                result in the same estimated production volumes as less stringent
                standards with very low technology costs. For this analysis, as in the
                proposal, the CAFE model begins with the first-year production volumes
                (i.e., MY 2017 for today's analysis) and adjusts ensuing sales mix year
                by year (between cars and trucks, and between manufacturers)
                endogenously as part of the analysis, rather than using external
                forecasts of future car/truck split and future manufacturer sales
                volumes. This leads the model to produce different estimates of future
                production volumes under different standards and in response to
                different inputs, reflecting the expectation that regulatory standards
                and other external factors will, in fact, impact the market.
                (4) Comments on Other Analysis Fleet Data
                 Some commenters suggest that the CAFE model should run as a full
                consumer choice model (and this idea is discussed in more detail in
                Section VI.D.1). While this sounds like a reasonable request on the
                surface, such an approach would place enormous new demands on the data
                characterized in the fleet (and preceding fleets, which may be needed
                to calibrate a model properly). For instance, some model concepts may
                depend on a bevy of product features, such as interior cargo room,
                artistic appeal of the design, and perceived quality of the vehicle.
                But product features alone may not be sufficient. Additional
                information about dealership channels, product awareness and
                advertising effectiveness, and financing terms also may be required.
                Such information could dramatically increase the scope of work needed
                to characterize the analysis fleet for future rulemakings. As described
                in Section VI.D.1.b)(2)(d) Using Vehicle Choice Models in Rulemaking
                Analysis. Accordingly, the agencies decided not to develop such a model
                for this rulemaking.
                2. Treatment of Compliance Credit Provisions
                 Today's final rule involves a variety of provisions regarding
                ``credits'' and other compliance flexibilities. Some recently
                introduced regulatory provisions allow a manufacturer to earn
                ``credits'' that will be counted toward a vehicle's rated
                CO2 emissions level, or toward a fleet's rated average
                CO2 or CAFE level, without reference to required levels for
                these average levels of performance. Such flexibilities effectively
                modify emissions and fuel economy test procedures, or methods for
                calculating fleets' CAFE and average CO2 levels. Such
                provisions are discussed below in Section VI.B.2. Other provisions (for
                CAFE, statutory provisions) allow manufacturers to earn credits by
                achieving CAFE or average CO2 levels beyond required levels;
                these provisions may hence more appropriately be termed ``compliance
                credits.''
                 EPCA has long provided that, by exceeding the CAFE standard
                applicable to a given fleet in a given model year, a manufacturer may
                earn corresponding ``credits'' that the same manufacturer may, within
                the same regulatory class, apply toward compliance in a different model
                year. EISA amended these provisions by providing that manufacturers
                may, subject to specific statutory limitations, transfer compliance
                credits between regulatory classes, and trade compliance credits with
                other manufacturers. The CAA provides EPA with broad standard-setting
                authority for the CO2 program, with no specific directives
                regarding either CO2 standards or CO2 compliance
                credits.
                 EPCA also specifies that NHTSA may not consider the availability of
                CAFE credits (for transfer, trade, or direct application) toward
                compliance with new standards when establishing the standards
                themselves.\403\ Therefore, this analysis, like that presented in the
                NPRM, considers 2020 to be the last model year in which carried-forward
                or transferred credits can be applied for the CAFE program. Beginning
                in model year 2021, today's ``standard setting'' analysis for NHTSA's
                program is conducted assuming each fleet must comply with the CAFE
                standard separately in every model year.
                ---------------------------------------------------------------------------
                 \403\ 49 U.S.C. 32902(h)(3).
                ---------------------------------------------------------------------------
                 The ``unconstrained'' perspective acknowledges that these
                flexibilities exist as part of the program, and, while not considered
                by NHTSA in setting standards, are nevertheless important to consider
                when attempting to estimate the real impact of any alternative. Under
                [[Page 24304]]
                the ``unconstrained'' perspective, credits may be earned, transferred,
                and applied to deficits in the CAFE program throughout the full range
                of model years in the analysis. The Final Environmental Impact Analysis
                (FEIS) accompanying today's final rule, like the corresponding Draft
                EIS analysis, presents results of ``unconstrained'' modeling. Also,
                because the CAA provides no direction regarding consideration of any
                CO2 credit provisions, today's analysis, like the NPRM
                analysis, includes simulation of carried-forward and transferred
                CO2 credits in all model years.
                 Some commenters took issue broadly with this treatment of
                compliance credits. Michalek and Whitefoot wrote that ``we find this
                requirement problematic because the automakers use these flexibilities
                as a common means of complying with the regulation, and ignoring them
                will bias the cost-benefit analysis to overestimate costs.'' \404\
                ---------------------------------------------------------------------------
                 \404\ Michalek, J. and Whitefoot, K., NHTSA-2018-0067-11903, at
                10-11.
                ---------------------------------------------------------------------------
                 Counter to the above general claim, the CAFE model does provide
                means to simulate manufacturers' potential application of some
                compliance credits, and both the analysis of CO2 standards
                and the NEPA analysis of CAFE standards do make use of this aspect of
                the model. As discussed above, NHTSA does not have the discretion to
                consider the credit program--in fact, the agency is prohibited by
                statute from doing so--in establishing maximum feasible standards.
                Further, as discussed below, the agencies also continue to find it
                appropriate for the analysis largely to refrain from simulating two of
                the mechanisms allowing the use of compliance credits.
                 The model's approach to simulating compliance decisions accounts
                for the potential to earn and use CAFE credits as provided by EPCA/
                EISA. The model similarly accumulates and applies CO2
                credits when simulating compliance with EPA's standards. Like past
                versions, the current CAFE model can be used to simulate credit carry-
                forward (a.k.a. banking) between model years and transfers between the
                passenger car and light truck fleets but not credit carry-back (a.k.a.
                borrowing) from future model years or trading between manufacturers.
                 Regarding the potential to carry back compliance credits, UCS
                commented that, although past versions of the CAFE model had
                ``considered this flexibility in its approach to multiyear modeling,''
                NHTSA had, without explanation, ``abruptly discontinued support of this
                method of compliance,'' such that ``manufacturers are generally
                incentivized to over comply, regardless of whether carrying forward a
                deficit to be compensated by later overcompliance would be a more cost-
                effective method of compliance.'' \405\ Citing the potential that
                manufacturers could make use of carried back credits in the future, UCS
                also stated that ``NHTSA's decision to constrain it in the model is
                unreasonable and arbitrary.'' \406\ UCS effectively implies that the
                agencies should base standards on analysis that presumes manufacturers
                will take full theoretical advantage of provisions allowing credits to
                be borrowed.
                ---------------------------------------------------------------------------
                 \405\ UCS, NHTSA-2018-0067-12039, Technical Appendix, at 44.
                 \406\ UCS, op. cit., at 77.
                ---------------------------------------------------------------------------
                 The agencies have carefully considered these comments, and while
                EPA's decisions regarding CO2 standards can consider the
                potential to carry back compliance credits from later to earlier model
                years, and NHTSA's ``unconstrained'' evaluation could also do so, past
                examples of failed attempts to carry back CAFE credits (e.g., a MY2014
                carry back default leading to a civil penalty payment) underscore the
                riskiness of such ``borrowing.'' Recent evidence indicates
                manufacturers are disinclined to take such risks,\407\ and both
                agencies find it reasonable and prudent to refrain from attempting to
                simulate such ``borrowing'' in rulemaking analysis.
                ---------------------------------------------------------------------------
                 \407\ Section IX, below, reviews data regarding manufacturers'
                use of CAFE compliance credit mechanism during MYs 2011-2016, and
                shows that the use of ``carry back'' credits is, relative to the use
                of other compliance credit mechanisms, too small to discern.
                ---------------------------------------------------------------------------
                 Unlike past versions, the NPRM and current versions of CAFE model
                provide a basis to specify (in model inputs) CAFE credits available
                from model years earlier than those being explicitly simulated. For
                example, with this analysis representing model years 2017-2050
                explicitly, credits earned in model year 2012 are made available for
                use through model year 2017 (given the current five-year limit on
                carry-forward of credits). The banked credits are specific to both the
                model year and fleet in which they were earned.
                 In addition to the above-mentioned comments, UCS also cited as
                ``errors'' that ``the model does not accurately reflect the one-time
                exemption from the EPA 5-year credit life for credits earned in the MY
                2010-2015 timeframe'' and ``NHTSA assumes that there will be absolutely
                no credit trading between manufacturers.''
                 As discussed below, in the course of updating the analysis fleet
                from MY 2016 to MY 2017, the agencies have updated and expanded the
                manner in which the model accounts for credits earned prior to MY 2017,
                including credits earned as early as MY 2009. In order to increase the
                realism with which the model transitions between the early model year
                (MYs 2017-2020) and the later years that are the subject of this
                action, the agencies have accounted for the potential that some
                manufacturers might trade some of these pre-MY 2017 credits to other
                manufacturers. However, as with the NPRM, the analysis refrains from
                simulating the potential that manufacturers might continue to trade
                credits during and beyond the model years covered by today's action.
                The agencies remain concerned that any realistic simulation of such
                trading would require assumptions regarding which specific pairs of
                manufacturers might actually trade compliance credits, and the evidence
                to date makes it clear that the credit market is far from fully
                ``open.'' With respect to the FCA comment cited above, the agencies
                also remain concerned that to set standards based on an analysis that
                presumes the use of program flexibilities risks making the
                corresponding actions mandatory. Some flexibilities--credit carry-
                forward (banking) and transfers between fleets in particular--involve
                little risk, because they are internal to a manufacturer and known in
                advance. As discussed above, credit carry-back involves significant
                risk, because it amounts to borrowing against future improvements,
                standards, and production volume and mix--and anticipated market demand
                for fuel efficient vehicles often fail to materialize. Similarly,
                credit trading also involves significant risk, because the ability of
                manufacturer A to acquire credits from manufacturer B depends not just
                on manufacturer B actually earning the expected amount of credit, but
                also on manufacturer B being willing to trade with manufacturer A, and
                on potential interest by other manufacturers. Manufacturers' compliance
                plans have already evidenced cases of compliance credit trades that
                were planned and subsequently aborted, reinforcing the agencies'
                judgment that, like credit banking, credit trading involves too much
                risk to be included in an analysis that informs decisions about the
                stringency of future standards. Nevertheless, recognizing that some
                manufacturers have actually been trading credits, the agencies have, as
                in the NPRM, included in the sensitivity analysis a case that simulates
                ``perfect'' trading of compliance credits, focusing
                [[Page 24305]]
                on CO2 standards to illustrate the hypothetical maximum
                potential impact of trading. The FRIA summarizes results of this and
                other cases included in the sensitivity analysis.
                 As discussed in the CAFE model documentation, the model's default
                logic attempts to maximize credit carry-forward--that is, to ``hold
                on'' to credits for as long as possible. If a manufacturer needs to
                cover a shortfall that occurs when insufficient opportunities exist to
                add technology in order to achieve compliance with a standard, the
                model will apply credits. Otherwise the manufacturer carries forward
                credits until they are about to expire, at which point it will use them
                before adding technology that is not considered cost-effective. The
                model attempts to use credits that will expire within the next three
                years as a means to smooth out technology application over time to
                avoid both compliance shortfalls and high levels of over-compliance
                that can result in a surplus of credits. Although it remains impossible
                precisely to predict manufacturer's actual earning and use of
                compliance credits, and this aspect of the model may benefit from
                future refinement as manufacturers and regulators continue to gain
                experience with these provisions, this approach is generally consistent
                with manufacturers' observed practices.
                 NHTSA introduced the CAFE Public Information Center to provide
                public access to a range of information regarding the CAFE
                program,\408\ including manufacturers' credit balances. However, there
                is a data lag in the information presented on the CAFE PIC that may not
                capture credit actions across the industry for as much as several
                months. Furthermore, CAFE credits that are traded between manufacturers
                are adjusted to preserve the gallons saved that each credit
                represents.\409\ The adjustment occurs at the time of application
                rather than at the time the credits are traded. This means that a
                manufacturer who has acquired credits through trade, but has not yet
                applied them, may show a credit balance that is either considerably
                higher or lower than the real value of the credits when they are
                applied. For example, a manufacturer that buys 40 million credits from
                Tesla may show a credit balance in excess of 40 million. However, when
                those credits are applied, they may be worth only 1/10 as much--making
                that manufacturer's true credit balance closer to 4 million than 40
                million.
                ---------------------------------------------------------------------------
                 \408\ CAFE Public Information Center, http://www.nhtsa.gov/CAFE_PIC/CAFE_PIC_Home.htm (last visited June 22, 2018).
                 \409\ CO2 credits for EPA's program are denominated
                in metric tons of CO2 rather than gram/mile compliance
                credits and require no adjustment when traded between manufacturers
                or fleets.
                ---------------------------------------------------------------------------
                 For the NPRM, the agencies reviewed then-recent credit balances,
                estimated the potential that some manufacturers could trade credits,
                and developed inputs that make carried-forward credits available in
                each of model years 2011-2015, after subtracting credits assumed to be
                traded to other manufacturers, adding credits assumed to be acquired
                from other manufacturers through such trades, and adjusting any traded
                credits (up or down) to reflect their true value for the fleet and
                model year into which they were traded.\410\ For today's analysis, an
                additional model year's data was available in mid-2019, and the
                agencies updated these inputs, as summarized in Table VI-12, Table VI-
                13, and Table VI-14. While the CAFE model will transfer expiring
                credits into another fleet (e.g., moving expiring credits from the
                domestic car credit bank into the light truck fleet), some of these
                credits were moved into the initial banks to improve the efficiency of
                application and both to reflect better the projected shortfalls of each
                manufacturer's regulated fleets and to represent observed behavior. For
                context, a manufacturer that produces one million vehicles in a given
                fleet, and experiences a shortfall of 2 mpg, would need 20 million
                credits, adjusted for fuel savings, to offset the shortfall completely.
                ---------------------------------------------------------------------------
                 \410\ The adjustments, which are based upon the CAFE standard
                and model year of both the party originally earning the credits and
                the party applying them, were implemented assuming the credits would
                be applied to the model year in which they were set to expire. For
                example, credits traded into a domestic passenger car fleet for MY
                2014 were adjusted assuming they would be applied in the domestic
                passenger car fleet for MY 2019.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24306]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.097
                [GRAPHIC] [TIFF OMITTED] TR30AP20.098
                [[Page 24307]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.099
                BILLING CODE 4910-59-C
                 In addition to the inclusion of these existing credit banks, the
                CAFE model also updated its treatment of credits in the rulemaking
                analysis. EPCA requires that NHTSA set CAFE standards at maximum
                feasible levels for each model year without consideration of the
                program's credit mechanisms. However, as recent NHTSA CAFE/EPA tailpipe
                CO2 emissions rulemakings have evaluated effects of
                standards over longer time periods, the early actions taken by
                manufacturers required more nuanced representation. Accordingly, the
                CAFE model now provides for a setting to establish a ``last year to
                consider credits.'' This adjustment is set at the last year for which
                new standards are not being considered (MY 2020 in this analysis). This
                allows the model to replicate the practical application of existing
                credits toward compliance in early years but also to examine the impact
                of proposed standards based solely on fuel economy improvements in all
                years for which new standards are being considered.
                 Regarding the model's simulation of manufacturers' potential
                earning and application of compliance credits, UCS commented that the
                model ``inexplicably lets credits expire'' because ``all technologies
                which pay for themselves within the assumed payback period are applied
                to all manufacturers, regardless of credit status.'' UCS also claimed
                that ``NHTSA did not accurately reflect unique attributes of EPA's
                credit bank,'' that ``credits are not traded between manufacturers,''
                and that ``NHTSA does not model credit carryback for compliance.''
                \411\ Relatedly, as discussed above, UCS attributes modeling outcomes
                to the ``effective cost'' metric used to select from among available
                fuel-saving technologies.\412\ As discussed in Section VI.B.1, the
                agencies expect that manufacturers are likely to improve fuel economy
                voluntarily insofar as doing so ``pays back'' economically within a
                short period (30 months), and the agencies note that periods of
                regulatory stability have, in fact, been marked by CAFE levels
                exceeding requirements. As discussed above, the agencies have excluded
                simulation of credit trading (except in MYs prior to those under
                consideration, aside from an idealized case presented in the
                sensitivity analysis) and likewise excluded simulation of potential
                ``carryback'' provisions. The agencies have excluded modeling these
                scenarios not just because of the analytical complexities involved (and
                rejecting, for example, the random number generator analysis suggested
                by UCS), but also because the agencies agree that the actual provisions
                regarding trading and borrowing of compliance credits create too much
                risk to be used in the analysis underlying consideration of standards.
                However, as discussed above, the agencies have revised the ``metric''
                used to prioritize available options to apply fuel-saving technologies.
                As discussed below, the agencies have revised model inputs to include
                the large quantity of ``legacy'' compliance credits EPA has made
                available under its CO2 standards.
                ---------------------------------------------------------------------------
                 \411\ UCS, NHTSA-2018-0067-12039, Technical Appendix, at 35-46.
                 \412\ UCS, NHTSA-2018-0067-12039, Technical Appendix, at 28-30.
                ---------------------------------------------------------------------------
                 The CAFE model has also been modified to include a similar
                representation of existing credit banks in EPA's CO2
                program. While the life of a CO2 credit, denominated in
                metric tons of CO2, has a five-year life, matching the
                lifespan of CAFE credits, such credits earned in the early MY 2009-2011
                years of the EPA program, may be used through MY 2021.\413\ The CAFE
                model was not modified to allow
                [[Page 24308]]
                exceptions to the life-span of compliance credits, and, to reflect
                statutory requirements, treated them as if they may be carried forward
                for no more than five years, so the initial credit banks were modified
                to anticipate the years in which those credits might be needed. MY 2016
                was simulated explicitly in the NPRM analysis to prohibit the inclusion
                of banked credits in MY 2016 (which could be carried forward from MY
                2016 to MY 2021), and thus underestimated the extent to which
                individual manufacturers, and the industry as a whole, could rely on
                these early credits to comply with EPA standards between MY 2016 and MY
                2021. However, as indicated in the NPRM, the final rule's model inputs
                updated the analysis fleet's basis to MY 2017, such that these
                additional banked credits can be included. The credit banks with which
                the simulations in this analysis were conducted are presented in the
                following Tables:
                ---------------------------------------------------------------------------
                 \413\ In the 2010 rule, EPA placed limits on credits earned in
                MY 2009, which expired prior to this rule. However, credits
                generated in MYs 2010-2011 may be carried forward, or traded, and
                applied to deficits generated through MY 2021.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.100
                [[Page 24309]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.101
                BILLING CODE 4910-59-C
                 While the CAFE model does not simulate the ability to trade credits
                between manufacturers, it does simulate the strategic accumulation and
                application of compliance credits, as well as the ability to transfer
                credits between fleets to improve the compliance position of a less
                efficient fleet by leveraging credits earned by a more efficient fleet.
                The model prefers to hold on to earned compliance credits within a
                given fleet, carrying them forward into the future to offset potential
                future deficits. This assumption is consistent with observed strategic
                manufacturer behavior dating back to 2009.
                 From 2009 to present, no manufacturer has transferred CAFE credits
                into a fleet to offset a deficit in the same year in which they were
                earned. This has occurred with credits acquired from other
                manufacturers via trade but not with a manufacturer's own credits.
                Therefore, the current representation of credit transfers between
                fleets--where the model prefers to transfer expiring, or soon-to-be-
                expiring credits rather than newly earned credits--is both appropriate
                and consistent with observed industry behavior.
                 This may not be the case for CO2 standards, though it is
                difficult to be certain at this point. The CO2 program
                seeded the industry with a large quantity of early compliance credits
                (earned in MYs 2009-2011) \414\ prior to the existence formal
                CO2 standards. Early credits from MYs 2010 and 2011,
                however, do not expire until 2021. Thus, for manufacturers looking to
                offset deficits, it is more sensible to exhaust credits that were
                generated during later model years (which are set to expire within the
                next five years), rather than relying on the initial bank of credits
                from MYs 2010 and 2011. The first model year for which earned credits
                outlive the initial bank is MY 2017, for which final manufacturer
                CO2 performance data (and hence, banked credits) has not yet
                been released. However, considering that under the CO2
                program manufacturers simultaneously comply with passenger car and
                light truck fleets, to more accurately represent the CO2
                credit system the CAFE model allows (and encourages) intra-year
                transfers between regulated fleets for the purpose of simulating
                compliance with the CO2 standards.
                ---------------------------------------------------------------------------
                 \414\ In response to public comment, EPA eliminated the possible
                use of credits earned in MY 2009 for future model years. However,
                credits earned in MY 2010 and MY 2011 remain available for use.
                ---------------------------------------------------------------------------
                a) Off-Cycle and A/C Efficiency Adjustments to CAFE and Average
                CO2 Levels
                 In addition to more rigorous accounting of CAFE and CO2
                credits, the model now also accounts for air conditioning efficiency
                and off-cycle adjustments. NHTSA's program considers those adjustments
                in a manufacturer's compliance calculation starting in MY 2017, and the
                NPRM version of the model used the adjustments claimed by each
                manufacturer in MY 2016 as the starting point for all future years.
                Because air conditioning efficiency and off-cycle adjustments are not
                credits in NHTSA's program, but rather adjustments to compliance fuel
                economy (much like the Flexible Fuel Vehicle adjustments due to phase
                out in MY 2019), they may be included under either a ``standard
                setting'' or ``unconstrained'' analysis perspective.
                 The manner in which the CAFE model treats the EPA and CAFE A/C
                efficiency and off-cycle credit programs is similar, but the model also
                accounts for A/C leakage (which is not part of NHTSA's program). When
                determining the compliance status of a
                [[Page 24310]]
                manufacturer's fleet (in the case of EPA's program, PC and LT are the
                only fleet distinctions), the CAFE model weighs future compliance
                actions against the presence of existing (and expiring) CO2
                credits resulting from over-compliance with earlier years' standards,
                A/C efficiency credits, A/C leakage credits, and off-cycle credits.
                 Another aspect of credit accounting, implemented in the NPRM
                version of the CAFE model, involved credits related to the application
                of off-cycle and A/C efficiency adjustments, which manufacturers earn
                by taking actions such as special window glazing or using reflective
                paints that provide fuel economy improvements in real-world operation
                but do not produce measurable improvements in fuel consumption on the
                2-cycle test.
                 NHTSA's inclusion of off-cycle and A/C efficiency adjustments began
                in MY 2017, while EPA has collected several years' worth of submissions
                from manufacturers about off-cycle and A/C efficiency technology
                deployment. Currently, the level of deployment can vary considerably by
                manufacturer, with several claiming extensive Fuel Consumption
                Improvement Values (FCIV) for off-cycle and A/C efficiency
                technologies, and others almost none. The analysis of alternatives
                presented here (and in the NPRM) does not attempt to project how future
                off-cycle and A/C efficiency technology use will evolve or speculate
                about the potential proliferation of FCIV proposals submitted to the
                agencies. Rather, this analysis uses the off-cycle credits submitted by
                each manufacturer for MY 2017 compliance, and, with a few exceptions,
                carries these forward to future years. Several of the technologies
                described below are associated with A/C efficiency and off-cycle FCIVs.
                In particular, stop-start systems, integrated starter generators, and
                full hybrids are assumed to generate off-cycle adjustments when applied
                to vehicles to improve their fuel economy. Similarly, higher levels of
                aerodynamic improvements are assumed to include active grille shutters
                on the vehicle, which also qualify for off-cycle FCIVs.
                 The NPRM analysis assumed that any off-cycle FCIVs that are
                associated with actions outside of the technologies discussed in
                Section VI.C (either chosen from the pre-approved ``pick list,'' or
                granted in response to individual manufacturer petitions) remained at
                the levels claimed by manufacturers in MY 2017. Any additional A/C
                efficiency and off-cycle adjustments that accrued as the result of
                explicit technology application calculated dynamically in each model
                year for each alternative. The NPRM version of the CAFE model also
                represented manufacturers' credits for off-cycle improvements, A/C
                efficiency improvements, and A/C leakage reduction in terms of values
                applicable across all model years.
                 Recognizing that application of these improvements thus far varies
                considerably among manufacturers, such that some manufacturers have
                opportunities to earn significantly more of the corresponding
                adjustments over time, the agencies have expanded the CAFE model's
                representation of these credits to provide for year-by-year
                specification of the amounts of each type of adjustment for each
                manufacturer, denominated in grams CO2 per mile,\415\ as
                summarized in the following table:
                ---------------------------------------------------------------------------
                 \415\ For estimating their contribution to CAFE compliance, the
                grams CO2/mile values in Table VI-1711 are converted to
                gallons/mile and applied to a manufacturer's 2-cycle CAFE
                performance. When calculating compliance with EPA's CO2
                program, there is no conversion necessary (as standards are also
                denominated in grams/mile).
                 \416\ These values are specified in the ``market_ref.xlsx''
                input file's ``Credits and Adjustments'' worksheet. The file is
                available with the archive of model inputs and outputs posted at
                https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24311]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.102
                [[Page 24312]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.103
                [[Page 24313]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.104
                BILLING CODE 4910-59-C
                 In addition to these refinements to the estimation of the
                quantities of adjustments earned over time by each manufacturer, the
                agencies revised the
                [[Page 24314]]
                CAFE model to apply estimates of the corresponding costs. For today's
                analysis, the agencies applied estimates developed previously by EPA,
                adjusting these values to 2019 dollars. The following table summarizes
                inputs through model year 2030:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.105
                 The model currently accounts for any off-cycle adjustments
                associated with technologies that are included in the set of fuel-
                saving technologies explicitly simulated as part of this proposal (for
                example, start-stop systems that reduce fuel consumption during idle or
                active grille shutters that improve aerodynamic drag at highway speeds)
                and accumulates these adjustments up to the 10 g/mi cap. As a practical
                matter, most of the adjustments for which manufacturers are claiming
                off-cycle FCIV exist outside of the technology tree, so the cap is
                rarely reached during compliance simulation. The agencies have
                considered the potential to model their application explicitly.
                However, doing so would require data regarding which vehicle models
                already possess these improvements as well as the cost and expected
                value of applying them to other models in the future. Such data is
                currently too limited to support explicit modeling of these
                technologies and adjustments.
                b) Alternative Fuel Vehicles
                 When establishing maximum feasible fuel economy standards, NHTSA is
                prohibited from considering the availability of alternatively fueled
                vehicles,\417\ and credit provisions related to AFVs that significantly
                increase their fuel economy for CAFE compliance purposes. Under the
                ``standard setting'' perspective, these technologies (pure battery
                electric vehicles and fuel cell vehicles) \418\ are not available in
                the compliance simulation to improve fuel economy. Under the
                ``unconstrained'' perspective, such as is documented in the DEIS and
                FEIS, the CAFE model considers these technologies in the same manner as
                other available technologies, and may apply them if they represent
                cost-effective compliance pathways. However, under both perspectives,
                the analysis continues to include dedicated AFVs that already exist in
                the MY 2017 fleet (and their projected future volumes). Also, because
                the CAA provides no direction regarding consideration of alternative
                fuels, the final rule's analysis includes simulation of the potential
                that some manufacturers might introduce new AFVs in response to
                CO2 standards. To represent the compliance benefit from such
                a response fully, NHTSA modified the CAFE model to include the specific
                provisions related to AFVs under the CO2 standards. In
                particular, the CAFE model now carries a full representation of the
                production multipliers related to electric vehicles, fuel cell
                vehicles, plug-in hybrids, and CNG vehicles, all of which vary by year
                through MY 2021.
                ---------------------------------------------------------------------------
                 \417\ 49 U.S.C. 32902(h).
                 \418\ Dedicated compressed natural gas (CNG) vehicles should
                also be excluded in this perspective but are not considered as a
                compliance strategy under any perspective in this analysis.
                ---------------------------------------------------------------------------
                 EPCA also provides that CAFE levels may, subject to limitations, be
                adjusted upward to reflect the sale of flexible fuel vehicles (FFVs).
                Although these adjustments end after model year 2020, the final rule's
                analysis, like the NPRM's, includes estimated potential use through MY
                2019, as summarized below:
                [[Page 24315]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.106
                 For its part, EPA has provided that manufacturers selling
                sufficient numbers of PHEVs, BEVs, and FCVs may, when calculating fleet
                average CO2 levels, ``count'' each unit of production as
                more than a single unit. The CAFE model accounts for these
                ``multipliers.'' As for the NPRM, the final rule's analysis applies the
                following multipliers:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.107
                 For example, under EPA's current regulation, when calculating the
                average CO2 level achieved by its MY 2019 passenger car
                fleet, a manufacturer may treat each 1,000 BEVs as 2,000 BEVs. When
                calculating the average level required of this fleet, the manufacturer
                must use the actual production volume (in this example, 1,000 units).
                Similarly, the manufacturer must use the actual production volume when
                calculating compliance credit balances.
                 There were no natural gas vehicles in the baseline fleet, and the
                analysis did not apply natural gas technology due to cost
                effectiveness. The application of a 2.0 multiplier for natural gas
                vehicles for MYs 2022-2026 would have no impact on the analysis because
                given the state of natural gas vehicle refueling infrastructure, the
                cost to equip vehicles with natural gas tanks, the outlook for
                petroleum prices, and the outlook for battery prices, we have little
                basis to project more than an inconsequential response to this
                incentive in the foreseeable future.
                 For the final rule's analysis, the CAFE model can be exercised in a
                manner that simulates these current EPA requirements, or that simulates
                two alternative approaches. The first includes the above-mentioned
                multipliers in the calculation of average requirements, and the second
                also includes the multipliers in the calculation of credit balances.
                The central analysis reflects current regulations. The sensitivity
                analysis presented in the FRIA includes a case
                [[Page 24316]]
                applying multipliers to the calculation of achieved and required
                average CO2 levels, and calculation of credit balances.
                c) Civil Penalties
                 Throughout the history of the CAFE program, some manufacturers have
                consistently achieved fuel economy levels below applicable standards,
                electing instead to pay civil penalties as specified by EPCA. As in
                previous versions of the CAFE model, the current version allows the
                user to specify inputs identifying such manufacturers and to consider
                their compliance decisions as if they are willing to pay civil
                penalties for non-compliance with the CAFE program. As with the NPRM,
                the civil penalty rate in the current analysis is $5.50 per 1/10 of a
                mile per gallon, per vehicle manufactured for sale.
                 NHTSA notes that treating a manufacturer as if it is willing to pay
                civil penalties does not necessarily mean that it is expected to pay
                penalties in reality. Doing so merely implies that the manufacturer
                will only apply fuel economy technology up to a point, and then stop,
                regardless of whether or not its corporate average fuel economy is
                above its standard. In practice, the agencies expect that many of these
                manufacturers will continue to be active in the credit market, using
                trades with other manufacturers to transfer credits into specific
                fleets that are challenged in any given year, rather than paying
                penalties to resolve CAFE deficits. The CAFE model calculates the
                amount of penalties paid by each manufacturer, but it does not simulate
                trades between manufacturers. In practice, some (possibly most) of the
                total estimated penalties may be a transfer from one OEM to another.
                 Although EPCA, as amended in 2007 by the Energy Independence and
                Security Act (EISA), prescribes these specific civil penalty provisions
                for CAFE standards, the Clean Air Act (CAA) does not contain similar
                provisions. Rather, the CAA's provisions regarding noncompliance
                prohibit sale of a new motor vehicle that is not covered by an EPA
                certificate of conformity, and in order to receive such a certificate
                the new motor vehicle must meet EPA's Section 202 regulations,
                including applicable emissions standards. Therefore, inputs regarding
                civil penalties--including inputs regarding manufacturers' potential
                willingness to treat civil penalty payment as an economic choice--apply
                only to simulation of CAFE standards. On the other hand, some of the
                same manufacturers recently opting to pay civil penalties instead of
                complying with CAFE standards have also recently led adoption of lower-
                GWP refrigerants, and the ``A/C leakage'' credits count toward
                compliance only with CO2 standards, not CAFE standards. The
                model accounts for this difference between the programs.
                 When considering technology applications to improve fleet fuel
                economy, the model will add technology up to the point at which the
                effective cost of the technology (which includes technology cost,
                consumer fuel savings, consumer welfare changes, and the cost of
                penalties for non-compliance with the standard) is less costly than
                paying civil penalties or purchasing credits. Unlike previous versions
                of the model, the current implementation further acknowledges that some
                manufacturers experience transitions between product lines where they
                rely heavily on credits (either carried forward from earlier model
                years or acquired from other manufacturers) or simply pay penalties in
                one or more fleets for some number of years. The model now allows the
                user to specify, when appropriate for the regulatory program being
                simulated, on a year-by-year basis, whether each manufacturer should be
                considered as willing to pay penalties for non-compliance. This
                provides additional flexibility, particularly in the early years of the
                simulation. As discussed above, this assumption is best considered as a
                method to allow a manufacturer to under-comply with its standard in
                some model years--treating the civil penalty rate and payment option as
                a proxy for other actions it may take that are not represented in the
                CAFE model (e.g., purchasing credits from another manufacturer, carry-
                back from future model years, or negotiated settlements with NHTSA to
                resolve deficits).
                 For the NPRM, NHTSA relied on past compliance behavior and
                certified transactions in the credit market to designate some
                manufacturers as willing to pay CAFE penalties in some model years. The
                full set of NPRM assumptions regarding manufacturer behavior with
                respect to civil penalties is presented in Table VI-21, which shows all
                manufacturers were assumed to be willing to pay civil penalties prior
                to MY 2020. This was largely a reflection of either existing credit
                balances (which manufacturers will use to offset CAFE deficits until
                the credits reach their expiration dates) or inter-manufacturer trades
                assumed likely to happen in the near future, based on previous
                behavior. The manufacturers in the table whose names appear in bold all
                had at least one regulated fleet (of three) whose CAFE was below its
                standard in MY 2016. Because the NPRM analysis began with the MY 2016
                fleet, and no technology could be added to vehicles that are already
                designed and built, all manufacturers could generate civil penalties in
                MY 2016. However, once a manufacturer is designated as unwilling to pay
                penalties, the CAFE model will attempt to add technology to the
                respective fleets to avoid shortfalls.
                [[Page 24317]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.108
                 Several of the manufacturers in Table VI-21 that were presumed to
                be willing to pay civil penalties in the early years of the program
                have no history of paying civil penalties. However, several of those
                manufacturers have either bought or sold credits--or transferred
                credits from one fleet to another to offset a shortfall in the
                underperforming fleet. As the CAFE model does not simulate credit
                trades between manufacturers, providing this additional flexibility in
                the modeling avoids the outcome where the CAFE model applies more
                technology than needed in the context of the full set of compliance
                flexibilities at the industry level. By statute, NHTSA cannot consider
                credit flexibilities when setting standards, so most manufacturers
                (those without a history of civil penalty payment) are assumed to
                comply with their standards through fuel economy improvements for the
                model years being considered in this analysis. The notable exception to
                this assumption is Fiat Chrysler Automobiles (FCA), which could still
                satisfy the requirements of the program through a combination of credit
                application and civil penalties through MY 2025 before eventually
                complying exclusively through fuel economy improvements in MY 2026.
                 As mentioned above, the CAA does not provide civil penalty
                provisions similar to those provisions specified in EPCA/EISA, and the
                above-mentioned corresponding inputs apply only to simulation of
                compliance with CAFE standards.
                 Some stakeholders offering comments related to the analytical
                treatment of civil penalties indicated that NHTSA should tend toward
                assuming manufacturers will take advantage of this EPCA provision as an
                economically attractive alternative to compliance. Other commenters
                implied that NHTSA should tend toward not relying on compliance
                flexibilities in the analysis used to determine the maximum feasible
                stringency of CAFE standards. For example, New York University's
                Institute for Policy Integrity (IPI) offered the following comments:
                 NHTSA assumes that most manufacturers will be unwilling to pay
                penalties based in part on the fact that most manufacturers have not
                paid penalties in recent years. The Proposed Rule cites the
                statutory prohibition on NHTSA considering credit trading as a
                reason to assume manufacturers without a history of paying penalties
                will comply through technology alone, whatever the cost. But this is
                an arbitrary assumption and is in no way dictated by the statute.
                NHTSA knows as much, since elsewhere in the proposed rollback, the
                agency explains ``EPCA is very clear as to which flexibilities are
                not to be considered'' and NHTSA is allowed to consider off-cycle
                adjustments because they are not specifically mentioned. But
                considering penalties are not mentioned as off-limits for NHTSA in
                setting the standards either. Instead, the prohibition focuses on
                credit trading and transferring. The penalty safety valve has
                existed in EPCA for decades, and Congress clearly would have known
                how to add penalties to the list of trading and transferring. The
                fact that Congress did not bar NHTSA from considering penalties as a
                safety valve means that NHTSA must consider manufacturer's efficient
                use of penalties as a cost minimizing compliance option. Besides,
                NHTSA does consider penalties for some of the manufacturers making
                its statutory justification even less rational.\419\
                ---------------------------------------------------------------------------
                 \419\ Institute for Policy Integrity, NHTSA-2018-0067-12213, at
                24.
                 On the other hand, in more general comments about NHTSA's
                analytical treatment of program flexibilities, FCA stated that ``when
                flexibilities are considered while setting targets, they cease to be
                flexibilities and become simply additional technology mandates.'' \420\
                ---------------------------------------------------------------------------
                 \420\ FCA, Docket #NHTSA-2018-0067-11943, at 6.
                ---------------------------------------------------------------------------
                 NHTSA agrees with IPI that EPCA does not expressly prohibit NHTSA,
                when conducting analysis supporting determinations of the maximum
                feasible stringency of future CAFE standards, from including
                manufacturers' potential tendency to pay civil penalties rather than
                complying with those standards. However, EPCA also does not require
                NHTSA to include this tendency in its analysis. NHTSA also notes, as
                does IPI, that EPCA does prohibit NHTSA from including credit trading,
                transferring, or the availability of credits in such
                [[Page 24318]]
                analysis (although NHTSA interprets this prohibition to apply only to
                the model years for which standards are being set). This statutory
                difference is logical based on the way credits and penalties function
                differently under EPCA. Because credits help manufacturers achieve
                compliance with CAFE standards, absent the statutory prohibition,
                credits would be relevant to the feasibility of a standard.\421\
                Penalties, on the other hand, do not enable a manufacturer to comply
                with an applicable standard; penalties are for noncompliance.\422\ When
                Congress added credit trading provisions to EPCA in 2007, NHTSA
                anticipated that competitive considerations would make manufacturers
                reluctant to engage in such trades. Since that time, manufacturers
                actually have demonstrated otherwise, although the reliance on
                trading--especially between specific pairs of OEMs--appears to vary
                widely. At this time, NHTSA considers it most likely that manufacturers
                will shift away from paying civil penalties and toward compliance
                credit trading. Consequently, for NHTSA to include civil penalty
                payment in its analysis would increasingly amount to using civil
                penalty payment as an analytical proxy for credit trading. Having
                further considered the question, NHTSA's current view is, therefore,
                that including civil penalty payment beyond MY 2020 would effectively
                subvert EPCA's prohibition against considering credit trading.
                Therefore, for today's announcement, NHTSA has modified its analysis to
                assume that BMW, Daimler, FCA, JLR, and Volvo would consider paying
                civil penalties through MY 2020, and that all manufacturers would apply
                as much technology as would be needed in order to avoid paying civil
                penalties after MY 2020.
                ---------------------------------------------------------------------------
                 \421\ See 49 U.S.C. 32911(b) (``Compliance is determined after
                considering credits available to the manufacturer . . . . '').
                 \422\ See id.
                ---------------------------------------------------------------------------
                3. Technology Effectiveness Values
                 The next input required to simulate manufacturers' decision-making
                processes for the year-by-year application of technologies to specific
                vehicles is estimates of how effective each technology would be at
                reducing fuel consumption. In the NPRM, the agencies used full-vehicle
                modeling and simulation to estimate the fuel economy improvements
                manufacturers could make to a fleet of vehicles, considering those
                vehicles' technical specifications and how combinations of technologies
                interact. Full-vehicle modeling and simulation uses computer software
                and physics-based models to predict how combinations of technologies
                perform as a full system under defined conditions.
                 A model is a mathematical representation of a system, and
                simulation is the behavior of that mathematical representation over
                time. In this analysis, the model is a mathematical representation of
                an entire vehicle,\423\ including its individual components such as the
                engine and transmission, overall vehicle characteristics such as mass
                and aerodynamic drag, and the environmental conditions, such as ambient
                temperature and barometric pressure. The agencies simulated the model's
                behavior over test cycles, including the 2-cycle laboratory compliance
                tests (or 2-cycle tests),\424\ to determine how the individual
                components interact. 2-cycle tests are test cycles that are used to
                measure fuel economy and emissions for CAFE and CO2
                compliance, and therefore are the relevant test cycles for determining
                technology effectiveness when establishing standards. In the
                laboratory, 2-cycle testing involves sophisticated test and measurement
                equipment, carefully controlled environmental conditions, and precise
                procedures to provide the most repeatable results possible with human
                drivers. Measurements using these structured procedures serve as a
                yardstick for fuel economy and CO2 emissions.
                ---------------------------------------------------------------------------
                 \423\ Our full vehicle model was composed of sub-models, which
                is why the full vehicle model could also be referred to as a full
                system model, composed of sub-system models.
                 \424\ EPA's compliance test cycles are used to measure the fuel
                economy of a vehicle. For readers unfamiliar with this process, it
                is like running a car on a treadmill following a program--or more
                specifically, two programs. The ``programs'' are the ``urban
                cycle,'' or Federal Test Procedure (abbreviated as ``FTP''), and the
                ``highway cycle,'' or Highway Fuel Economy Test (abbreviated as
                ``HFET''), and they have not changed substantively since 1975. Each
                cycle is a designated speed trace (of vehicle speed versus time)
                that all certified vehicles must follow during testing. The FTP is
                meant roughly to simulate stop and go city driving, and the HFET is
                meant roughly to simulate steady flowing highway driving at about 50
                mph. For further details on compliance testing, see the discussion
                in Section VI.B.3.a)(7).
                ---------------------------------------------------------------------------
                 Full-vehicle modeling and simulation was initially developed to
                avoid the costs of designing and testing prototype parts for every new
                type of technology. For example, if a truck manufacturer has a concept
                for a lightweight tailgate and wants to determine the fuel economy
                impact for the weight reduction, the manufacturer can use physics-based
                computer modeling to estimate the impact. The vehicle, modeled with the
                proposed change, can be simulated on a defined test route and under a
                defined test condition, such as city or highway driving in warm ambient
                temperature conditions, and compared against the baseline reference
                vehicle. Full-vehicle modeling and simulation allows the consideration
                and evaluation of different designs and concepts before building a
                single prototype. In addition, full vehicle modeling and simulation is
                beneficial when considering technologies that provide small incremental
                improvements. These improvements are difficult to measure in laboratory
                tests due to variations in how vehicles are driven over the test cycle
                by human drivers, variations in emissions measurement equipment, and
                variations in environmental conditions.\425\
                ---------------------------------------------------------------------------
                 \425\ Difficulty with controlling for such variability is
                reflected, for example, in 40 CFR 1065.210, Work input and output
                sensors, which describes complicated instructions and
                recommendations to help control for variability in real world (non-
                simulated) test instrumentation set up.
                ---------------------------------------------------------------------------
                 Full-vehicle modeling and simulation requires detailed data
                describing the individual technologies and performance-related
                characteristics. Those specifications generally come from design
                specifications, laboratory measurements, and other subsystem
                simulations or modeling. One example of data used as an input to the
                full vehicle simulation are engine maps for each engine technology that
                define how much fuel is consumed by the engine technology across its
                operating range.
                 Using full-vehicle modeling and simulation to estimate technology
                efficiency improvements has two primary advantages over using single or
                limited point estimates. An analysis using single or limited point
                estimates may assume that, for example, one fuel economy improving
                technology with an effectiveness value of 5 percent by itself and
                another technology with an effectiveness value of 10 percent by itself,
                when applied together achieve an additive improvement of 15 percent.
                Single point estimates generally do not provide accurate effectiveness
                values because they do not capture complex relationships among
                technologies. Technology effectiveness often differs significantly
                depending on the vehicle type (e.g., sedan versus pickup truck) and how
                the technology interacts with other technologies on the vehicle, as
                different technologies may provide different incremental levels of fuel
                economy improvement if implemented alone or in tandem with other
                technologies. Any oversimplification of these complex interactions
                leads to less accurate and often overestimated effectiveness estimates.
                 In addition, because manufacturers often implement several fuel-
                saving
                [[Page 24319]]
                technologies simultaneously when redesigning a vehicle, it is difficult
                to isolate the effect of individual technologies using laboratory
                measurement of production vehicles alone. Modeling and simulation
                offers the opportunity to isolate the effects of individual
                technologies by using a single or small number of baseline vehicle
                configurations and incrementally adding technologies to those baseline
                configurations. This provides a consistent reference point for the
                incremental effectiveness estimates for each technology and for
                combinations of technologies for each vehicle type. Vehicle modeling
                also reduces the potential for overcounting or undercounting technology
                effectiveness.
                 An important feature of this analysis is that the incremental
                effectiveness of each technology and combinations of technologies be
                accurate and relative to a consistent baseline vehicle. The absolute
                fuel economy values of the full vehicle simulations are used only to
                determine incremental effectiveness and are never used directly to
                assign an absolute fuel economy value to any vehicle model or
                configuration for the rulemaking analysis.
                 For this analysis, absolute fuel economy levels are based on the
                individual fuel economy values from CAFE compliance data for each
                vehicle in the baseline fleet. The incremental effectiveness from the
                full vehicle simulations performed in Autonomie, a physics-based full-
                vehicle modeling and simulation software developed and maintained by
                the U.S. Department of Energy's Argonne National Laboratory, are
                applied to baseline fuel economy to determine the absolute fuel economy
                of applying the first technology change. For subsequent technology
                changes, incremental effectiveness is applied to the absolute fuel
                economy level of the previous technology configuration.
                 For example, if a Ford F150 2-wheel drive crew cab and short bed in
                the baseline fleet has a fuel economy value of 30 mpg for CAFE
                compliance, 30 mpg will be considered the reference absolute fuel
                economy value. A similar full vehicle model in the Autonomie simulation
                may begin with an average fuel economy value of 32 mpg, and with
                incremental addition of a specific technology X its fuel economy
                improves to 35 mpg, a 9.3 percent improvement. In this example, the
                incremental fuel economy improvement (9.3 percent) from technology X
                would be applied to the F150's 30 mpg absolute value.
                 For this analysis, the agencies determined the incremental
                effectiveness of technologies as applied to the 2,952 unique vehicle
                models in the analysis fleet. Although, as mentioned above, full-
                vehicle modeling and simulation reduces the work and time required to
                assess the impact of moving a vehicle from one technology state to
                another, it would be impractical--if not impossible--to build a unique
                vehicle model for every individual vehicle in the analysis fleet.
                Therefore, as explained further below, vehicle models are built in a
                way that maintains similar attributes to the analysis fleet vehicles,
                which ensures key components are reasonably represented.
                 We received a wide array of comments regarding the full-vehicle
                modeling and simulation performed for the NPRM, but there was general
                agreement that full-vehicle modeling and simulation was the appropriate
                method to determine technology effectiveness.\426\ Stakeholders
                commented on other areas, such as full vehicle simulation tools,
                inputs, and assumptions, and these comments will be discussed in the
                following sections. For this final rule, the agencies continued to use
                the same full-vehicle simulation approach to estimate technology
                effectiveness for technology adoption in the rulemaking timeframe. The
                next sections will discuss the details of the explicit input
                specifications and assumptions used for the final rule analysis.
                ---------------------------------------------------------------------------
                 \426\ See NHTSA-2018-0067-12039; NHTSA-2018-0067-12073. UCS and
                AAM both agreed that full vehicle simulation can significantly
                improve the estimates of technology effectiveness.
                ---------------------------------------------------------------------------
                a) Why This Rulemaking Used Autonomie Full-Vehicle Modeling and
                Simulation To Determine Technology Effectiveness
                 The NPRM and final rule analysis use effectiveness estimates for
                technologies developed using Autonomie, a physics-based full-vehicle
                modeling and simulation software developed and maintained by the U.S.
                Department of Energy's Argonne National Laboratory.\427\ Autonomie was
                designed to serve as a single tool to meet requirements of automotive
                engineering throughout the vehicle development process, and has been
                under continuous improvement by Argonne for over 20 years. Autonomie is
                commercially available and widely used in the automotive industry by
                suppliers, automakers, and academic researchers (who publish findings
                in peer reviewed academic journals).\428\ DOE and manufacturers have
                used Autonomie and its ability to simulate a large number of powertrain
                configurations, component technologies, and vehicle-level controls over
                numerous drive cycles to support studies on fuel efficiency, cost-
                benefit analysis, and carbon dioxide emissions,\429\ and other topics.
                ---------------------------------------------------------------------------
                 \427\ More information about Autonomie is available at https://www.anl.gov/technology/project/autonomie-automotive-system-design
                (last accessed June 21, 2018). As mentioned in the preliminary
                regulatory impact analysis (PRIA) for this rule, the agencies used
                Autonomie version R15SP1, the same version used for the 2016 Draft
                TAR.
                 \428\ Rousseau, A. Shidore, N. Karbowski, D. Sharer, ``Autonomie
                Vehicle Validation Summary.'' https://www.nhtsa.gov/sites/nhtsa.dot.gov/files/anl-autonomie-vehicle-model-validation-1509.pdf.
                 \429\ Delorme et al. 2008, Rousseau, A, Sharer, P, Pagerit, S.,
                & Das, S. ``Trade-off between Fuel Economy and Cost for Advanced
                Vehicle Configurations,'' 20th International Electric Vehicle
                Symposium (EVS20), Monaco (April 2005); Elgowainy, A., Burnham, A.,
                Wang, M., Molburg, J., & Rousseau, A. ``Well-To-Wheels Energy Use
                and Greenhouse Gas Emissions of Plug-in Hybrid Electric Vehicles,''
                SAE 2009-01-1309, SAE World Congress, Detroit, April 2009.
                ---------------------------------------------------------------------------
                 Autonomie has also been used to provide the U.S. government with
                data to make decisions about future research, and is used by DOE for
                analysis supporting budget priorities and plans for programs managed by
                its Vehicle Technologies Office (VTO), and to support decision making
                among competing vehicle technology research and development
                projects.\430\ In addition, Autonomie is the primary vehicle simulation
                tool used by DOE to support its U.S. DRIVE program, a government-
                industry partnership focused on advanced automotive and related energy
                infrastructure technology research and development.\431\
                ---------------------------------------------------------------------------
                 \430\ U.S. DOE Benefits & Scenario Analysis publications is
                available at https://www.autonomie.net/publications/fuel_economy_report.html (last accessed September 11, 2019).
                 \431\ For more information on U.S. Drive, see https://www.energy.gov/eere/vehicles/us-drive.
                ---------------------------------------------------------------------------
                 Autonomie is a MathWorks-based software environment and framework
                for automotive control-system design, simulation, and analysis.\432\ It
                is designed for rapid and easy integration of models with varying
                levels of detail (low to high fidelity), abstraction (from subsystems
                to systems and entire architectures), and processes (e.g., calibration,
                validation). By building models automatically, Autonomie allows the
                quick simulation of many component technologies and powertrain
                configurations, and, in this case, to assess the energy consumption of
                advanced powertrain technologies. Autonomie simulates subsystems,
                [[Page 24320]]
                systems, or entire vehicles; evaluates and analyzes fuel efficiency and
                performance; performs analyses and tests for virtual calibration,
                verification, and validation of hardware models and algorithms;
                supports system hardware and software requirements; links to
                optimization algorithms; and supplies libraries of models for
                propulsion architectures of conventional powertrains as well as hybrid
                and electric vehicles.
                ---------------------------------------------------------------------------
                 \432\ Halbach, S. Sharer, P. Pagerit, P., Folkerts, C. &
                Rousseau, A. ``Model Architecture, Methods, and Interfaces for
                Efficient Math-Based design and Simulation of Automotive Control
                Systems,'' SAE 2010-01-0241, SAE World Congress, Detroit, April,
                2010.
                ---------------------------------------------------------------------------
                 With hundreds of pre-defined powertrain configurations along with
                vehicle level control strategies developed from dynamometer test data,
                Autonomie is a highly capable tool for analyzing advantages and
                drawbacks of applying different technology options within each
                technology family, including conventional, parallel hybrid, power-split
                hybrid electric vehicles (HEVs), plug-in hybrid electric vehicles
                (PHEVs), battery electric vehicles (BEV) and fuel cell vehicles (FCVs).
                Autonomie also allows users to evaluate the effect of component sizing
                on fuel consumption for different powertrain technologies as well as to
                define component requirements (e.g., power, energy) to maximize fuel
                displacement for a specific application.\433\ To evaluate properly any
                powertrain-configuration or component-sizing influence, vehicle-level
                control models are critical, especially for electric drive vehicles
                like hybrids and plug-in hybrids. Argonne has extensive expertise in
                developing vehicle-level control models based on different approaches,
                from global optimization to instantaneous optimization, rule-based
                optimization, and heuristic optimization.\434\
                ---------------------------------------------------------------------------
                 \433\ Nelson, P., Amine, K., Rousseau, A., & Yomoto, H. (EnerDel
                Corp.), ``Advanced Lithium-ion Batteries for Plug-in Hybrid-electric
                Vehicles,'' 23rd International Electric Vehicle Symposium (EVS23),
                Anaheim, CA, (Dec. 2007); Karbowski, D., Haliburton, C., & Rousseau,
                A. ``Impact of Component Size on Plug-in Hybrid Vehicles Energy
                Consumption using Global Optimization,'' 23rd International Electric
                Vehicle Symposium (EVS23), Anaheim, CA, (Dec. 2007).
                 \434\ Karbowski, D., Kwon, J., Kim, N., & Rousseau, A.,
                ``Instantaneously Optimized Controller for a Multimode Hybrid
                Electric Vehicle,'' SAE paper 2010-01-0816, SAE World Congress,
                Detroit, April 2010; Sharer, P., Rousseau, A., Karbowski, D., &
                Pagerit, S. ``Plug-in Hybrid Electric Vehicle Control Strategy--
                Comparison between EV and Charge-Depleting Options,'' SAE paper
                2008-01-0460, SAE World Congress, Detroit (April 2008); and
                Rousseau, A., Shidore, N., Carlson, R., & Karbowski, D. ``Impact of
                Battery Characteristics on PHEV Fuel Economy,'' AABC08.
                ---------------------------------------------------------------------------
                 Autonomie has been developed to consider real-world vehicle metrics
                like performance, hardware limitations, utility, and drivability
                metrics (e.g., towing capability, shift busyness, frequency of engine
                on/off transitions), which are important to producing realistic
                estimates of fuel economy and CO2 emission rates. This
                increasing realism has, in turn, steadily increased confidence in the
                appropriateness of using Autonomie to make significant investment
                decisions. Autonomie has also been validated for a number of powertrain
                configurations and vehicle classes using Argonne's Advanced Mobility
                Technology Laboratory (AMTL) (formerly Advanced Powertrain Research
                Facility, or APRF) vehicle test data.\435\
                ---------------------------------------------------------------------------
                 \435\ Jeong, J., Kim, N., Stutenberg, K., Rousseau, A.,
                ``Analysis and Model Validation of the Toyota Prius Prime.'' SAE
                2019-01-0369, SAE World Congress, Detroit, April 2019; Kim, N,
                Jeong, J. Rousseau, A. & Lohse-Busch, H. ``Control Analysis and
                Thermal Model Development of PHEV,'' SAE 2015-01-1157, SAE World
                Congress, Detroit, April 2015; Kim, N., Rousseau, A. & Lohse-Busch,
                H. ``Advanced Automatic Transmission Model Validation Using
                Dynamometer Test Data,'' SAE 2014-01-1778, SAE World Congress,
                Detroit, Apr. 14; Lee, D. Rousseau, A. & Rask, E. ``Development and
                Validation of the Ford Focus BEV Vehicle Model,'' 2014-01-1809, SAE
                World Congress, Detroit, Apr. 14; Kim, N., Kim, N., Rousseau, A., &
                Duoba, M. ``Validating Volt PHEV Model with Dynamometer Test Data
                using Autonomie,'' SAE 2013-01-1458, SAE World Congress, Detroit,
                Apr. 13; Kim, N., Rousseau, A., & Rask, E. ``Autonomie Model
                Validation with Test Data for 2010 Toyota Prius,'' SAE 2012-01-1040,
                SAE World Congress, Detroit, Apr. 12; Karbowski, D., Rousseau, A,
                Pagerit, S., & Sharer, P. ``Plug-in Vehicle Control Strategy--From
                Global Optimization to Real Time Application,'' 22th International
                Electric Vehicle Symposium (EVS22), Yokohama, (October 2006).
                ---------------------------------------------------------------------------
                 Argonne has spent several years developing, applying, and expanding
                the means to use distributed computing to exercise its Autonomie full-
                vehicle simulation tool over the scale necessary for realistic analysis
                to provide data for CAFE and CO2 standards rulemaking. The
                NPRM and PRIA detailed how Argonne used Autonomie to estimate the fuel
                economy impacts for roughly a million combinations of technologies and
                vehicle types.436 437 Argonne developed input parameters for
                Autonomie to represent every combination of vehicle, powertrain, and
                component technologies considered in this rulemaking. The sequential
                addition of more than 50 fuel economy-improving technologies to ten
                vehicle types generated more than 140,000 unique technology and vehicle
                combinations. Running the Autonomie powertrain sizing algorithms to
                determine the appropriate amount of engine downsizing needed to
                maintain overall vehicle performance when vehicle mass reduction is
                applied and for certain engine technology changes (discussed further,
                below) increased the total number of simulations to more than one
                million. The result of these simulations is a useful dataset
                identifying the impacts of combinations of vehicle technologies on
                energy consumption--a dataset that can be referenced as an input to the
                CAFE model for assessing regulatory compliance alternatives.
                ---------------------------------------------------------------------------
                 \436\ As part of the Argonne simulation effort, individual
                technology combinations simulated in Autonomie were paired with
                Argonne's BatPAC model to estimate the battery cost associated with
                each technology combination based on characteristics of the
                simulated vehicle and its level of electrification. Information
                regarding Argonne's BatPAC model is available at http://www.cse.anl.gov/batpac/.
                 \437\ Additionally, the impact of engine technologies on fuel
                consumption, torque, and other metrics was characterized using GT
                POWER simulation modeling in combination with other engine modeling
                that was conducted by IAV Automotive Engineering, Inc. (IAV). The
                engine characterization ``maps'' resulting from this analysis were
                used as inputs for the Autonomie full-vehicle simulation modeling.
                Information regarding GT Power is available at https://www.gtisoft.com/gt-suite-applications/propulsion-systems/gt-power-engine-simulation-software.
                ---------------------------------------------------------------------------
                 The following sections discuss the full-vehicle modeling and
                simulation inputs and data assumptions, and comments received on the
                NPRM analysis. The discussion is necessarily technical, but also
                important to understand the agencies' decisions to modify (or not) the
                Autonomie analysis for the final rule.
                (1) Full-Vehicle Modeling, Simulation Inputs and Data Assumptions
                 The agencies provided extensive documentation that quantitatively
                and qualitatively described the over 50 technologies considered as
                inputs to the Autonomie modeling.438 439 These inputs
                consisted of engine technologies, transmission technologies, powertrain
                electrification, light-weighting, aerodynamic improvements, and tire
                rolling resistance improvements.\440\ The PRIA provided an overview of
                the sub-models for each technology, including the internal combustion
                engine model, automatic transmission model, and others.\441\ The
                Argonne NPRM model documentation expanded on these sub-models in detail
                to show the interaction of each sub-model input and output.\442\
                [[Page 24321]]
                For example, as shown in Figure VI-2, the input for Autonomie's driver
                model (i.e., the model used to approximate the driving behavior of a
                real driver) is vehicle speed, and outputs are accelerator pedal, brake
                pedal, and torque demand.
                ---------------------------------------------------------------------------
                 \438\ NHTSA-2018-0067-12299. Preliminary Regulatory Impact
                Analysis (July 2018).
                 \439\ NHTSA-2018-0067-0007. Islam, E., S, Moawad, A., Kim, N,
                Rousseau, A. ``A Detailed Vehicle Simulation Process To Support CAFE
                Standards 04262018--Report'' ANL Autonomie Documentation. Aug 21,
                2018. NHTSA-2018-0067-0004. ANL Autonomie Data Dictionary. Aug 21,
                2018. NHTSA-2018-0067-0003. ANL Autonomie Summary of Main Component
                Assumptions. Aug 21, 2018. NHTSA-2018-0067-0005. ANL Autonomie Model
                Assumptions Summary. Aug 21, 2018. NHTSA-2018-0067-1692. ANL BatPac
                Model 12 55. Aug 21, 2018.
                 \440\ SAFE Rule for MY2021-2026 PRIA Chapter 6.2.3 Technology
                groups in Autonomie simulations and CAFE model.
                 \441\ PRIA at 189.
                 \442\ NHTSA-2018-0067-0007. Islam, E., S, Moawad, A., Kim, N,
                Rousseau, A. ``A Detailed Vehicle Simulation Process To Support CAFE
                Standards 04262018--Report'' ANL Autonomie Documentation. Aug 21,
                2018.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.109
                 Effectiveness inputs for the NPRM and the final rule analysis were
                specifically developed to consider many real world and compliance test
                cycle constraints, to the extent a computer model could capture them.
                Examples include the advanced engine knock model discussed below, in
                addition to other constraints like allowing cylinder deactivation to
                occur in ways that would not negatively impact noise-vibration-
                harshness (NVH), and similarly optimizing the number of engine on/off
                events (e.g., from start/stop 12V micro hybrid systems) to balance
                between effectiveness and NVH.
                 One major input used in the effectiveness modeling that the
                agencies provided key specifications for in the PRIA are engine fuel
                maps that define how an engine equipped with specific technologies
                operates over a variety of engine load (torque) and engine speed
                conditions. The engine maps used as inputs to the Autonomie modeling
                portion of the analysis were developed by starting with a base map and
                then modifying that base map, incrementally, to model the addition of
                engine technologies. These engine maps, developed using the GT-Power
                modeling tool by IAV, were based off real-world engine designs.
                Simulated operation of these engines included the application of an IAV
                knock model, also developed from real-world engine
                data.443 444 Using this process, which incorporated real-
                world data, ensured that real-world constraints were considered for
                each vehicle type. Although the same type of engine map is used for all
                technology classes, the effectiveness varies based on the
                characteristics of each vehicle type. For example, a compact car with a
                turbocharged engine will have different fuel economy and performance
                values than a pickup truck with the same engine technology type. The
                engine map specifications are discussed further in Section VI.C.1 of
                this preamble and Section VI of FRIA.
                ---------------------------------------------------------------------------
                 \443\ Engine knock in spark ignition engines occurs when
                combustion of some of the air/fuel mixture in the cylinder does not
                result from propagation of the flame front ignited by the spark
                plug, but one or more pockets of air/fuel mixture explodes outside
                of the envelope of the normal combustion front.
                 \444\ See IAV material submitted to the docket; IAV_20190430_Eng
                22-26 Updated_Docket.pdf,
                IAV_Engine_tech_study_Sept_2016_Docket.pdf, IAV_Study for 4 Cylinder
                Gas Engines_Docket.pdf.
                ---------------------------------------------------------------------------
                 The agencies also provided key details about input assumptions for
                various vehicle specifications like transmission gear ratios, tire
                size, final drive ratios, and individual component weights.\445\ Each
                of these assumptions, to some extent, varied between the ten technology
                classes to capture appropriately real-world vehicle specifications like
                wheel mass or fuel tank mass. These specific input assumptions were
                developed based on the latest test data and current market fleet
                information.\446\ The agencies relied on default assumptions developed
                by the Autonomie team, based on test data and technical publication
                review, for other model inputs required by Autonomie, such as throttle
                time response and shifting strategies for different transmission
                technologies. The Autonomie modeling tool did not simulate vehicle
                attributes determined to have minimal impacts, like whether a vehicle
                had a sun roof or hood scoops, as those attributes would have trivial
                impact in the overall analysis.
                ---------------------------------------------------------------------------
                 \445\ ANL Autonomie Model Assumptions Summary. Aug 21, 2018,
                NHTSA-2018-0067-0005. ANL--Summary of Main Component Performance and
                Assumptions NPRM. Aug 21, 2018, NHTSA-2018-0067-0003.
                 \446\ See further details in Section VI.B.1 Analysis Fleet.
                ---------------------------------------------------------------------------
                 Because the agencies model ten different vehicle types to represent
                the 2,952 vehicles in the baseline fleet, improper assumptions about an
                advanced technology could lead to errors in estimating effectiveness.
                Autonomie is a sophisticated full-vehicle modeling tool that requires
                extensive technology characteristics based on both physical and
                intangible data, like proprietary software. With a few technologies,
                the agencies did not have publicly available data, but had received
                confidential business information confirming such technologies
                potential availability in the market during the rulemaking time frame.
                For such technologies, including advanced cylinder deactivation, the
                agencies adopted a method in the CAFE model to represent the
                effectiveness of the technology, and did not explicitly simulate the
                technologies in the Autonomie model. For this limited set of
                technologies, the agencies determined that effectiveness could
                reasonably be represented as a fixed value.\447\ Effectiveness values
                for technologies not explicitly simulated in Autonomie are discussed
                further in the individual technology sections of this preamble.
                ---------------------------------------------------------------------------
                 \447\ For final rule, 9 out of 50 plus technologies use fixed
                offset effectiveness values. The total effectiveness of these
                technologies cannot be captured on the 2-cycle test or, like ADEAC,
                they are a new technology where robust data that could be used as an
                input to the technology effectiveness modeling does not yet exist.
                Specifically, these nine technologies are LDB, SAX, EPS, IACC, EFR,
                ADEAC, DSLI, DSLIAD and TURBOAD.
                ---------------------------------------------------------------------------
                 The agencies sought comments on all effectiveness inputs and input
                assumptions, including the specific data used to characterize the
                technologies,
                [[Page 24322]]
                such as data to build the technology input, data representing operating
                range of technologies, and data for variation among technology inputs.
                The agencies also sought comment on the effectiveness values used for
                technologies not explicitly defined in Autonomie.
                 Meszler Engineering Services, commenting on behalf of the Natural
                Resources Defense Council, and ICCT questioned the accuracy of the
                effectiveness estimates in the Argonne database, and as an example
                Meszler analyzed the fuel economy impacts of a 10-speed automatic
                transmission relative to a baseline 8-speed automatic transmission,
                concluding that the widely ranging effectiveness estimates were
                unexpected. ICCT questioned the accuracy of the IAV engine maps that
                serve as an input to the Autonomie effectiveness modeling, and asked
                whether those could ``reasonably stand as a foundation for automotive
                developments and technology combinations'' discussed elsewhere in their
                comments. ICCT also questioned whether Autonomie realistically and
                validly modeled synergies between technologies, using the effectiveness
                values from CEGR and transmissions as an example. Meszler stated that
                the agencies have an obligation to validate the Autonomie estimates
                before using them to support the NPRM or any other rulemaking. The
                agencies also received comments on the specific effectiveness estimates
                generated by Autonomie; however, those comments will be discussed in
                each individual technology section, below.
                 Despite these criticisms, Meszler stated that the critiques of the
                Autonomie technology database were not meant to imply that the
                Autonomie vehicle simulation model used to develop the database was
                fundamentally flawed, or that the model could not be used to derive
                accurate fuel economy impact estimates. Meszler noted that, as with any
                model, estimates derived with Autonomie are only valid for a given set
                of modeling parameters and if those parameters are well defined, the
                estimates should be accurate and reliable. Conversely, if those
                parameters are not well defined, the estimates would be inaccurate and
                unreliable. Meszler stated that the agencies must make the full set of
                modeling assumptions used for the Autonomie database available for
                review and comment.
                 We agree with Meszler that, in general, when inputs to a model are
                inaccurate, output effectiveness results may be too high or too low.
                The technology effectiveness estimates from modeling results often vary
                with the type of vehicle and the other technologies that are on that
                vehicle.\448\ The Autonomie output database consists of permutations of
                over 50 technologies for each of the ten technology classes simulated
                by the CAFE model. A wide range of effectiveness is expected when going
                from a baseline technology to an advanced technology across different
                technology classes because there are significant differences in how
                much power is required from the powertrain during 2-cycle testing
                across the ten vehicle types. This impacts powertrain operating
                conditions (e.g., engine speed and load) during 2-cycle testing. Fuel
                economy improving technologies have different effectiveness at each of
                those operating conditions so vehicles that have higher average power
                demands will have different effectiveness than vehicles with lower
                average power demands. Further, the differences in effectiveness at
                higher power and lower power vary by technology so the overall
                relationship is complex. Large-scale full-vehicle modeling and
                simulation account for these interactions and complexities.
                ---------------------------------------------------------------------------
                 \448\ The PRIA Chapter 6.2.2.1, Table 6-2 and Table 6-3 defined
                the characteristics of the reference technology classes that
                representative of the analysis fleet.
                ---------------------------------------------------------------------------
                 Before conducting any full-vehicle modeling and simulation, the
                agencies spent a considerable amount of time and effort developing the
                specific inputs used for the Autonomie analysis. The agencies believe
                that these technology inputs provide reasonable estimates for the
                light-duty vehicle technologies the agencies expect to be available in
                the market in the rulemaking timeframe. As discussed earlier, these
                inputs vary in effectiveness due to how different vehicles, like
                compact cars and pickup trucks, operate on the 2-cycle test and in the
                real world. Some technologies, such as 10-speed automatic transmissions
                (AT10) relative to 8-speed automatic transmissions (AT8), can and
                should have different effectiveness results in the analysis between two
                different technology classes.\449\ These unique synergistic effects can
                only be taken into account through conducting full-vehicle modeling and
                simulation, which the agencies did here.
                ---------------------------------------------------------------------------
                 \449\ Separately, the agencies modified specific transmission
                modeling parameters for the final rule after additional review,
                including a thorough review of public comments, and this review is
                discussed in detail in Section VI.C.2.
                ---------------------------------------------------------------------------
                 With regards to Meszler's comment that the agencies have an
                obligation to validate the Autonomie estimates before using them to
                support the NPRM or any other rulemaking, the agencies would like to
                point Meszler to the description of the Argonne Autonomie team's robust
                process for vehicle model validation that was contained in the
                PRIA.\450\ To summarize, the NPRM and final rule analysis leveraged
                extensive vehicle test data collected by Argonne National
                Laboratory.\451\ Over the past 20 years, the Argonne team has developed
                specific instrumentation lists and test procedures for collecting
                sufficient information to develop and validate full vehicle models. In
                addition, the agencies described the Argonne team's efforts to validate
                specific component models as well, such as the advanced automatic
                transmission and dual clutch transmission models.\452\
                ---------------------------------------------------------------------------
                 \450\ PRIA at 216-7. See also N. Kim, A. Rousseau, E. Rask,
                ``Autonomie Model Validation with Test Data for 2010 Toyota Prius,''
                SAE 2012-01-1040, SAE World Congress, Detroit, Apr12. https://www.autonomie.net/docs/5%20-%20Presentations/Validation/SAE%202012-01-1040.pdf; Vehicle Validation Status, February 2010 https://www.autonomie.net/docs/5%20-%20Presentations/Validation/vehicle_validation_status.pdf; Tahoe HEV Model Development in PSAT,
                SAE paper 2009-01-1307, April 2009 https://www.autonomie.net/docs/5%20-%20Presentations/Validation/tahoe_hev.pdf; PHEV Model
                Validation, U.S.DOE Merit Review 2008 https://www.autonomie.net/docs/5%20-%20Presentations/Validation/phev_model_validation.pdf ;
                PHEV HyMotion Prius model validation and control improvements, 23rd
                International Electric Vehicle Symposium (EVS23), Dec. 2007 https://www.autonomie.net/docs/5%20-%20Presentations/Validation/phev_hymotion_prius.pdf; Integrating Data, Performing Quality
                Assurance, and Validating the Vehicle Model for the 2004 Prius Using
                PSAT, SAE paper 2006-01-0667, April 2006; https://www.autonomie.net/docs/5%20-%20Presentations/Validation/integrating_data.pdf.
                 \451\ A list of the vehicles that have been tested at the APRF
                can be found under http://www.anl.gov/energy-systems/group/downloadable-dynamometer-database.
                 \452\ Kim, N., Rousseau, N., Lohse-Bush, H. ``Advanced Automatic
                Transmission Model Validation Using Dynamometer Test Data,'' SAE
                2014-01-1778, SAE World Congress, Detroit, April 2014; Kim, N.,
                Lohse-Bush, H., Rousseau, A. ``Development of a model of the dual
                clutch transmission in Autonomie and validation with dynamometer
                test data,'' International Journal of Automotive Technologies, March
                2014, Volume 15, Issue 2, pp 263-71.
                ---------------------------------------------------------------------------
                 The agencies also described the process for validating inputs used
                to develop the IAV engine maps,453 454 another input to the
                Autonomie simulations. As discussed in the PRIA, IAV's engine model
                development relied on a collection of sub-models that controlled
                independent combustion characteristics such as heat release, combustion
                knock, friction, heat flow, and other combustion optimization tools.
                These sub-models and other
                [[Page 24323]]
                computational fluid dynamics models were utilized to convert test data
                for use in the IAV engine map development. Specific combustion
                parameters, like from test data for the coefficient of variation for
                the indicated mean effective pressure (COV of IMEP), which is a common
                variable for combustion stability in a spark ignited engine, was used
                to assure final engine models were reasonable. The assumptions and
                inputs used in the modeling and validation of engine model results
                leveraged IAV's global engine database, which included benchmarking
                data, engine test data, single cylinder test data and prior modeling
                studies, and also technical publications and information presented at
                conferences. The agencies referenced in the PRIA that engine maps were
                validated with engine dynamometer test data to the maximum extent
                possible.\455\ Because the NPRM and the final rule analysis considered
                some technologies not yet in production, the agencies relied on
                technical publications and engine modeling by IAV to develop and
                corroborate inputs and input assumptions where engine dynamometer test
                data was not available.
                ---------------------------------------------------------------------------
                 \453\ See PRIA at 251.
                 \454\ See IAV material submitted to the docket; IAV_20190430_Eng
                22-26 Updated_Docket.pdf,
                IAV_Engine_tech_study_Sept_2016_Docket.pdf, IAV_Study for 4 Cylinder
                Gas Engines_Docket.pdf.
                 \455\ See PRIA at 288.
                ---------------------------------------------------------------------------
                 In addition, as described earlier in this section, the full set of
                NPRM modeling assumptions used for the Autonomie database were
                available for review and comment in the docket for this
                rulemaking.\456\ The full set of modeling assumptions used for the
                final rule are also available in the docket.\457\
                ---------------------------------------------------------------------------
                 \456\ NHTSA-2018-0067-0007. Islam, E., S, Moawad, A., Kim, N,
                Rousseau, A., ``A Detailed Vehicle Simulation Process To Support
                CAFE Standards 04262018--Report'' ANL Autonomie Documentation. Aug
                21, 2018. NHTSA-2018-0067-0004. ANL Autonomie Data Dictionary. Aug
                21, 2018. NHTSA-2018-0067-0003. ANL Autonomie Summary of Main
                Component Assumptions. Aug 21, 2018. NHTSA-2018-0067-0005. ANL
                Autonomie Model Assumptions Summary. Aug 21, 2018. NHTSA-2018-0067-
                1692. ANL BatPac Model 12 55. Aug 21, 2018. Preliminary Regulatory
                Impact Analysis (July 2018). Posted July 2018 and updated August 23
                and October 16, 2018.
                 \457\ The CAFE Model is available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system with documentation and all inputs and outputs supporting
                today's notice.
                ---------------------------------------------------------------------------
                 Both ICCT and Meszler also commented on the availability of
                technologies within the Autonomie database, with Meszler stating that
                with limited exceptions, technologies were not included in the NPRM
                CAFE model if they were not included in the simulation modeling that
                underlay the Argonne database, and accordingly if a combination of
                technologies was not modeled during the development of the Argonne
                database, that package (or combination) of technologies was not
                available for adoption in the CAFE model. Meszler stated that these
                constraints limited the slate of technologies available to respond to
                fuel economy standards, and independently expanding the model to
                include additional technologies or technology combinations is not
                trivial.
                 ICCT gave specific examples of key efficiency technologies that it
                stated Autonomie did not include, like advanced DEAC, VCR, Miller
                Cycle, e-boost, and HCCI. ICCT argued that this was especially
                problematic as the agencies appeared to have available engine maps from
                IAV on advanced DEAC, VCR, Miller Cycle, E-boost (and from advanced
                DEAC, VCR, Miller Cycle, E-boost, HCCI from EPA) that Argonne or the
                agencies have been unable to or opted not to include in their modeling.
                ICCT stated that the agencies must disclose how Autonomie had been
                updated to incorporate ``cutting edge'' 2020-2025 automotive
                technologies to ensure they reflect available improvements.\458\
                ---------------------------------------------------------------------------
                 \458\ ICCT also made the same request of EPA's ALPHA model, and
                the agencies' response to that comment is discussed in Section
                VI.C.1 Engine Paths, below.
                ---------------------------------------------------------------------------
                 The agencies have updated the final rule analysis to include
                additional technologies. In the NPRM, the agencies presented the engine
                maps for all of the technologies that ICCT listed, except HCCI, and
                sought comment on the engine maps, technical assumptions and the
                potential use of the technologies for the final rule analysis. Based on
                the available technical information and the ICCT and Meszler comments,
                for the final rule analysis, VCR, Miller Cycle (VTG), and e-boost (VTGe
                with 48V BISG) technologies have been added and included in the
                Autonomie modeling and simulations, and advanced DEAC technology has
                been added using fixed point effectiveness estimates in the CAFE model
                analysis. The agencies disagree with ICCT's assessment of HCCI and do
                not believe it will be available for wide-scale application in the
                rulemaking timeframe, and therefore have not included it as a
                technology. HCCI technology has been in the research phase for several
                decades, and the only production applications to date use a highly-
                limited version that restricts HCCI combustion to a very narrow range
                of engine operating conditions.459 460 461 Additional
                discussion of how Autonomie-modeled and non-modeled technologies are
                incorporated into the CAFE Model is located in Section VI.B.3.c),
                below.
                ---------------------------------------------------------------------------
                 \459\ Mazda introduced Skyactiv-X in Europe with a mild hybrid
                technology to assist the engine.
                 \460\ Mazda News. ``Revolutionary Mazda Skyactiv-X engine
                details confirmed as sales start,'' May 6, 2019. https://www.mazda-press.com/eu/news/2019/revolutionary-mazda-skyactiv-x-engine-details-confirmed-as-sales-start/. Last accessed Dec. 2, 2019.
                 \461\ Confer. K. Kirwan, J. ``Ultra Efficient Light-Duty
                Powertrain with Gasoline Low-Temperature Combustion.'' DOE Merit
                Review. June 9, 2017. https://www.energy.gov/sites/prod/files/2017/06/f34/acs094_confer_2017_o.pdf. Last accessed Dec. 2, 2019.
                ---------------------------------------------------------------------------
                 ICCT and Meszler also commented that the agencies overly limited
                the availability of several technologies in the NPRM analysis. In
                response, the agencies reconsidered the restrictions that were applied
                in the NPRM analysis, and agree with the commenters for several
                technologies and technology classes. Many technologies identified by
                the commenters are now in production for the MY2017 as well as MY2018
                and MY2019. The agencies also think that the baseline fleet compliance
                data reflects adoption of many of these technologies. For the final
                rule analysis, the agencies have expanded the availability of several
                technologies. In the CAFE model, the agencies are now allowing parallel
                hybrids (SHEVP2) to be adopted with high compression Atkinson mode
                engines (HCR0 and HCR1). In addition, as mentioned above, the Autonomie
                full-vehicle modeling included Variable Compression Ratio engine (VCR),
                Miller Cycle Engine (VTG), E-boost (VTGe) technologies, and cylinder
                deactivation technologies (DEAC) to be applied to turbocharged engines
                (TURBO1). As these changes relate to the technology effectiveness
                modeling, the CAFE model analysis now includes effectiveness estimates
                based on full vehicle simulations for all of these technology
                combinations.
                 We disagree with comments stating the agencies should allow every
                technology to be available to every vehicle class.\462\ Discussed
                earlier in this section, Autonomie models key aspects of vehicle
                operation that are most relevant to assessing fuel economy, vehicle
                performance and certain aspects of drivability (like EPA 2-cycle tests,
                EPA US06 cycle tests, gradability, low speed acceleration time from 0-
                to-60 mph, passing acceleration time from 50 to 80 mph, and number of
                transmission shifts). However, there are other critical aspects of
                vehicle functionality and operation that the agencies considered beyond
                those criteria, that cannot necessarily be reflected in the Autonomie
                modeling. For example, a pickup truck can be modeled with a
                [[Page 24324]]
                continuously variable transmission (CVT) and show improvements on the
                2-cycle tests. However, pickup trucks are designed to provide high load
                towing utility.\463\ CVTs lack the torque levels needed to provide that
                towing utility, and would fail mechanically if subject to high load
                towing.\464\ The agencies provided discussions of some of these
                technical considerations in the PRIA, and explained why the agencies
                had limited technologies for certain vehicle classes, such as limiting
                CVTs on pickups as in the example above. These and other limitations
                are discussed further in the individual technology sections.
                ---------------------------------------------------------------------------
                 \462\ NHTSA-2018-0067-11723. NRDC Attachment2 at p. 4.
                 \463\ SAE J2807. ``Performance Requirements for Determining Tow-
                Vehicle Gross Combination Weight Rating and Trailer Weight Rating.''
                Feb. 4, 2016.
                 \464\ PRIA at p. 223 and 340.
                ---------------------------------------------------------------------------
                 The agencies also received a variety of comments that conflated
                aspects of the Autonomie models with technology inputs and input
                assumptions. For example, commenters expressed concern about the
                transmission gear set and final drive values used for the NPRM
                analysis, or more specifically, that the gear ratios were held constant
                across applications.\465\ In this case, both the inputs (gear set and
                final drive ratio) and input assumption (ratios held constant) were
                discussed by the commenters. Because these comments are actually about
                technology inputs to the Autonomie model, for these and similar cases,
                the agencies are addressing the comments in the individual technology
                sections which discuss the technology inputs and input assumptions that
                impact the effectiveness values for those technologies.
                ---------------------------------------------------------------------------
                 \465\ NHTSA-2018-0067-11873. Comments from Roush Industries,
                Attachment 1, at p. 14-15. NHTSA-2018-0067-11873. Comments from
                CARB, at p.110.
                ---------------------------------------------------------------------------
                 For the NPRM analysis, the agencies prioritized using inputs that
                were based on data for identifiable technology configurations and that
                reflected practical real world constraints. The agencies provided
                detailed information on the NPRM analysis inputs and input assumptions
                in the NPRM Preamble, PRIA and Argonne model documentation for engine
                technologies, transmission technologies, powertrain electrification,
                light-weighting, aerodynamic improvements, tire rolling resistance
                improvements, and other vehicle technologies. Comments and the
                agencies' assessment of comments for each technology are discussed in
                the individual technology sections below. Through careful consideration
                of the comments, the agencies have updated analytical inputs associated
                with several technologies, and as discussed above, have included
                several advanced technologies for which technical information was
                included in the NPRM. However, for most technologies, the agencies have
                determined that the technology inputs and input assumptions that were
                used in the NPRM analysis remain reasonable and the best available for
                the final rule analysis.
                (2) How The Agencies Defined Different Vehicle Types in Autonomie
                 As described in the NPRM, Argonne produced full-vehicle models and
                ran simulations for many combinations of technologies, on many types of
                vehicles, but it did not simulate literally every single vehicle model/
                configuration in the analysis fleet because it would be impractical to
                assemble the requisite detailed information--much of which would likely
                only be provided on a confidential basis--specific to each vehicle
                model/configuration and because the scale of the simulation effort
                would correspondingly increase by orders of magnitude. Instead, Argonne
                simulated 10 different vehicle types, corresponding to the five
                ``technology classes'' generally used in CAFE analysis over the past
                several rulemakings, each with two performance levels and corresponding
                vehicle technical specifications (e.g., small car, small performance
                car, pickup truck, performance pickup truck, etc.).
                 Technology classes are a means of specifying common technology
                input assumptions for vehicles that share similar characteristics.
                Because each vehicle technology class has unique characteristics, the
                effectiveness of technologies and combinations of technologies is
                different for each technology class. Conducting Autonomie simulations
                uniquely for each technology class provides a specific set of
                simulations and effectiveness data for each technology class. Like the
                Draft TAR analysis, there are separate technology classes for compact
                cars, midsize cars, small SUVs, large SUVs, and pickup trucks. However,
                new for the NPRM analysis and carried into this final rule analysis,
                each of those vehicle types has been split into ``low'' (or
                ``standard'') performance and a ``high'' performance versions, which
                represent two classes with similar body styles but different levels of
                performance attributes (for a total of 10 technology classes). The
                separate technology classes for high performance and low performance
                vehicles better account for performance diversity across the fleet.
                 NHTSA directed Argonne to develop a vehicle assumptions database to
                capture vehicle attributes that would comprise the full vehicle models.
                For each vehicle technology class, representative vehicle attributes
                and characteristics were identified from publicly available information
                and automotive benchmarking databases like A2Mac1,\466\ Argonne's
                Downloadable Dynamometer Database (D\3\),\467\ and EPA compliance and
                fuel economy data,\468\ EPA's guidance on the cold start penalty on 2-
                cycle tests.\469\ The resulting vehicle assumptions database consists
                of over 100 different attributes like vehicle frontal area, drag
                coefficient, fuel tank weight, transmission housing weight,
                transmission clutch weight, hybrid vehicle component weights, and
                weights for components that comprise engines and electric machines,
                tire rolling resistance, transmission gear ratios and final drive
                ratio. Each of the 10 different vehicle types was assigned a set of
                these baseline attributes and characteristics, to which combinations of
                fuel-saving technologies were added as inputs for the Autonomie
                simulations. For example, the characteristics of the MY 2016 Honda Fit
                were considered along with a wide range of other compact cars to
                identify representative characteristics for the Autonomie simulations
                for the base compact car technology class. The simulations determined
                the fuel economy achieved when applying each combination of
                technologies to that vehicle type, given its baseline characteristics.
                ---------------------------------------------------------------------------
                 \466\ A2Mac1: Automotive Benchmarking. (Proprietary data).
                Retrieved from https://a2mac1.com.
                 \467\ Downloadable Dynamometer Database (D\3\). ANL Energy
                Systems Division. https://www.anl.gov/es/downloadable-dynamometer-database. Last accessed Oct. 31, 2019.
                 \468\ Data on Cars used for Testing Fuel Economy. EPA Compliance
                and Fuel Economy Data. https://www.epa.gov/compliance-and-fuel-economy-data/data-cars-used-testing-fuel-economy. Last accessed Oct.
                31, 2019.
                 \469\ EPA PD TSD at p.2-265--2-266.
                ---------------------------------------------------------------------------
                 For each vehicle technology class and for each vehicle attribute,
                Argonne estimated the attribute value using statistical distribution
                analysis of publicly available data and data obtained from the A2Mac1
                benchmarking database.\470\ Some
                [[Page 24325]]
                vehicle attributes were also based on test data and vehicle
                benchmarking, like the cold-start penalty for the FTP test cycle and
                vehicle electrical accessories load. The analysis of vehicle attributes
                used in the NPRM was discussed in the Argonne model documentation,\471\
                and values for each vehicle technology class were provided with the
                NPRM for public review.\472\
                ---------------------------------------------------------------------------
                 \470\ A2Mac1 is subscription-based benchmarking service that
                conducts vehicle and component teardown analyses. Annually, A2Mac1
                removes individual components from production vehicles such as oil
                pans, electric machines, engines, transmissions, among the many
                other components. These components are weighed and documented for
                key specifications which is then available to their subscribers.
                 \471\ NHTSA-2018-0067-0007, at 131. Islam, E., S, Moawad, A.,
                Kim, N, Rousseau, A., ``A Detailed Vehicle Simulation Process To
                Support CAFE Standards 04262018--Report'' ANL Autonomie
                Documentation. Aug 21, 2018.
                 \472\ NHTSA-2018-0067-0003. ANL Autonomie Summary of Main
                Component Assumptions. Aug 21, 2018.
                ---------------------------------------------------------------------------
                 The agencies did not believe it was appropriate to assign one
                single engine mass for each vehicle technology class in the NPRM
                analysis. To account for the difference in weight for different engine
                types, Argonne performed a regression analysis of engine peak power
                versus weight, based on attribute data taken from the A2Mac1
                benchmarking database. For example, to account for weight of different
                engine sizes like 4-cylinder versus 8-cylinder, Argonne developed a
                relationship curve between peak power and engine weight based on the
                A2Mac1 benchmarking data. For the NPRM analysis, this relationship was
                used to estimate mass for all engine types regardless of technology
                type (e.g., variable valve lift and direct injection). Secondary weight
                reduction associated with changes in engine technology was applied by
                using this linear relationship between engine power and engine weight
                from the A2Mac1 benchmarking database. When a vehicle in the analysis
                fleet with an 8-cylinder engine adopted a more fuel efficient 6-
                cylinder engine, the total vehicle weight would reflect the updated
                engine weight with two less cylinders based on the peak power versus
                engine weight relationship. The impact of engine mass reduction on
                effectiveness is accounted for directly in the Autonomie simulation
                data through the application of the above relationship. Engine mass
                reduction through downsizing is, therefore, appropriately not included
                as part of vehicle mass reduction technology that is discussed in
                Section VI.C.4 because doing so would result in double counting the
                impacts. As discussed further below, for the final rule the agencies
                improved upon the precision of engine weights by creating two curves to
                separately represent naturally aspirated engine designs and
                turbocharged engine designs.
                 In addition, certain attributes were held at constant levels within
                each technology class to maintain vehicle functionality, performance
                and utility including noise, vibration, and harshness (NVH), safety,
                performance and other utilities important for customer satisfaction.
                For example, in addition to the vehicle performance constraints
                discussed in Section VI.B.3.a)(6), the analysis does not allow the
                frontal area of the vehicle to change, in order to maintain utility
                like ground clearance, head-room space, and cargo space, and a cold-
                start penalty is used to account for fuel economy degradation for
                heater performance and emissions system catalyst light-off.\473\ This
                allows us to capture the discrete improvement in technology
                effectiveness while maintaining vehicle attributes that are important
                vehicle utility, consumer acceptance and compliance with criteria
                emission standards, and considering these constraints similar to how
                manufacturers do in the real world.
                ---------------------------------------------------------------------------
                 \473\ The catalyst light-off is the temperature necessary to
                initiate the catalytic reaction and this energy is generated from
                engine.
                ---------------------------------------------------------------------------
                 The agencies sought comment on the analytical approach used to
                determine vehicle attributes and characteristics for the Autonomie
                modeling. In response, the agencies received a wide variety of comments
                on vehicle attributes ranging from discussions of performance increase
                from technology adoption (e.g., if a vehicle adopting an electrified
                powertrain improved its time to accelerate from 0-60 mph), to comments
                on vehicle attributes not modeled in Autonomie, like heated seats and
                cargo space.
                 Toyota and the Alliance commented that the inclusion of performance
                vehicle classes addressed the market reality that some consumers will
                purchase vehicles for their performance attributes and will accept the
                corresponding reduction in fuel economy. Furthermore, Toyota commented
                that some gain in performance is more realistic, and that ``dedicating
                all powertrain improvements to fuel efficiency is inconsistent with
                market reality.'' Toyota ``supports the agencies' inclusion of
                performance classes in compliance modeling where a subset of certain
                models is defined to have higher performance and a commensurate
                reduction in fuel efficiency.'' \474\ Also, in support of the addition
                of performance vehicle classes, the Alliance commented that ``vehicle
                categories have been increased to 10 to better recognize the range of
                0-60 performance characteristics within each of the 5 previous
                categories, in recognition of the fact that many vehicles in the
                baseline fleet significantly exceeded the previously assumed 0-60
                performance metrics. This provides better resolution of the baseline
                fleet and more accurate estimates of the benefits of technology.''
                \475\
                ---------------------------------------------------------------------------
                 \474\ Toyota, Attachment 1, Docket No. NHTSA-2018-0067-12098, at
                p. 6.
                 \475\ Alliance of Automobile Manufacturers, Attachment ``Full
                Comment Set,'' Docket No. NHTSA-2018-0067-12073, at p.135.
                ---------------------------------------------------------------------------
                 UCS commented that the CAFE model incorporates technology
                improvements to each vehicle by applying the effectiveness improvement
                of the average vehicle in the technology class, leading to discrete
                ``stepped'' effectiveness levels for technologies across the different
                vehicle types. UCS stated that in contrast, the OMEGA model takes into
                account a vehicle's performance characteristics through response-
                surface modeling based on relative deviation from the class average
                modeled in ALPHA.\476\
                ---------------------------------------------------------------------------
                 \476\ NHTSA-2018-0067-12039, at p.24.
                ---------------------------------------------------------------------------
                 Although differences between the ALPHA and Autonomie models are
                discussed in more detail below, for the NPRM vehicle simulation
                analysis the agencies expanded the number of vehicle classes from the
                five classes used in the Draft TAR to ten classes, to represent better
                the diversity of vehicle characteristics across the fleet. Each of
                these ten vehicle technology classes are empirically built from
                benchmarking data and other information from various sources, amounting
                to hundreds of vehicle characteristics data points to develop each
                vehicle class. The agencies expand on these vehicle classes and
                characteristics in Section VI.B.3.(a)(2) Vehicle Types in Autonomie and
                Section VI.B.3.(a)(3) How Vehicle Models are Built in Autonomie and
                Optimized for Simulation. The agencies believe that the real-world data
                used to define vehicle characteristics for each of the ten vehicle
                classes, in addition to the ten vehicle technology classes themselves,
                ensures the analysis reasonably accounts for the diversity in vehicle
                characteristics across the fleet.
                 The agencies believe that UCS's characterization of how technology
                improvements are applied in the analysis is a misleading
                oversimplification. While the analysis approach in the final rule uses
                a representative effectiveness value, the value is not linked solely to
                the vehicle technology class, as the UCS implies. The entire technology
                combination, or technology key, which includes the vehicle technology
                class, is used to
                [[Page 24326]]
                determine the value for the platform being considered. Within each
                vehicle class, the interactions between the added technology and the
                full vehicle system (including other technologies and substantial road
                load characteristics) are considered in the effectiveness values
                calculated for each technology during compliance modeling. As discussed
                under each of the technology pathways sections, the effectiveness for
                most technologies is reported as a range rather than a single value.
                The range exists because the effectiveness for each technology is
                adjusted based on the technologies it is coupled with and the major
                road load characteristics of the full vehicle system. This approach, in
                combination with using the baseline vehicle's initial performance
                values as a starting point for performance improvement, results in a
                widely variable level of improvement for the system, dependent on
                individual vehicle platform characteristics. As a result, the
                application of a response-surface approach would likely result in
                minimal improvement in accuracy for the Autonomie and CAFE model
                analysis approach.
                 For the final rule analysis, the agencies used the same process to
                obtain the vehicle attributes and characteristics for the vehicle
                technology classes. Data was acquired from publicly available sources,
                Argonne D\3\, EPA compliance and fuel economy data, and A2mac1
                benchmarking data. Accordingly, the attributes and characteristics of
                the modeled vehicles reflect actual vehicles that meet customer
                expectations and automakers' capabilities to manufacture the vehicles.
                In addition, for the final rule, the agencies improved the NPRM
                analysis by updating some of the attribute values to account for
                changes in the fleet. For example, the agencies have updated vehicle
                electrical accessory load on the test cycle to reflect higher
                electrical loads associated with contemporary vehicle features.
                (3) How This Rulemaking Builds Vehicle Models for Autonomie and
                Optimize Them for Simulation
                 Before any simulation is initiated in Autonomie, Argonne must
                ``build'' a vehicle by assigning reference technologies and initial
                attributes to the components of the vehicle model representing each
                technology class.\477\ The reference technologies are baseline
                technologies that represent the first step on each technology pathway
                used in the analysis. For example, a compact car is built by assigning
                it a baseline engine, a baseline 6-speed automatic transmission (AT6),
                a baseline level of aerodynamic improvement (AERO0), a baseline level
                of rolling resistance improvement (ROLL0), a baseline level of mass
                reduction technology (MR0), and corresponding attributes from the
                Argonne vehicle assumptions database like individual component
                weights.\478\ A baseline vehicle will have a unique starting point for
                the simulation and a unique set of assigned inputs and attributes,
                based on its technology class.
                ---------------------------------------------------------------------------
                 \477\ For the NPRM analysis, Chapter 8 Vehicle-Sizing Process in
                the ANL Model Documentation had discussed this process in detail.
                Further discussion of this process is located in Chapter 8 of the
                ANL Model Documentation for this final rule.
                 \478\ See Section VI.A.7.
                ---------------------------------------------------------------------------
                 The next step in the process is to run a powertrain sizing
                algorithm that ensures the built vehicle meets or exceeds defined
                performance metrics, including low-speed acceleration (i.e., time
                required to accelerate from 0-60 mph), high-speed passing acceleration
                (time required to accelerate from 50-80 mph), gradeability (e.g. the
                ability of the vehicle to maintain constant 65 miles per hour speed on
                a six percent upgrade), and towing capacity. Together, these
                performance criteria are widely used by industry as metrics to quantify
                vehicle performance attributes that consumers observe and that are
                important for vehicle utility and customer satisfaction.
                 In the compact car example used above, the agencies assigned an
                initial specific engine design and engine power, transmission, AERO,
                ROLL, and MR technologies, and other attributes like vehicle weight. If
                the built vehicle does not meet all the performance criteria in the
                first iteration, then the engine power is increased to meet the
                performance requirement. This increase in power is from higher engine
                displacement, which could involve an increase in number of cylinders,
                leading to an increase in the engine weight. The iterative process
                continues to check whether the compact car with updated engine power,
                and corresponding updated engine weight, meets its defined performance
                metrics. The loop stops once all the metrics are met, and at this
                point, a compact car technology class vehicle model becomes ready for
                simulation. For further discussion of the vehicle performance metrics,
                see Section VI.B.3.(a).
                 Autonomie then adopts a single fuel saving technology to the
                baseline vehicle model, keeping everything else the same except for
                that one technology and the attributes associated with it. For example,
                the model would apply an 8-speed automatic transmission in place of the
                baseline 6-speed automatic transmission, which would lead to either an
                increase or decrease in the total weight of the vehicle based on the
                technology class assumptions. At this point, Autonomie confirms whether
                performance metrics are met for this new vehicle model through the
                previously discussed sizing algorithm. Once a technology has been
                assigned to the vehicle model and the resulting vehicle meets its
                performance metrics, those vehicle models will be used as inputs to the
                full vehicle simulations. So, in the example of the 6-speed to 8-speed
                automatic transmission technology update, the agencies now have the
                initial ten vehicle models (one for each technology class), plus the
                ten new vehicle models with the updated 8-speed automatic transmission,
                which adds up to 20 different vehicle models for simulation. This
                permutation process is conducted for each of the over 50 technologies
                considered, and for all ten technology classes, which results in more
                than one million optimized vehicle models.
                 Figure VI-3 shows the process for building vehicles in Autonomie
                for simulation.
                [[Page 24327]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.110
                 Some of the technologies require extra steps for optimization
                before the vehicle models are built for simulation; for example, the
                sizing and optimization process is more complex for the electrified
                vehicles (i.e., HEVs, PHEVs) compared to vehicles with internal
                combustion engines, as discussed further, below. Throughout the vehicle
                building process, the following items are considered for optimization:
                 Vehicle weight is decreased or increased in response to
                switching from one type of technology to another for the technologies
                for which the agencies consider weight, such as different engine and
                transmission types;
                 Vehicle performance is decreased or increased in response
                to the addition of mass reduction technologies when switching from one
                vehicle model to another vehicle model for the same engine;
                 Vehicle performance is decreased or increased in response
                to the addition of a new technology when switching from one vehicle
                model to another vehicle model for the same hybrid electric machine;
                and
                 Electric vehicle battery size is decreased or increased in
                response to the addition of mass, aero and/or tire rolling resistance
                technologies when switching from one vehicle model to another vehicle
                model.
                 Every time a vehicle adopts a new technology, the vehicle weight is
                updated to reflect the new component weight. For some technologies, the
                direct weight change is easy to assess. For example, in the NPRM the
                agencies designated weights for transmissions so, when a vehicle is
                updated to a higher geared transmission, the weight of the original
                transmission is replaced with the corresponding transmission weight
                (e.g., the weight of a vehicle moving from a 5-speed automatic
                transmission to an 8-speed automatic transmission will be updated based
                on the 8-speed transmission weight).
                 For other technologies, like engine technologies, assessing the
                updated vehicle weight is much more complex. Discussed earlier,
                modeling a change in engine technology involves both the new technology
                adoption and a change in power (because the reduction in vehicle weight
                leads to lower engine loads, and a resized engine). When a new engine
                technology is adopted on a vehicle the agencies account for the
                associated weight change to the vehicle based on the earlier discussed
                regression analysis of weight versus power. For the NPRM engine weight
                regression analysis, the agencies considered 19 different engine
                technologies that consisted of unique components to achieve fuel
                economy improvements. This regression analysis is technology agnostic
                by taking the approach of using engine peak power versus engine weight
                because it removed biases to any specific engine technology in the
                analysis. Although the agencies do not estimate the specific weight for
                each individual engine technology, such as VVT and SGDI, this process
                provides a reasonable estimate of the weight differences among engine
                technologies.
                [[Page 24328]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.111
                 For the final rule analysis, the agencies used the same process to
                assign initial weights to the original 19 engines, plus the added
                engines. However, the agencies improved upon precision of the weights
                by creating two separate curves separately to represent naturally
                aspirated engine designs and turbocharged engine designs.\479\ This
                update resulted in two benefits. First, small naturally aspirated 4-
                cylinder engines that adopted turbocharging technology reflected the
                increased weight of associated components like ducting, clamps, the
                turbocharger itself, a charged air cooler, wiring, fasteners, and a
                modified exhaust manifold. Second, larger cylinder count engines like
                naturally aspirated 8-cylinder and 6-cylinder engines that adopted
                turbocharging and downsized technologies would have lower weight due to
                having fewer engine cylinders. For example, a naturally aspirated 8-
                cylinder engine that adopts turbocharging technology when downsized to
                a 6-cylinder turbocharged engine appropriately reflects the added
                weight of turbocharging components, and the lower weight of fewer
                cylinders.
                ---------------------------------------------------------------------------
                 \479\ ANL Model Documentation for the final rule analysis,
                Chapter 5.2.9 Engine Weight Determination.
                ---------------------------------------------------------------------------
                 As with conventional vehicle models, electrified vehicle models
                were built from the ground up. For the NPRM analysis, Argonne used data
                from the A2mac1 database and vehicle test data to define different
                attributes like weights and power. Argonne used one electric motor
                specific power for each type of hybrid and electric vehicle.\480\ For
                MY2017, the U.S. market has an expanded number of available hybrid and
                electric vehicle models. To capture appropriately the improvements for
                electrified vehicles for the final rule analysis, the agencies applied
                the same regression analysis process that considers electric motor
                weight versus electric motor power for vehicle models that have adopted
                electric motors. Benchmarking data for hybrid and electric vehicles
                from the A2Mac1 database was analyzed to develop a regression curve of
                electric motor peak power versus electric motor weight.\481\
                ---------------------------------------------------------------------------
                 \480\ NHTSA-2018-0067-0005. ANL Autonomie Model Assumptions
                Summary. Aug 21, 2018. Non_Vehicle_Attributes tab. Specific power
                for PS and P2 HEVs was set to 2750 watts/kg, plug-in HEVs were set
                to 375 watts/kg, and electric vehicles were set to 1400 watts/kg.
                 \481\ ANL Model Documentation for the final rule analysis,
                Chapter 5.2.10 Electric Machines System Weight.
                ---------------------------------------------------------------------------
                (4) How Autonomie Sizes Powertrains for Full Vehicle Simulation
                 The agencies maintain performance neutrality of the full vehicle
                simulation analysis by resizing engines, electric machines, and hybrid
                electric vehicle battery packs at specific incremental technology
                steps. To address product complexity and economies of scale, engine
                resizing is limited to specific incremental technology changes that
                would typically be associated with a major vehicle or engine
                redesign.\482\ Manufacturers have repeatedly told the agencies that the
                high costs for redesign and the increased manufacturing complexity that
                would result from resizing engines for small technology changes
                preclude them from doing so. It would be unreasonable and unaffordable
                to resize powertrains for every unique combination of technologies, and
                exceedingly so for every unique combination of technologies across
                every vehicle model due to the extreme manufacturing complexity that
                would be required to do so. The agencies reiterated in the NPRM that
                the analysis should not include engine resizing with the application of
                every technology or for combinations of technologies that drive small
                performance changes so that the analysis better reflects what is
                feasible for manufacturers.\483\
                ---------------------------------------------------------------------------
                 \482\ See 83 FR 43027 (Aug. 24, 2018).
                 \483\ For instance, a vehicle would not get a modestly bigger
                engine if the vehicle comes with floor mats, nor would the vehicle
                get a modestly smaller engine without floor mats. This example
                demonstrates small levels of mass reduction. If manufacturers
                resized engines for small changes, manufacturers would have
                dramatically more part complexity, potentially losing economies of
                scale.
                ---------------------------------------------------------------------------
                 When a powertrain does need to be resized, Autonomie attempts to
                mimic manufacturers' development approaches to the extent possible.
                Discussed earlier, the Autonomie vehicle building process is initiated
                by building a baseline vehicle model with a baseline engine,
                transmission, and other baseline vehicle technologies. This baseline
                vehicle model (for each technology class) is sized to meet a specific
                set of
                [[Page 24329]]
                performance criteria, including acceleration and gradeability.
                 The modeling also accounts for the industry practice of platform,
                engine, and transmission sharing to manage component complexity and the
                associated costs.\484\ At a vehicle refresh cycle, a vehicle may
                inherit an already resized powertrain from another vehicle within the
                same engine-sharing platform that adopted the powertrain in an earlier
                model year. In the Autonomie modeling, when a new vehicle adopts fuel
                saving technologies that are inherited, the engine is not resized (the
                properties from the baseline reference vehicle are used directly and
                unchanged) and there may be a small change in vehicle performance. For
                example, in Figure VI-3, Vehicle 2 inherits Eng01 from Vehicle 1 while
                updating the transmission. Inheritance of the engine with new
                transmission may change performance. This example illustrates how
                manufacturers generally manage manufacturing complexity for engines,
                transmissions, and electrification technologies.
                ---------------------------------------------------------------------------
                 \484\ Ford EcoBoost Engines are shared across ten different
                models in MY2019. https://www.ford.com/powertrains/ecoboost/. Last
                accessed Nov. 05, 2019.
                ---------------------------------------------------------------------------
                 Autonomie implements different powertrain sizing algorithms
                depending on the type of powertrain being considered because different
                types of powertrains contain different components that must be
                optimized.\485\ For example, the conventional powertrain resizing
                considers the reference power of the conventional engine (e.g., Eng01,
                a basic VVT engine, is rated at 108 kilowatts and this is the starting
                reference power for all technology classes) against the power-split
                hybrid (SHEVPS) resizing algorithm that must separately optimize engine
                power, battery size (energy and power), and electric motor power. An
                engine's reference power rating can either increase or decrease
                depending on the architecture, vehicle technology class, and whether it
                includes other advanced technologies.
                ---------------------------------------------------------------------------
                 \485\ ANL Model Documentation for the final rule Analysis,
                Chapter 8.3.1 Conventional-Vehicle Sizing Algorithm; Chapter 8.3.2
                Split-HEV Sizing Algorithm; 8.3.4 Blended PHEV sizing Algorithm;
                8.3.5 Voltec PHEV (Extended Range) Vehicle Sizing Algorithm; Chapter
                8.3.6 BEV Sizing Algorithm.
                ---------------------------------------------------------------------------
                 Performance requirements also differ depending on the type of
                powertrain because vehicles with different powertrain types may need to
                meet different criteria. For example, a plug-in hybrid electric vehicle
                (PHEV) powertrain that is capable of traveling a certain number of
                miles on its battery energy alone (referred to as all-electric range,
                or AER, or as performing in electric-only mode) is also sized to ensure
                that it can meet the performance requirements of a US06 cycle in
                electric-only mode.
                 The powertrain sizing algorithm is an iterative process that
                attempts to optimize individual powertrain components at each step. For
                example, the sizing algorithm for conventional powertrains estimates
                required power to meet gradeability and acceleration performance and
                compares it to the reference engine power for the technology class. If
                the power required to meet gradeability and acceleration performance
                exceeds the reference engine power, the engine power is updated to the
                new value. Similarly, if the reference engine power exceeds the
                gradeability and acceleration performance power, it will be decreased
                to the lower power rating. As the change in power requires a change
                design of the engine, like increasing displacement (e.g., going from a
                5.2-liter to 5.6-liter engine, or vice versa) or increasing cylinder
                count (e.g., going from an I4 to a V6 or vice versa), the engine weight
                will also change. The new engine power is used to update the weight of
                the engine.
                 Next, the conventional powertrain sizing algorithm enters an
                acceleration algorithm loop to verify low-speed acceleration
                performance (time it takes to go from 0 mph to 60 mph). In this step,
                Autonomie adjusts engine power to maintain a performance attribute for
                the given technology class and updates engine weight accordingly. Once
                the performance criteria are met, Autonomie ends the low-speed
                acceleration performance algorithm loop and enters a high-speed
                acceleration (time it takes to go from 50 mph to 80 mph) algorithm
                loop. Again, Autonomie might need to adjust engine power to maintain a
                performance attribute for the given technology, and it exits this loop
                once the performance criteria have been met. At this point, the sizing
                algorithm is complete for the conventional powertrain based on the
                designation for engine type, transmissions type, aero type, mass
                reduction technology and low rolling resistance technology.
                 Figure VI-5 below shows the sizing algorithm for conventional
                powertrains.
                [[Page 24330]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.112
                 Depending on the type of powertrain considered, the sizing
                algorithms may also size to meet different performance criteria in
                different order. The powertrain sizing algorithms for electrified
                vehicles are considerably more complex, and are discussed in further
                detail in Section VI.C.3, below.
                (5) How the Agencies Considered Maintaining Vehicle Attributes
                 For this rulemaking analysis, consistent with past CAFE and
                CO2 rulemakings, the agencies have analyzed technology
                pathways manufacturers could use for compliance that attempt to
                maintain vehicle attributes, utility, and performance. Using this
                approach allows the agencies to assess costs and benefits of potential
                standards under a scenario where consumers continue to get the similar
                vehicle attributes and features, other than changes in fuel economy.
                The purpose of constraining vehicle attributes is to simplify the
                analysis and reduce variance in other attributes that consumers value
                across the analyzed regulatory alternatives. This allows for a more
                streamlined accounting of costs and benefits by not requiring the
                values of other vehicle attributes that trade off with fuel economy.
                 Several examples of vehicle attributes, utility and performance
                that could be impacted by adoption of fuel economy improving technology
                include the following.
                [[Page 24331]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.113
                 Consequences for the agencies not fully considering or accounting
                for potential changes in vehicle attributes, utility, and performance
                are degradation in vehicle attributes, utility, and performance that
                lead to consumer acceptance issues without accounting for the
                corresponding costs and/or not accounting for the costs of technology
                designs that maintain vehicle attributes, utility, and performance. The
                agencies incorporated changes in the NPRM analysis and that are carried
                into this final rule that address deficiencies in past analyses,
                including the Draft TAR and Proposed Determination analyses. These
                changes were discussed in the NPRM and are repeated in the discussion
                of individual technologies in this Preamble, the FRIA, and supporting
                documents. The following are several examples of technologies that did
                not maintain vehicle attributes, utility, and performance in the Draft
                TAR and Proposed Determination analyses.
                 For the EPA Draft TAR and Proposed Determination analyses, HCR
                engine and downsized and turbocharged engine technologies effectiveness
                was estimated using Tier 2 certification fuel, which has a higher
                octane rating compared to regular octane fuel.486 487 This
                does not maintain functionality because consumers would incur higher
                costs for using premium fuel in order to achieve the modeled fuel
                economy improvements, compared to baseline engines that were replaced,
                which operated on lower cost regular octane fuel. By not maintaining
                the fuel octane functionality and vehicle attributes, the EPA Draft TAR
                and Proposed Determination analyses applied higher effectiveness for
                these technologies than could be achieved had regular octane fuel been
                assumed for the HCR and downsized turbocharged engines. The Draft TAR
                and Proposed Determination analyses also did not account for the higher
                costs that would be incurred by consumers to pay for high octane fuel.
                These issues were addressed in the
                [[Page 24332]]
                NPRM and this final rule analysis, and account for some of the
                effectiveness and cost differences between the Draft TAR/Proposed
                Determination and the NPRM/final rule.\488\
                ---------------------------------------------------------------------------
                 \486\ Tier 2 fuel has an octane rating of 93. Typical regular
                grade fuel has an octane rating of 87 ((R+M)/2 octane.
                 \487\ EPA Proposed Determination at 2-209 to 2-212.
                 \488\ For more details, see Section VI.C.1 Engine Paths.
                ---------------------------------------------------------------------------
                 Another example is mass reduction technology. As background, the
                agencies characterize mass reduction as either primary mass reduction
                or secondary mass reduction. Primary mass reduction involves reducing
                mass of components that can be done independently of the mass of other
                components. For example, the mass of a hood (e.g., replacing a steel
                hood with an aluminum hood) or reducing the mass of a seat are examples
                of primary mass reduction because each can be implemented
                independently. When there is a significant level of primary mass
                reduction, other components that are designed based on the mass of
                primary components, may be redesigned and have lower mass. An example
                of secondary mass reduction is the brake system. If the mass of primary
                components is reduced sufficiently, the resulting lighter weight
                vehicle could maintain braking performance and attributes, and safety
                with a lighter weight brake system. Mass reduction in the brake system
                is secondary mass reduction because it requires primary mass reduction
                before it can be incorporated. For the EPA Draft TAR and Proposed
                Determination analyses, secondary mass reduction was applied
                exclusively based on cost, with no regard to whether sufficient primary
                mass reduction was applied concurrently. The analyses did not account
                for the degraded functionality of the secondary components and systems
                and also understated the costs for lower levels of mass reduction.\489\
                These issues were addressed in the NPRM and this final rule analysis,
                and account for some of the cost differences between the Draft TAR/
                Proposed Determination and the NPRM/final rule.
                ---------------------------------------------------------------------------
                 \489\ For more details, see Section VI.C.4 Mass Reduction.
                ---------------------------------------------------------------------------
                 The agencies note that for some technologies it is not reasonable
                or practicable to match exactly the baseline vehicle's attributes,
                utility, and performance. For example, when engines are resized to
                maintain acceleration performance, if the agencies applied a criterion
                that allowed no shift in performance whatsoever, there would be an
                extreme proliferation of unique engine displacements. Manufacturers
                have repeatedly and consistently told the agencies that the high costs
                for redesign and the increased manufacturing complexity that would
                result from resizing engines for small technology changes preclude them
                from doing so. It would be unreasonable and unaffordable to resize
                powertrains for every unique combination of technologies, and
                exceedingly so for every unique combination technologies across every
                vehicle model due to the extreme manufacturing complexity that would be
                required to do so.\490\ For the NPRM and final rule analyses, engine
                resizing is limited to specific incremental technology changes that
                would typically be associated with a major vehicle or engine redesign
                to address product complexity and economies of scale considerations.
                The EPA Draft TAR and Proposed Determination analyses adjusted the
                effectiveness of every technology combination assuming performance
                could be held constant for every combination, and the analysis did not
                recognize or account for the extreme complexity nor the associated
                costs for that impractical assumption. The NPRM and final rule analyses
                account for these real-world practicalities and constraints, and doing
                so explains some of the effectiveness and cost differences between the
                Draft TAR/Proposed Determination and the NPRM/final rule.
                ---------------------------------------------------------------------------
                 \490\ For more details, see Section VI.B.3.a)(6) Performance
                Neutrality.
                ---------------------------------------------------------------------------
                 The subsections for individual technologies discuss the technology
                assumptions and constraints that were considered to maintain vehicle
                attributes, utility, and performance as closely as possible. The
                agencies believe that any minimal remaining differences, which may
                directionally either improve or degrade vehicle attributes, utility and
                performance are small enough to have de minimis impact on the analysis.
                (6) How the Agencies Considered Performance Neutrality
                 The CAFE model examines technologies that can improve fuel economy
                and reduce CO2 emissions. An improvement in efficiency can
                be realized by improving the powertrain that propels the vehicle (e.g.,
                replacing a 6-cylinder engine with a smaller, turbocharged 4-cylinder
                engine), or by reducing the vehicle's loads or burdens (e.g., lowering
                aerodynamic drag, reducing vehicle mass and/or rolling resistance).
                Either way, these changes reduce energy consumption and create a range
                of choices for automobile manufacturers. At the two ends of the range,
                the manufacturer can choose either:
                 (A) To design a vehicle that does same the amount of work as before
                but uses less fuel.
                 For example, a redesigned pickup truck would receive a turbocharged
                V6 engine in place of the outgoing V8. The pickup would offer no
                additional towing capacity, acceleration, larger wheels and tires,
                expanded infotainment packages, or customer convenience features, but
                would achieve a higher fuel economy rating (and correspondingly lower
                CO2 emissions).
                 (B) To design a vehicle that does more work and uses the same
                amount of fuel as before.
                 For example, a redesigned pickup truck would receive a turbocharged
                V6 engine in place of the outgoing V8, but with engine efficiency
                improvements that allow the same amount of fuel to do more work. The
                pickup would offer improved towing capacity, improved acceleration,
                larger wheels and tires, an expanded (heavier) infotainment package,
                and more convenience features, while maintaining (not improving) the
                fuel economy rating of the previous year's model.
                 In other words, automakers weigh the trade-offs between vehicle
                performance/utility and fuel economy, and they choose a blend of these
                attributes to balance meeting fuel economy and emissions standards and
                suiting the demands of their customers.
                 Historically, vehicle performance has improved over the years. The
                average horsepower is the highest that it has ever been; all vehicle
                types have improved horsepower by at least 49 percent compared to the
                1975 model year, and pickup trucks have improved by 141 percent.\491\
                Since 1978, the 0-60 acceleration time of vehicles has improved by 39-
                47 percent depending on vehicle type.\492\ Also, to gain consumer
                acceptance of downsized turbocharged engines, manufacturers have stated
                they often offer an increase in performance.\493\ Fuel economy has also
                improved, but the horsepower and acceleration trends show that not 100
                percent of technological improvements have been applied to fuel
                savings. While future trends are uncertain, the past trends suggest
                vehicle performance is unlikely to decrease, as it seems reasonable to
                assume that customers
                [[Page 24333]]
                will at a minimum demand vehicles that offer the same utility as
                today's fleet.
                ---------------------------------------------------------------------------
                 \491\ The 2018 EPA Automotive Trends Report (EPA-420-R-19-002
                March 2019) https://www.epa.gov/automotive-trends/download-automotive-trends-report.
                 \492\ The 2018 EPA Automotive Trends Report (EPA-420-R-19-002
                March 2019) https://www.epa.gov/automotive-trends/download-automotive-trends-report.
                 \493\ Alliance of Automobile Manufacturers, Attachment
                ``Comment,'' Docket No. EPA-HQ-OAR-2015-0827-4089, at p. 122.
                ---------------------------------------------------------------------------
                 For this rulemaking analysis, consistent with past CAFE and
                CO2 rulemakings, the agencies have analyzed technology
                pathways manufacturers could use for compliance that attempt to
                maintain vehicle attributes, utility and performance. NHTSA's analysis
                in the Draft TAR used the same approach for performance neutrality as
                was used for the NPRM and is being carried into this final rule. This
                approach is described throughout this section and further in FRIA
                Section VI. For the Draft TAR and Proposed Determination, the EPA
                analyses used an approach that maintained 0-60 mph acceleration time
                for every technology package. However, that approach did not account
                for the added development, manufacturing, assembly and service parts
                complexity and associated costs that would be incurred by manufacturers
                to produce the substantial number of engine variants that would be
                required to achieve those CO2 improvements.\494\ Using the
                NPRM approach, which is carried into this final rule, allows the
                agencies to assess costs and benefits of potential standards under a
                scenario where consumers continue to get the same vehicle attributes
                and features, other than changes in fuel economy (approaching the
                scenario in example ``A'' above). This approach also eliminates the
                need to assess the value of changes in vehicle attributes and features.
                As discussed later in this section, while some small level of
                performance increase is unavoidable when conducting this type of
                analysis, the added technology results almost exclusively in improved
                fuel economy. This allows the cost of these technologies to reflect
                almost entirely the cost of compliance with standards with nearly
                neutral vehicle performance.
                ---------------------------------------------------------------------------
                 \494\ Each variant would require a unique engine displacement,
                requiring unique internal engine components, such as crankshaft,
                connecting rods and others.
                ---------------------------------------------------------------------------
                 The CAFE model maintains the initial performance and utility levels
                of the analysis vehicle fleet, while considering real world constraints
                faced by manufacturers.
                 To maintain performance neutrality when applying fuel economy
                technologies, it is first necessary to characterize the performance
                levels of each of the nearly 3000 vehicle models in the MY 2017
                baseline fleet. As discussed in Section VI.B.1.b) Assigning Vehicle
                Technology Classes, above, each individual vehicle model in the
                analysis fleet was assigned to one of ten vehicle ``technology
                classes''--the class that is most similar to the vehicle model. The
                technology classes include five standard class vehicles (compact car,
                midsize car, small SUV, midsize SUV, pickup) plus five ``performance''
                versions of these same body styles.\495\ Each vehicle class has a
                unique set of attributes and characteristics, including vehicle
                performance metrics, that describe the typical characteristics of the
                vehicles in that class.
                ---------------------------------------------------------------------------
                 \495\ Separate technology classes were created for high
                performance and low performance vehicles to better account for
                performance diversity across the fleet.
                ---------------------------------------------------------------------------
                 The analysis used four criteria to characterize vehicle performance
                attributes and utility:
                 Low-speed acceleration (time required to accelerate from 0-60
                mph)
                 High-speed acceleration (time required to accelerate from 50-
                80 mph)
                 Gradeability (the ability of the vehicle to maintain constant
                65 miles per hour speed on a six percent upgrade)
                 Towing capacity
                 Low-speed and high-speed acceleration target times are typical of
                current production vehicles and range from 6 to 10 seconds depending on
                the vehicle class; for example, the midsize SUV performance class has a
                low- and high-speed acceleration target of 7 seconds.\496\ The
                gradeability criterion requires that the vehicle, given its attributes
                of weight, engine power, and transmission gearing, be capable of
                maintaining a minimum of 65 mph while going up a six percent grade. The
                towing criterion, which is applicable only to the pickup truck and
                performance pickup truck vehicle technology classes, is the same as the
                gradeability requirement but adds an additional payload/towing mass
                (3,000 lbs. for pickups, or 4,350 lbs for performance pickups) to the
                vehicle, essentially making the vehicle heavier.
                ---------------------------------------------------------------------------
                 \496\ Note, for all vehicle classes, the low and high-speed
                acceleration targets use the same value. See section VI.B.1.b)(1)
                Assigning Vehicle Technology Classes for a list of low-speed
                acceleration target by vehicle technology class.
                ---------------------------------------------------------------------------
                 In addition, to maintain the capabilities of certain electrified
                vehicles in the 2017 baseline fleet, the analysis required that those
                vehicles be capable of achieving the accelerations and speeds of
                certain standard driving cycles. The agencies use the US06 ``aggressive
                driving'' cycle and the UDDS ``city driving'' cycle to ensure that core
                capabilities of BEVs and PHEVs, such as driving certain speeds and/or
                distances in electric-only mode, are maintained. In addition to the
                four criteria discussed above, the following performance criteria are
                applied to these electrified vehicles:
                 Battery electric vehicles (BEV) are sized to be capable of
                completing the US06 ``aggressive driving'' cycle.
                 Plug-in hybrid vehicles with 50 mile all-electric range
                (PHEV50) are sized to be capable of completing the US06 ``aggressive
                driving'' cycle in electric-only mode.
                 Plug-in hybrid vehicles with 20 mile all-electric range
                (PHEV20) are sized to be capable of completing the UDDS ``city
                driving'' cycle in electric-only (charge depleting) mode.\497\
                ---------------------------------------------------------------------------
                 \497\ PHEV20's are blended-type plug-in hybrid vehicles, which
                are capable of completing the UDDS cycle in charge depleting mode
                without assistance from the engine. However, under higher loads,
                this charge depleting mode may use supplemental power from the
                engine.
                ---------------------------------------------------------------------------
                 Together, these performance criteria are widely used by industry as
                metrics to quantify vehicle performance attributes that consumers
                observe and that are important for vehicle utility and customer
                satisfaction.\498\
                ---------------------------------------------------------------------------
                 \498\ Conlon, B., Blohm, T., Harpster, M., Holmes, A. et al.,
                ``The Next Generation ``Voltec'' Extended Range EV Propulsion
                System,'' SAE Int. J. Alt. Power. 4(2):2015, doi:10.4271/2015-01-
                1152. Kapadia, J., Kok, D., Jennings, M., Kuang, M., et al.,
                ``Powersplit or Parallel--Selecting the Right Hybrid Architecture,''
                SAE Int. J. Alt. Power. 6(1):2017, doi:10.4271/2017-01-1154. Islam,
                E., A. Moawad, N. Kim, and A. Rousseau, 2018a, An Extensive Study on
                Vehicle Sizing, Energy Consumption and Cost of Advance Vehicle
                Technologies, Report No. ANL/ESD-17/17, Argonne National Laboratory,
                Lemont, Ill., Oct 2018.
                ---------------------------------------------------------------------------
                 When certain fuel-saving technologies are applied that affect
                vehicle performance to a significant extent, such as replacing a pickup
                truck's V8 engine with a turbocharged V6 engine, iterative resizing of
                the vehicle powertrain (engine, electric motors, and/or battery) is
                performed in the Autonomie simulation such that the above performance
                criteria is maintained. For example, if the aforementioned engine
                replacement caused an improvement in acceleration, the engine may be
                iteratively resized until vehicle acceleration performance is shifted
                back to the initial target time for that vehicle technology class. For
                the low and high-speed acceleration criteria, engine resizing
                iterations continued until the acceleration time was within plus or
                minus 0.2 seconds of the target time,499 500 which is judged
                to balance
                [[Page 24334]]
                reasonably the precision of engine resizing with the number of
                simulation iterations needed to achieve performance within the 0.2
                second window, and the associated computer resources and time required
                to perform the iterative simulations. Engine resizing is explained
                further in Section VI.B.3.a)(4) How Autonomie Sizes Powertrains for
                Full Vehicle Simulation and the Argonne Model Documentation for the
                final rule analysis.
                ---------------------------------------------------------------------------
                 \499\ For example, if a vehicle has a target 0-60 acceleration
                time of 6 seconds, a time within 5.8-6.2 seconds was accepted.
                 \500\ With the exception of a few performance electrified
                vehicle types which, based on observations in the marketplace, use
                different criteria to maintain vehicle performance without battery
                assist. Performance PHEV20, and Performance PHEV50 resize to the
                performance of a conventional six-speed automatic (CONV 6AU).
                Performance SHEVP2, engines/electric-motors were resized if the 0-60
                acceleration time was worse than the target, but not resized if the
                acceleration time was better than the target time.
                ---------------------------------------------------------------------------
                 The Autonomie simulation resizes until the least capable of the
                performance criteria is met, to ensure the pathways do not degrade any
                of the vehicle performance metrics. It is possible that as one
                criterion target is reached after the application of a specific
                technology or technology package, other criteria may be better than
                their target values. For example, if the engine size is decreased until
                the low speed acceleration target is just met, it is possible that the
                resulting engine size would cause high speed acceleration performance
                to be better than its target.\501\ Or, a PHEV50 may have an electric
                motor and battery appropriately sized to operate in all electric mode
                through the repeated accelerations and high speeds in the US06 driving
                cycle, but the resulting motor and battery size enables the PHEV50
                slightly to over-perform in 0-60 acceleration, which utilizes the power
                of both the electric motor and combustion engine.
                ---------------------------------------------------------------------------
                 \501\ The Autonomie simulation databases include all of the
                estimated performance metrics for each combination of technology as
                modeled.
                ---------------------------------------------------------------------------
                 To address product complexity and economies of scale, engine
                resizing is limited to specific incremental technology changes that
                would typically be associated with a major vehicle or engine
                redesign.\502\ Manufacturers have repeatedly and consistently told the
                agencies that the high costs for redesign and the increased
                manufacturing complexity that would result from resizing engines for
                small technology changes preclude them from doing so. It would be
                unreasonable and unaffordable to resize powertrains for every unique
                combination of technologies, and exceedingly so for every unique
                combination technologies across every vehicle model due to the extreme
                manufacturing complexity that would be required to do so. Engine
                displacements are further described in Section VI.C.1 Engine Paths.
                ---------------------------------------------------------------------------
                 \502\ See 83 FR 43027 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 To address this issue, and consistent with past rulemakings, the
                NPRM simulation allowed engine resizing when mass reductions of 7.1
                percent, 10.7 percent, 14.2 percent (and 20 percent for the final rule
                analysis) were applied to the vehicle curb weight,\503\ and when one
                powertrain architecture was replaced with another architecture during a
                redesign cycle.\504\ At its refresh cycle, a vehicle may also inherit
                an already resized powertrain from another vehicle within the same
                engine-sharing platform. The analysis did not re-size the engine in
                response to adding technologies that have smaller effects on vehicle
                performance. For instance, if a vehicle's curb weight is reduced by 3.6
                percent (MR1), causing the 0-60 mile per hour time to improve slightly,
                the analysis would not resize the engine. The criteria for resizing
                used for the analysis better reflects what is feasible for
                manufacturers to do.\505\
                ---------------------------------------------------------------------------
                 \503\ These correspond, respectively, to reductions of 10%, 15%,
                20%, and 28.2% of the vehicle glider mass. For more detail on glider
                mass calculation, see section VI.C.4 Mass Reduction.
                 \504\ Some engine and accessory technologies may be added to an
                engine without an engine architecture change. For instance,
                manufacturers may adapt, but not replace engine architectures to
                include cylinder deactivation, variable valve lift, belt-integrated
                starter generators, and other basic technologies. However, switching
                from a naturally aspirated engine to a turbo-downsized engine is an
                engine architecture change typically associated with a major
                redesign and radical change in engine displacement.
                 \505\ For instance, a vehicle would not get a modestly bigger
                engine if the vehicle comes with floor mats, nor would the vehicle
                get a modestly smaller engine without floor mats. This example
                demonstrates small levels of mass reduction. If manufacturers
                resized engines for small changes, manufacturers would have
                dramatically more part complexity, potentially losing economies of
                scale.
                ---------------------------------------------------------------------------
                 Automotive manufacturers have commented that the CAFE model's
                consideration of the constraints faced in relation to vehicle
                performance and economies of scale are realistic.
                 Industry associations and individual manufacturers widely supported
                the use of the performance metrics used in the NPRM analysis, the use
                of standard and higher performance technology classes, and the
                representation in the analysis of the real-world manufacturing
                complexity constraints and criteria for powertrain redesign.
                 The Alliance of Automobile Manufacturers (Alliance), Ford, and
                Toyota stated that the inclusion of additional performance metrics such
                as gradeability are appropriate. Specifically in support of the
                gradeability performance criteria, the Alliance commented that
                ``performance metrics related to vehicle operation in top gear are just
                as critical to customer acceptance as are performance metrics such as
                0-60 mph times that focus on performance in low-gear ranges.'' \506\
                The Alliance also commented specifically on the relationship between
                gradeability and downsized engines, stating that as ``engine downsizing
                levels increase, top-gear gradeability becomes more and more
                important,'' and further that the consideration of gradeability ``helps
                prevent the inclusion of small displacement engines that are not
                commercially viable and that would artificially inflate fuel savings.''
                \507\
                ---------------------------------------------------------------------------
                 \506\ Alliance of Automobile Manufacturers, Attachment ``Full
                Comment Set,'' Docket No. NHTSA-2018-0067-12073, at 139.
                 \507\ Alliance of Automobile Manufacturers, Attachment ``Full
                Comment Set,'' Docket No. NHTSA-2018-0067-12073, at 135.
                ---------------------------------------------------------------------------
                 Ford and Toyota similarly commented in support of the CAFE model's
                consideration of multiple performance criteria. Ford stated that this
                model ``takes a more realistic approach to performance modeling'' and
                ``better replicates OEM attribute-balancing practices.'' Ford stated
                furthermore that ``OEMs must ensure that each individual performance
                measure--and not an overall average--meets its customer's
                requirements,'' and that, in contrast, previous analyses did ``not
                align with product planning realities.'' \508\ Toyota commented in
                support of including gradeability as a performance metric ``to avoid
                underpowered engines and overestimated fuel savings.'' \509\
                ---------------------------------------------------------------------------
                 \508\ Ford, Attachment 1, Docket No. NHTSA-2018-0067-11928, at
                8.
                 \509\ Toyota, Attachment 1, Docket No. NHTSA-2018-0067-12098, at
                6.
                ---------------------------------------------------------------------------
                 Toyota and the Alliance commented that the inclusion of performance
                vehicle classes addressed the market reality that some consumers will
                purchase vehicles for their performance attributes and will accept the
                corresponding reduction in fuel economy. Furthermore, Toyota commented
                that most consumers consider more than just fuel economy when
                purchasing a vehicle, and that ``dedicating all powertrain improvements
                to fuel efficiency is inconsistent with market reality.'' Toyota
                ``supports the agencies' inclusion of performance classes in compliance
                modeling where a subset of certain models is defined to have higher
                performance and a commensurate reduction in fuel efficiency.'' \510\
                Also in support of the addition of performance vehicle classes, the
                Alliance commented that ``vehicle categories have been increased to 10
                to better recognize the range of 0-60 performance
                [[Page 24335]]
                characteristics within each of the 5 previous categories, in
                recognition of the fact that many vehicles in the baseline fleet
                significantly exceeded the previously assumed 0-60 performance metrics.
                This provides better resolution of the baseline fleet and more accurate
                estimates of the benefits of technology.'' \511\
                ---------------------------------------------------------------------------
                 \510\ Toyota, Attachment 1, Docket No. NHTSA-2018-0067-12098, at
                6.
                 \511\ Alliance of Automobile Manufacturers, Attachment ``Full
                Comment Set,'' Docket No. NHTSA-2018-0067-12073, at 135.
                ---------------------------------------------------------------------------
                 Toyota also commented in support of various real-world
                manufacturing complexity constraints employed in the analysis for
                powertrain redesigns. Toyota commented that model parameters such as
                redesign cycles and engine sharing across vehicle models place a more
                realistic limit on the number of engines and transmissions that a
                manufacturer is capable of introducing. Toyota also commented in
                support of the constraints that the CAFE model placed on engine
                resizing, stating that ``there are now more realistic limits placed on
                the number of engines and transmissions in a powertrain portfolio which
                better recognizes [how] manufacturers must manage limited engineering
                resources and control supplier, production, and service costs.
                Technology sharing and inheritance between vehicle models tends to
                limit the rate of improvement in a manufacturer's fleet.'' Toyota
                pointed out that this is in contrast to previous analyses in which
                resizing was too unconstrained, which created an ``unmanageable number
                of engine configurations within a vehicle platform'' and spawned cases
                where ``engine downsizing and power reduction sometimes exceeded limits
                beyond basic acceleration requirements needed for vehicle safety and
                customer satisfaction.'' \512\
                ---------------------------------------------------------------------------
                 \512\ Toyota, Attachment 1, Docket No. NHTSA-2018-0067-12098, at
                6.
                ---------------------------------------------------------------------------
                 The above comments from the Alliance, Ford, and Toyota support the
                methodologies the agencies employed to conduct a performance neutral
                analysis. These methodologies helped to ensure that multiple
                performance criteria, including gradeability, are all individually
                accounted for and maintained when a vehicle powertrain is resized, and
                that real-world manufacturing complexity constraints are factored in to
                the agencies' analysis of feasible pathways manufacturers could take to
                achieve compliance with CAFE standards. The agencies continue to
                believe this is a reasonable approach for the aforementioned reasons.
                 Environmental advocacy groups and CARB criticized the CAFE model's
                engine resizing constraints and how they affected the acceleration
                performance criteria.
                 CARB, The International Council on Clean Transportation (ICCT), the
                Union of Concerned Scientists (UCS), and the American Council for an
                Energy-Efficient Economy (ACEEE) commented that the CAFE model was not
                performance neutral, allowing an improvement in performance which
                reduced the effectiveness of applied fuel-saving technologies and/or
                increased the cost of compliance. Specifically, ACEEE stated that there
                appeared to be a shortfall in the fuel economy effectiveness of
                technology packages, potentially resulting from the effectiveness being
                ``consumed'' by additional vehicle performance rather than improvement
                of fuel economy. Several of these same commenters conducted analyses
                attempting to quantify the magnitude of these changes in vehicle
                performance for various vehicle technology classes.
                 CARB commented on the performance shift of several vehicle types.
                Analyzing the 0-60 acceleration for the medium car non-performance
                technology class and looking at all cases with resized engines, CARB
                claimed that ``effectively half of the simulations resulted in improved
                performance.'' \513\ Focusing on electrified vehicles in that same
                technology class, CARB stated that ``the data from the Argonne
                simulations shows that 76 of the 88 strong electrified packages
                (including P2HPV, SHEVPS, BEV, FCEV, PHEV), where Argonne purposely
                resized the system to maintain performance neutrality, resulted in
                notably faster 0 to 60 mph acceleration times and passing times.''
                Specifically regarding parallel hybrid electric vehicles (SHEVP2), CARB
                stated that all modeled packages resulted in improved performance.\514\
                UCS commented that the NPRM analysis allowed too much change in vehicle
                performance, stating that ``while some performance creep may be
                reasonable'' many performance values show ``an overlap between
                performance and non-performance vehicles'' within the compact car
                technology class.\515\
                ---------------------------------------------------------------------------
                 \513\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 180. Note that the target acceleration
                time for medium car non-performance is in fact 9.0 seconds, as
                indicated in ANL documentation, but was incorrectly reported as 9.4s
                in NPRM table II-7 in the NPRM.
                 \514\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 186.
                 \515\ Union of Concerned Scientists, Attachment 2, Docket No.
                NHTSA-2018-0067-12039, at 24.
                ---------------------------------------------------------------------------
                 The agencies carefully considered these comments. For the NPRM
                analysis, the SHEVP2 engines/electric-motors were resized if the 0-60
                acceleration time was worse than the target, but not resized if the
                acceleration time was better than the target. This approach maintained
                vehicle performance with a depleted battery (without electric assist)
                in order to maintain fully the performance and utility characteristics
                under all conditions, and improved performance when electric assist was
                available (when the battery is not depleted), such as during the 0-60
                mph acceleration. The agencies found that this resulted in some
                parallel hybrid vehicles having improved 0-60 acceleration times. This
                approach was initially chosen for the NPRM because the resulting level
                of improved performance was consistent with observations of how
                industry had applied SHEVP2 technology. However, in assessing the CARB
                comment, the agencies balanced the NPRM approach for SHEVP2 performance
                with the agencies' criteria of maintaining vehicle functionality and
                performance when technology is applied. Both could not be fully
                achieved under all conditions for the case of the SHEVP2.
                 The agencies concluded it is reasonable to maintain performance
                including electric assist when SHEVP2 technology is applied to a
                standard (non-performance) vehicle, and therefore the analysis for the
                final rule allows upsizing and downsizing of the parallel hybrid
                powertrain (SHEVP2) using the 0.2 seconds window around the
                target.\516\ For performance vehicles, the agencies concluded that it
                remains reasonable to maintain vehicle performance with a depleted
                battery (without electric assist) in order to maintain fully the
                performance characteristics under all conditions, and continued to use
                the NPRM methodology.
                ---------------------------------------------------------------------------
                 \516\ To represent marketplace trends better, the performance
                class of SHEVP2's allow acceleration time below 0.2 seconds less
                than the target, and PHEV20's and PHEV50's inherit combustion engine
                size from the conventional powertrain they are replacing. Further
                discussion of resizing targets can be found in Chapter 8 of the ANL
                Model Documentation for the final rule analysis.
                ---------------------------------------------------------------------------
                 The refinement for the standard performance SHEVP2 resolved the
                electrified packages issue identified by CARB, and also addressed most
                of the change in performance in the overall fleet, including with
                compact cars as mentioned by UCS. As explained further below, the
                agencies assessed performance among the alternatives for the final rule
                analysis. That assessment showed that, with the final rule refinements,
                245 out of 255 total resized vehicles (96 percent of vehicles) in the
                medium non-performance class (same
                [[Page 24336]]
                class focused on by CARB), had 0-60 mph acceleration times within the
                plus-or-minus 0.2 second window (8.8 to 9.2 seconds).\517\ The only
                vehicles outside the window were certain strong electrified vehicles
                which exceeded 0-60 the acceleration target as a result of achieving
                other performance criteria, such as the US06 driving cycles in all-
                electric-mode.\518\
                ---------------------------------------------------------------------------
                 \517\ This includes 135 strong electrified vehicles.
                 \518\ As noted earlier, electrified vehicles had to be capable
                of successfully completing UDDS or US06 driving cycles in all-
                electric mode, and in some cases the resulting motor size produced
                improved acceleration times.
                ---------------------------------------------------------------------------
                 The assessment also showed that for the small car class (mentioned
                by UCS) the acceleration times of performance and non-performance
                vehicles do not go beyond each other's targets. For example, the
                vehicle in the small car class with the very best 0-60 mph time and a
                conventional powertrain achieves an 8.38 second 0-60 mph time, which is
                slower than the performance small car baseline of 8 seconds. This
                vehicle had multiple incremental technologies applied, including for
                example aerodynamic improvements, and has not reached the threshold for
                engine resizing.\519\ After engine resizing, the ``fastest''
                conventional small car has a 0-60 mph time of 9.9 seconds, only 0.1
                seconds from the target of 10 seconds.\520\
                ---------------------------------------------------------------------------
                 \519\ Discussion of engine resizing can be found in Section
                VI.B.3.a)(5).
                 \520\ See NPRM Autonomie simulation database for Small cars,
                Docket ID NHTSA-2018-0067-1855.
                ---------------------------------------------------------------------------
                 CARB also commented on the improvement of ``passing times,'' or 50-
                80 mph high-speed acceleration times. As stated above, an improvement
                in one or more of the performance criteria is an expected outcome when
                using the rulemaking analysis methodology that resizes powertrains such
                that there is no degradation in any of the performance metrics.
                Consistent with past rulemakings, the agencies do not believe it is
                appropriate for the rulemaking analysis to show pathways that degrade
                vehicle performance or utility for one or more of the performance
                criteria, as doing so would adversely impact functional capability of
                the vehicle and could lead to customer dissatisfaction. The agencies
                agree there is very small increase in passing performance for some
                technology combinations, and believe this is an appropriate outcome.
                High-speed acceleration is rarely the least-capable performance
                criteria.
                 CARB, ICCT, UCS, and H-D Systems (HDS), in an attempt to identify a
                potential cause for changes in performance, commented that the CAFE
                model should have placed fewer constraints on engine resizing. CARB and
                ICCT commented that engine resizing should have been allowed even at
                low levels of mass reduction. Comments from CARB, UCS, HDS, and ICCT
                stated that engine resizing should also have been allowed for other
                incremental technologies, and within their comments they conducted
                performance analysis of non-resized cases.
                 CARB claimed that requiring a minimum of 7.1 percent curb weight
                reduction before engine resizing is a constraint that ``limits the
                optimization of the technologies being applied.'' \521\ UCS stated that
                ``a significant share of the benefit of a few percent reduction in mass
                has gone towards improved performance rather than improved fuel
                economy, leaving a substantial benefit of mass reduction underutilized
                and/or uncounted.'' \522\ ICCT also commented that ``when vehicle
                lightweighting is deployed at up to a 7 percent mass reduction, the
                engine is not resized even though less power would be needed for the
                lighter vehicle, meaning any such vehicles inherently are higher
                performance.'' \523\
                ---------------------------------------------------------------------------
                 \521\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 178. Note, a 7.1% curb weight reduction
                equates to the agencies' third level of mass reduction (MR3);
                additional discussion of engine resizing for mass reduction can be
                found in Section VI.B.3.a)(4) Autonomie Sizes Powertrains for Full
                Vehicle Simulation] and in the ANL Model Documentation for the final
                rule analysis.
                 \522\ Union of Concerned Scientists, Attachment 2, Docket No.
                NHTSA-2018-0067-12039, at 11.
                 \523\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-50.
                ---------------------------------------------------------------------------
                 UCS and HDS commented on the lack of resizing for technologies
                other than mass reduction, with HDS stating that ``the Agencies
                incorrectly limited the efficacy of technologies that reduce tractive
                load because their modeling does not re-optimize engine performance
                after applying these technologies.'' \524\ CARB also commented that the
                lack of resizing when a BISG or CISG system is added ``results in a
                less than optimized system that does not take full advantage of the
                mild hybrid system.'' Similarly, ICCT noted a case in which a Dodge RAM
                ``did not apply engine downsizing with the BISG system on that truck,
                so there are also significant performance benefits that should be
                accounted for, meaning that for constant-performance the fuel
                consumption reduction would be even greater.'' \525\
                ---------------------------------------------------------------------------
                 \524\ H-D Systems, Attachment 1, Docket No. NHTSA-2018-0067-
                12395, at 4. For reference, technologies that reduce tractive road
                load include mass reduction, aerodynamic drag reduction, and tire
                rolling resistance reduction.
                 \525\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-24.
                ---------------------------------------------------------------------------
                 CARB further commented on the performance improvement in cases
                without engine resizing by stating that ``94 percent of the packages
                modeled result in improved performance,'' and that for these non-
                resized cases that were actually adopted by a vehicle in the
                simulation, ``fewer than 20 percent maintained baseline performance
                with gains of 2 percent or less in acceleration time.'' \526\ Referring
                specifically to non-resized electrified vehicles, CARB also stated that
                ``44,878 of the 53,818 packages, or greater than 83 percent, result in
                improved performance.'' \527\ CARB also commented that engine sharing
                across different vehicles within a platform, which in some cases may
                constrain resizing for a member of that platform, should not dictate
                that these engines must remain identical in all aspects, and that
                ``this overly restrictive sharing of identical engines newly imposed in
                the CAFE Model is not consistent with today's industry practices and
                results in less optimal engine sizing and causes a systematic
                overestimation of technology costs to meet the existing standards.''
                \528\
                ---------------------------------------------------------------------------
                 \526\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 183.
                 \527\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 187.
                 \528\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 185.
                ---------------------------------------------------------------------------
                 The agencies note broadly, in response to these comments, that when
                conducting an analysis which balances performance neutrality against
                the realities faced by manufacturers, such as manufacturing complexity,
                economies of scale, and maintaining the full range of performance
                criteria, it is inevitable to observe at least some minor shift in
                vehicle performance. For example, if a new transmission is applied to a
                vehicle, the greater number of gear ratios helps the engine run in its
                most efficient range which improves fuel economy, but also helps the
                engine to run in the optimal ``power band'' which improves performance.
                Thus, the technology can provide both improved fuel economy and
                performance. Another example is applying a small amount of mass
                reduction that improves both fuel economy and performance by a small
                amount. Resizing the engine to maintain performance in these examples
                would require a unique engine displacement that is only slightly
                different than the baseline engine. While engine resizing in these
                incremental cases could have some small benefit to fuel economy, the
                [[Page 24337]]
                gains may not justify the costs of producing unique niche engines for
                each combination of technologies. If manufacturers were to produce
                marginally downsized engines to complement every small increment of
                mass reduction or technology, the resulting large number of engine
                variants that would need to be manufactured would cause a substantial
                increase in manufacturing complexity, and require significant changes
                to manufacturing and assembly plants and equipment.\529\ The high costs
                would be economically infeasible.
                ---------------------------------------------------------------------------
                 \529\ For example, each unique engine would require unique
                internal components such as crankshafts, pistons, and connecting
                rods, as well as unique engine calibrations for each displacement.
                Assembly plants would need to stock and feed additional unique
                engines to the stations where engines are dressed and inserted into
                vehicles.
                ---------------------------------------------------------------------------
                 Also, as noted in the NPRM, the 2015 NAS report stated that ``[f]or
                small (under 5 percent [of curb weight]) changes in mass, resizing the
                engine may not be justified, but as the reduction in mass increases
                (greater than 10 percent [of curb weight]), it becomes more important
                for certain vehicles to resize the engine and seek secondary mass
                reduction opportunities.'' \530\ In consideration of both the NAS
                report and comments received from manufacturers, the agencies
                determined it would be reasonable to allow allows engine resizing upon
                adoption of 7.1 percent, 10.7 percent, 14.2 percent, and 20 percent
                curb weight reduction, but not at 3.6 percent and 5.3 percent.\531\
                Resizing is also allowed upon changes in powertrain type or the
                inheritance of a powertrain from another vehicle in the same platform.
                The increments of these higher levels of mass reduction, or complete
                powertrain changes, more appropriately match the typical engine
                displacement increments that are available in a manufacturer's engine
                portfolio.
                ---------------------------------------------------------------------------
                 \530\ National Research Council. 2011. Assessment of Fuel
                Economy Technologies for Light-Duty Vehicles. Washington, DC--The
                National Academies Press. http://nap.edu/12924.
                 \531\ These curb weight reductions equate to the following
                levels of mass reduction as defined in the analysis: MR3, MR4, MR5
                and MR6, but not MR1 and MR2; additional discussion of engine
                resizing for mass reduction can be found in Section VI.B.3.a)(6)
                Autonomie Sizes Powertrains for Full Vehicle Simulation.
                ---------------------------------------------------------------------------
                 The agencies point to the comments from manufacturers, discussed
                further above, which support the agencies' assertion that the CAFE
                model's resizing constraints are appropriate. As discussed previously,
                Toyota commented that this approach better considers the constraints of
                engineering resources and manufacturing costs and results in a more
                realistic number of engines and transmissions.\532\ The Alliance also
                commented on the benefit of constraining engine resizing, stating that
                ``the platform and engine sharing methodology in the model better
                replicates reality by making available to each manufacturer only a
                finite number of engine displacements, helping to prevent
                unrealistically `over-optimized' engine sizing.'' \533\
                ---------------------------------------------------------------------------
                 \532\ Toyota, Attachment 1, Docket No. NHTSA-2018-0067-12098, at
                6.
                 \533\ Alliance of Automobile Manufacturers, Attachment ``Full
                Comment Set,'' Docket No. NHTSA-2018-0067-12073, at 140.
                ---------------------------------------------------------------------------
                 Another comment from CARB stated that engine resizing ``was only
                simulated for cases where those levels of mass reduction were applied,
                in the absence of virtually all other technology or efficiency
                improvements.'' \534\ The agencies do not agree that resizing should be
                simulated in all cases which involve small incremental technologies. In
                the final rule analysis, vehicles can have engines resized at four (out
                of six) levels of mass reduction technology, during a vehicle redesign
                cycle which changes powertrain architecture, and by inheritance during
                a vehicle refresh cycle. As discussed previously, the application of
                small incremental technologies such as reductions in aerodynamic drag
                or rolling resistance does not justify the high cost and complexity of
                producing additional varieties of engine sizes. Accordingly, for each
                curb weight reduction level of 7.1 percent or above and for each
                vehicle technology class, Autonomie sized a baseline engine by running
                a simulation of a vehicle without incremental technologies applied;
                then, those baseline engines were inherited by all other simulations
                using the same levels of curb weight reduction, which also added any
                variety of incremental technologies.\535\ For further clarification, in
                any case in which a vehicle adopts a 7.1 percent or more curb weight
                reduction, no matter what other technologies were already present or
                are added to the vehicle in conjunction with the mass reduction, that
                vehicle will receive an engine which has been appropriately sized for
                the newly applied mass reduction level.\536\ This can be observed in
                the Autonomie simulation databases by tracking the ``EngineMaxPower''
                column (not the ``VehicleSized'' column).
                ---------------------------------------------------------------------------
                 \534\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 178.
                 \535\ In the Autonomie simulation database files, the
                simulations which establish baseline sized engines are marked
                ``yes'' in the ``VehicleSized'' column, and the subsequent
                simulations which use this engine and add other incremental
                technologies are marked ``inherited.'' For a list of Autonomie
                simulation database files, see Table VI-4 Autonomie Simulation
                Database Output Files in Section VI.A.7 Structure of Model Inputs
                and Outputs.
                 \536\ For example, if a vehicle possesses MR2, AERO1, and ROLL1
                and subsequently adopts MR3, AERO1, ROLL2, the vehicle will adopt
                the lower engine power level associated with MR3. As a counter
                example, if a vehicle possesses MR3, ROLL1, and AERO1 and
                subsequently adopts MR3, ROLL1, AERO2, the engine will not be
                resized and it will retain the power level associated with MR3.
                ---------------------------------------------------------------------------
                 Finally, ICCT claimed that the agencies did not sufficiently report
                performance-related vehicle information. ICCT commented that the output
                files did not show data on ``engine displacement, the maximum power of
                each engine, the maximum torque of each engine, the initial and final
                curb weight of each vehicle (in absolute terms), and estimated 0-60 mph
                acceleration.'' ICCT claimed that because this data was not found, the
                agencies are ``showing that they have not even attempted to analyze
                accurately the future year fleet for their performance'' and that ``the
                agencies are intentionally burying a critical assumption, whereby their
                future fleet has not been appropriately downsized, and it therefore has
                greatly increased utility and performance characteristics.'' \537\
                ---------------------------------------------------------------------------
                 \537\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-74.
                ---------------------------------------------------------------------------
                 In fact, for the NPRM, and again for this final rule, the agencies
                did analyze vehicle performance and have made the data available to the
                public. An indication of the actual engine displacement change is
                available by noting the displacements used in Automonie simulation
                database for each of the technology states. The displacements reported
                in Autonomie are used by the full-vehicle-simulation within the
                Autonomie model, and while they do not directly represent each specific
                vehicle's actual engine sizes, they do fully reflect the relative
                change in engine size that is applied to each vehicle. It is the
                relative change in engine size that is relevant for the analysis.
                Similarly, the vehicle power and torque used by the full vehicle
                simulations are reported in the Autonomie simulation databases; their
                values and relative change across an engine resizing event can be
                observed. Initial and final curb weights for the analysis fleet are
                reported in Vehicles Report output file column titled ``CW Initial''
                and ``CW,'' respectively. The time required for 0-60 mph acceleration
                is reported in the Autonomie simulation database files. A detailed
                description of the engine resizing methodology is available in the
                Argonne Model
                [[Page 24338]]
                Documentation, which explains how vehicle characteristics are used to
                calculate powertrain size.\538\ These data and information that are
                available in the Autonomie and CAFE model documentation provide the
                information needed to analyze performance, and in fact, this is
                evidenced by the statements of numerous commenters discussed in this
                section. The agencies have conducted their own performance analysis,
                which is discussed further below, using the same data documentation
                mentioned here.
                ---------------------------------------------------------------------------
                 \538\ See Chapter 8 of the ANL Model documentation for the final
                rule analysis.
                ---------------------------------------------------------------------------
                 Updates to the CAFE model have minimized performance shift over the
                simulated model years, and have eliminated performance differences
                between simulated standards.
                 The Autonomie simulation updates, discussed previously, were
                included in the final rule analysis, and have resulted in average
                performance that is similar across the regulatory alternatives. Because
                the regulatory analysis compares differences in impacts among the
                alternatives, the agencies believe that having consistent performance
                across the alternatives is an important aspect of performance
                neutrality. If the vehicle fleet had performance gains which varied
                significantly depending on the alternative, performance differences
                would impact the comparability of the simulations. Using the NPRM CAFE
                model data, the agencies analyzed the sales-weighted average 0-60
                performance of the entire simulated vehicle fleet for MYs 2016 and
                2029, and identified that the Augural standards had 4.7 percent better
                0-60 mph acceleration time compared to the NPRM preferred alternative,
                which had no changes in standards in MYs 2021-2026.\539\ This
                assessment confirmed the observations of the various commenters. With
                the refinements that were incorporated for the final rule, similar
                analysis showed that the Augural standards had a negligible 0.1 percent
                difference in 0-60 mph acceleration time compared to the NPRM preferred
                alternative.\540\
                ---------------------------------------------------------------------------
                 \539\ The agencies' analysis matched all MY 2016 and MY 2029
                vehicles in the NPRM Vehicles Report output file, under both the
                Augural standards and preferred alternative, with the appropriate 0-
                60 mph acceleration time from the NPRM Autonomie simulation
                databases. This was done by examining each vehicle's assigned
                technologies, finding the Autonomie simulation with the
                corresponding set of technologies, and extracting that simulation's
                0-60 mph acceleration time. This process effectively assigned a 0-60
                time to every vehicle in the fleet for four scenarios: (1) MY 2016
                under augural standards, (2) MY 2016 under the preferred
                alternative, (3) MY 2029 under augural standards, and (4) MY 2029
                under the preferred alternative. For each scenario, an overall
                fleet-wide weighted average 0-60 time was calculated, using each
                vehicle's MY2016 sales volumes as the weight. For more information,
                see the FRIA Section VI.
                 \540\ This updated analysis used the FRM CAFE Model Vehicles
                Report output file and the FRM Autonomie simulation databases. The
                final rule analysis introduced an updated MY 2017 fleet as a
                starting point, replacing the NPRM 2016MY fleet. For more
                information, see the FRIA Chapter VI.
                ---------------------------------------------------------------------------
                 The updates applied to the final rule Autonomie simulations also
                resulted in further minimizing the performance change across model
                years. As the agencies attempted to minimize this performance shift
                occurring ``over time,'' it was also acknowledged that a small increase
                would be expected and would be reasonable. This increase is attributed
                to the analysis recognizing the practical constraints on the number of
                unique engine displacements manufacturers can implement, and therefore
                not resizing powertrains for every individual technology and every
                combination of technologies when the performance impacts are small.
                Perfectly equal performance with 0 percent change would not be
                achievable while accounting for these real world resizing constraints.
                The performance analysis in the 2011 NAS report shared a similar view
                on performance changes, stating that ``truly equal performance involves
                nearly equal values . . . within 5 percent.'' \541\ In response to
                comments, using NPRM CAFE model data, the agencies analyzed the sales-
                weighted average 0-60 performance of the entire simulated vehicle
                fleet, and identified that the performance increase from MYs 2016 and
                2029 was 7.5 percent under Augural Standards and 3.1 percent under the
                NPRM preferred alternative standards. The agencies conducted a similar
                analysis using final rule data and found the performance increase over
                time from MYs 2017 to 2029 was 3.9 percent for Augural Standards and
                4.0 percent for the NPRM preferred alternative standards. The agencies
                determined this change in performance is reasonable and note it is
                within the 5 percent bound in discussed by NAS in its 2011 report.
                ---------------------------------------------------------------------------
                 \541\ National Research Council. 2011. Assessment of Fuel
                Economy Technologies for Light-Duty Vehicles. Washington, DC--The
                National Academies Press, at 62. http://nap.edu/12924.
                ---------------------------------------------------------------------------
                 This assessment shows that for the final rule analysis, performance
                is neutral across regulatory alternatives and across the simulated
                model years allowing for fair, direct comparison among the
                alternatives.
                (7) How the Agencies Simulated Vehicle Models on Test Cycles
                 After vehicle models are built for every combination of
                technologies and vehicle classes represented in the analysis, Autonomie
                simulates their performance on test cycles to calculate the
                effectiveness improvement of the fuel-economy-improving technologies
                that have been added to the vehicle. Discussed earlier, the agencies
                minimize the impact of potential variation in determining effectiveness
                by using a series of tests and procedures specified by federal law and
                regulations under controlled conditions.
                 Autonomie simulates vehicles in a very similar process as the test
                procedures and energy consumption calculations that manufacturers must
                use for CAFE and CO2 compliance.542 543 544
                Argonne simulated each vehicle model on several test procedures to
                evaluate effectiveness. For vehicles with conventional powertrains and
                micro hybrids, Autonomie simulates the vehicles on EPA 2-cycle test
                procedures and guidelines.\545\ For mild and full hybrid electric
                vehicles and FCVs, Autonomie simulates the vehicles using the same EPA
                2-cycle test procedure and guidelines, and the drive cycles are
                repeated until the initial and final state of charge are within a SAE
                J1711 tolerance. For PHEVs, Autonomie simulates vehicles in similar
                procedures and guidelines as SAE J1711.\546\ For BEVs Autonomie
                simulates vehicles in similar procedures and guidelines as SAE
                J1634.\547\
                ---------------------------------------------------------------------------
                 \542\ EPA, ``How Vehicles are Tested.'' https://www.fueleconomy.gov/feg/how_tested.shtml. Last accessed Nov 14,
                2019.
                 \543\ ANL model documentation for final rule Chapter 6. Test
                Procedures and Energy Consumption Calculations.
                 \544\ EPA Guidance Letter. ``EPA Test Procedures for Electric
                Vehicles and Plug-in Hybrids.'' Nov. 14, 2017. https://www.fueleconomy.gov/feg/pdfs/EPA%20test%20procedure%20for%20EVs-PHEVs-11-14-2017.pdf. Last accessed Nov. 7, 2019.
                 \545\ 40 CFR part 600.
                 \546\ PHEV testing is broken into several phased based on SAE
                J1711. Charge-Sustaining on the City cycle, Charge-Sustaining on the
                HWFET cycle, Charge-Depleting on the City and HWFET cycles.
                 \547\ SAE J1634. ``Battery Electric Vehicle Energy Consumption
                and Range Test Procedure.'' July 12, 2017.
                ---------------------------------------------------------------------------
                b) Selection of One Full-Vehicle Modeling and Simulation Tool
                 The NPRM described tools that the agencies previously used to
                estimate technology effectiveness. For the analysis supporting the 2012
                final rule for MYs 2017 and beyond, the agencies used technology
                effectiveness estimates from EPA's lumped parameter model (LPM). The
                LPM was calibrated using data from vehicle simulation work performed by
                Ricardo Engineering.\548\
                [[Page 24339]]
                The agencies also used full vehicle simulation modeling data from
                Autonomie vehicle simulations performed by Argonne for mild hybrid and
                advanced transmission effectiveness estimates.549 550
                ---------------------------------------------------------------------------
                 \548\ Response to Peer Review of: Ricardo Computer Simulation of
                Light-Duty Vehicle Technologies for Greenhouse Gas Emission
                Reduction in the 2020-2025 Timeframe, EPA-420-R-11-021 (December
                2011), available at https://nepis.epa.gov/Exe/ZyPDF.cgi/P100D5BX.PDF?Dockey=P100D5BX.PDF.
                 \549\ Joint TSD: Final Rulemaking for 2017-2025 Light-Duty
                Vehicle Greenhouse Emission Standards and Corporate Average Fuel
                Economy Standards. August 2012. EPA-420-R-12-901.3.3.1.3 Argonne
                National Laboratory Simulation Study p. 3-69.
                 \550\ Moawad, A. and Rousseau, A., ``Impact of Electric Drive
                Vehicle Technologies on Fuel Efficiency,'' Energy Systems Division,
                Argonne National Laboratory, ANL/ESD/12-7, August 2012.
                ---------------------------------------------------------------------------
                 For the 2016 Draft TAR analysis, EPA and NHTSA used two different
                full system simulation programs for complementary but separate
                analyses. NHTSA used Argonne's Autonomie tool, described in detail
                above, with engine map inputs developed by IAV using GT-Power in 2014,
                and updated in 2016.551 552 553 Argonne, in coordination
                with NHTSA, developed a methodology for large scale simulation using
                Autonomie and distributed computing, thus overcoming one of the
                challenges to full vehicle simulation that the NAS committee outlined
                in its 2015 report and implementing a recommendation that the agencies
                use full-vehicle simulation to improve the analysis method of
                estimating technology effectiveness.\554\ EPA used a limited number of
                full-vehicle simulations performed using its ALPHA model, an EPA-
                developed full-vehicle simulation model,\555\ to calibrate the LPM,
                used to estimate technology effectiveness. EPA also used the same
                modeling approach for its Proposed Determination analysis.\556\
                ---------------------------------------------------------------------------
                 \551\ GT-Power Engine Simulation Software. https://www.gtisoft.com/gt-suite-applications/propulsion-systems/gt-power-engine-simulation-software/. Last accessed Oct. 10, 2019.
                 \552\ 2016 Draft TAR Engine Maps by IAV Automotive Engineering
                using GT-Power. https://www.nhtsa.gov/staticfiles/rulemaking/pdf/cafe/IAV_EngineMaps_Details.xlsx. Lass accessed Oct. 10, 2019.
                 \553\ NHTSA-2018-0067-0003. ANL--Summary of Main Component
                Performance Assumptions NPRM.
                 \554\ See National Research Council. 2015. Cost, Effectiveness,
                and Deployment of Fuel Economy Technologies for Light-Duty Vehicles.
                Washington, DC: The National Academies Press [hereinafter ``2015 NAS
                Report''] at p. 263, available at https://www.nap.edu/catalog/21744/cost-effectiveness-and-deployment-of-fuel-economy-technologies-for-light-duty-vehicles (last accessed June 21, 2018). See also A.
                Moawad, A. Rousseau, P. Balaprakash, S. Wild, ``Novel Large Scale
                Simulation Process to Support DOT's CAFE Modeling System,''
                International Journal of Automotive Technology (IJAT), Paper No.
                220150349, Nov 2015; Pagerit, S., Sharper, P., Rousseau, A., Sun, Q.
                Kropinski, M. Clark, N., Torossian, J., Hellestrand, G., ``Rapid
                Partitioning, Automatic Assembly and Multicore Simulation of
                Distributed Vehicle Systems.'' ANL, General Motors, EST Embedded
                Systems Technology. 2015. https://www.autonomie.net/docs/5%20-%20Presentations/VPPC2015_ppt.pdf. Last accessed Dec. 9, 2019.
                 \555\ See Lee, B., S. Lee, J. Cherry, A. Neam, J. Sanchez, and
                E. Nam. 2013. Development of Advanced Light-Duty Powertrain and
                Hybrid Analysis Tool. SAE Technical Paper 2013-01-0808. doi:
                10.4271/2013-01-0808.
                 \556\ Proposed Determination on the Appropriateness of the Model
                Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions Standards
                under the Midterm Evaluation, EPA-420-R-16-020 (November 2016),
                available at https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100Q3DO.pdf; Final Determination on the
                Appropriateness of the Model Year 2022-2025 Light-Duty Vehicle
                Greenhouse Gas Emissions Standards under the Midterm Evaluation,
                EPA-420-R-17-001 (January 2017), available at https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100QQ91.pdf.
                ---------------------------------------------------------------------------
                 In the subsequent August 2017 Request for Comment on
                Reconsideration of the Final Determination of the Mid-Term Evaluation
                of Greenhouse Gas Emissions Standards for MY 2022-2025 Light-Duty
                Vehicles, the agencies requested comments on whether EPA should use
                alternative methodologies and modeling, including the Autonomie full-
                vehicle simulation tool and DOT's CAFE model, for the analysis that
                would accompany its revised Final Determination.\557\ As discussed in
                the NPRM, stakeholders questioned the efficacy of the combined outputs
                and assumptions of the LPM and ALPHA,\558\ especially as the tools were
                used to evaluate increasingly heterogeneous combinations of
                technologies in the vehicle fleet.\559\
                ---------------------------------------------------------------------------
                 \557\ 82 FR 39551 (Aug. 21, 2017).
                 \558\ 83 FR 43022 (``At NHTSA-2016-0068-0082, p. 49, FCA
                provided the following comments, ``FCA believes EPA is
                overestimating the benefits of technology. As the LPM is calibrated
                to those projections, so too is the LPM too optimistic.'' FCA also
                shared the chart, `LPM vs. Actual for 8 Speed Transmissions.' '').
                 \559\ 83 FR 43022 (referencing Automotive News ``CAFE math gets
                trickier as industry innovates'' (Kulisch), March 26, 2018.).
                ---------------------------------------------------------------------------
                 More specifically, the Auto Alliance noted that their previous
                comments to the midterm evaluation, in addition to comments from
                individual manufacturers, highlighted multiple concerns with EPA's
                ALPHA model that were unresolved, but addressed in Autonomie.\560\
                First, the Alliance expressed concern over ALPHA modeling errors
                related to road load reductions, stating that an error derived from how
                mass and coast-down coefficients were updated when mass, tire and aero
                improvements were made resulted in benefits overstated by 3 percent to
                11 percent for all vehicle types. Next, the Alliance repeated its
                concern that EPA should consider top-gear gradeability as one of its
                performance metrics to maintain functionality, noting that EPA had
                acknowledged the industry's comments in the Proposed Determination,
                ``but generally dismissed the auto industry concerns.'' Additional
                analysis by EPA in its Response to Comments document did not allay the
                Alliance's concerns,\561\ as the Alliance concluded that ``[c]onsistent
                with the National Academy of Sciences recommendation from 2011, EPA
                should monitor gradeability to ensure minimum performance.''
                ---------------------------------------------------------------------------
                 \560\ EPA-HQ-OAR-2015-0827-9194, at p. 36-44.
                 \561\ The Alliance noted that in higher-gear-count
                transmissions, like 8-speed automatics, modeled by ALPHA with an
                expanded ratio spread to achieve fuel economy, are concerning for
                gradeability. Additionally, infinite engine downsizing along with
                expanded ratio spread transmission, in real world gradeability may
                cause further deteriorate as modeled in ALPHA, which leads to
                inflated effectiveness values for powertrains that would not meet
                customer demands.
                ---------------------------------------------------------------------------
                 Furthermore, the Alliance stated that ALPHA vehicle technology
                walks provided in response to manufacturer comments on the Proposed
                Determination did not correctly predict cumulative effectiveness when
                compared to technologies in real world applications. The Alliance
                stated that many of the individual technologies and assumptions used by
                ALPHA overestimated technology effectiveness and were derived from
                questionable sources. As an example, the Alliance referenced an engine
                map used by EPA to represent the Honda L15B7 engine, where the engine
                map data was collected by ``(1) taking a picture of an SAE document
                containing an image of the engine map, and then (2) `digitizing' the
                image by `tracing image contours' '' (citing EPA's ALPHA
                documentation). The Alliance could not definitively state whether the
                ``digitization'' process, lack of detail in the source image, or
                another factor were the reasons that some regions of overestimated
                efficiency were observed in the engine map, but concluded that ``the
                use of this map should be discontinued within ALPHA,'' and ``any
                analysis conducted with it is highly questionable.'' Based on these
                concerns and others, the Alliance recommended that Autonomie be used to
                inform the downstream cost optimization models (i.e., the CAFE model
                and/or OMEGA).
                 Global Automakers argued that NHTSA's CAFE model, which
                incorporates data from Autonomie simulations, provided a more
                transparent and discrete step through each of the modeling
                scenarios.\562\ Global pointed out that the LPM is ``of particular
                concern due to its simplified technology projection processes,'' and it
                ``propagates fundamentally flawed
                [[Page 24340]]
                content into the ALPHA and OMEGA models and therefore cannot accurately
                assess the efficacy of fuel economy technologies.'' Global did note
                that EPA ``plans to abandon its reliance on LPM in favor of another
                modeling approach,'' referring to the RSE,\563\ but stated that ``EPA
                must provide stakeholders with adequate time to evaluate the updated
                modeling approach, ensure it is analytically robust, and provide
                meaningful feedback.'' Global Automakers concluded that EPA's engine
                mapping and tear-down analyses have played an important role in
                generating publicly-available information, and stated that the data
                should be integrated into the Autonomie model.
                ---------------------------------------------------------------------------
                 \562\ EPA-HQ-OAR-2015-0827-9728, at 14.
                 \563\ See Moskalik, A., Bolon, K., Newman, K., and Cherry, J.
                ``Representing GHG Reduction Technologies in the Future Fleet with
                Full Vehicle Simulation,'' SAE Technical Paper 2018-01-1273, 2018,
                doi:10.4271/2018-01-1273. Since 2018, EPA has employed vehicle-
                class-specific response surface equations automatically generated
                from a large number of ALPHA runs to more readily apply large-scale
                simulation results, which eliminated the need for manual calibration
                of effectiveness values between ALPHA and the LPM.
                ---------------------------------------------------------------------------
                 On the other hand, other stakeholders commented that EPA's ALPHA
                modeling should continue to be used, for procedural reasons like,
                ``[i]t would appear arbitrary for EPA now, after five years of modeling
                based on ALPHA, to declare it can no longer use its internally
                developed modeling tools and must rely solely on the Autonomie model,''
                and ``[t]he ALPHA model is inextricably built into the regulatory and
                technical process. It will require years of new analysis to replace the
                many ALPHA and OMEGA modeling inputs and outputs that permeate the
                entire rulemaking process, should EPA suddenly decide to change its
                models.'' \564\ Commenters also cited technical reasons to use ALPHA,
                like EPA's progress benchmarking and validating the ALPHA model to over
                fifteen various MY 2013-2015 vehicles,\565\ and that technologies like
                the ``Atkinson 2'' engine technology were not considered in NHTSA's
                compliance modeling.\566\ Commenters also cited that ALPHA was created
                to be publicly available, open-sourced, and peer-reviewed, ``to allow
                for transparency to both automakers and public stakeholders, without
                hidden and proprietary aspects that are present in commercial modeling
                products.'' \567\
                ---------------------------------------------------------------------------
                 \564\ EPA-HQ-OAR-2015-9826, at 39-40.
                 \565\ EPA-HQ-OAR-2015-9826, at 40.
                 \566\ EPA-HQ-OAR-2015-9197, at 28.
                 \567\ EPA-HQ-OAR-2015-9826, at 38.
                ---------------------------------------------------------------------------
                 The agencies described in the NPRM that after having reviewed
                comments about whether EPA should use alternative methodologies and
                modeling, and after having considered the matter fully, the agencies
                determined it was reasonable and appropriate to use Autonomie for full-
                vehicle simulation.\568\ The agencies stated that nothing in Section
                202(a) of the Clean Air Act (CAA) mandated that EPA use any specific
                model or set of models for analysis of potential CO2
                standards for light duty vehicles. The agencies also distinguished the
                models and the inputs used to populate them; specifically, comments
                presented as criticisms of the models, such as ``Atkinson 2'' engine
                technology not considered in the compliance modeling, actually
                concerned model inputs.\569\
                ---------------------------------------------------------------------------
                 \568\ 83 FR 43001.
                 \569\ 83 FR 43002.
                ---------------------------------------------------------------------------
                 With regards to modeling technology effectiveness, the agencies
                concluded that, although the CAFE model requires no specific approach
                to developing effectiveness inputs, the National Academy of Sciences
                recommended, and stakeholders have commented, that full-vehicle
                simulation provides the best balance between realism and practicality.
                As stated above, Argonne has spent several years developing, applying,
                and expanding means to use distributed computing to exercise its
                Autonomie full-vehicle simulation tool at the scale necessary for
                realistic analysis of technologies that could be used to comply with
                CAFE and CO2 standards, and this scalability and related
                flexibility (in terms of expanding the set of technologies to be
                simulated) makes Autonomie well-suited for developing inputs to the
                CAFE model.
                 In response to the NPRM, the Auto Alliance commented that NHTSA's
                modeling and analysis tools are superior to EPA's, noting that NHTSA's
                tools have had a significant lead in their development.\570\ The
                Alliance pointed out that Autonomie was developed from the beginning to
                address the complex task of combining two power sources in a hybrid
                powertrain, while EPA's ALPHA model had not been validated or used to
                simulate hybrid powertrains. While both models are physics-based
                forward looking vehicle simulators, the Alliance commented that
                Autonomie is fully documented with available training, while ALPHA
                ``has not been documented with any instructions making it difficult for
                users outside of EPA to run and interpret the model.'' The Alliance
                also mentioned specific improvements in the Autonomie simulations since
                the Draft TAR, including expanded performance classes to better
                consider vehicle performance characteristics, the inclusion of
                gradeability as a performance metric, as recommended by the NAS, the
                inclusion of new fuel economy technologies, and the removal of unproven
                technologies.
                ---------------------------------------------------------------------------
                 \570\ NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 The Alliance, Global Automakers, and other automakers writing
                separately all stated that the agencies should use one simulation and
                modeling tool for analysis.571 572 The Alliance stated that
                since both the Autonomie and ALPHA modeling systems answer essentially
                the same questions, using both systems leads to inconsistencies and
                conflicts, and is inefficient and counterproductive.
                ---------------------------------------------------------------------------
                 \571\ NHTSA-2018-0067-12073; NHTSA-2018-0067-12032. Comments of
                the Association of Global Automakers, Inc. on the Safer Affordable
                Fuel-Efficient Vehicles Rule Docket ID Numbers: NHTSA-2018-0067 and
                EPA-HQ-OAR-2018-0283 October 26, 2018.
                 \572\ NHTSA-2018-0067-11943. FCA Comments on The Safer
                Affordable Fuel-Efficient (SAFE) Vehicles Rule for Model Years 2021-
                2026 Passenger Cars and Light Trucks Notice of Proposed Rulemaking.
                ---------------------------------------------------------------------------
                 The agencies agree with the Alliance that the fully developed and
                validated Autonomie model fulfills the agencies' analytical needs for
                full-vehicle modeling and simulation. The agencies also agree that it
                is counterintuitive to have two separate models conducting the same
                work.
                 Some commenters stated that broadly, EPA was required to conduct
                its own technical analysis and rely on its own models to do so.\573\
                Those comments are addressed in Section IV.
                ---------------------------------------------------------------------------
                 \573\ NHTSA-2018-0067-12000; NHTSA-2018-0067-12039.
                ---------------------------------------------------------------------------
                 Regarding the merits of EPA's models, and based on previous inputs
                and assumptions used to populate those models, ICCT commented that
                ``[b]ased on the ICCT's global analysis of vehicle regulations, the
                EPA's physics-based ALPHA modeling offers the most sophisticated and
                thorough modeling of the applicable technologies that has ever been
                conducted.'' ICCT listed several reasons for this, including that the
                EPA modeling is based on systematic modeling of technologies and their
                synergies; it was built and improved upon by extensive modeling by and
                with Ricardo (an engineering consulting firm); it incorporated National
                Academies input at multiple stages; it has included many peer reviews
                at many stages of the modeling and the associated technical reports
                published by engineers in many technical journal articles and
                conference proceedings; and EPA's Draft TAR analysis, which used ALPHA,
                used state-of-the-art engine maps based on benchmarked high-efficiency
                engines. ICCT concluded
                [[Page 24341]]
                that ``[d]espite these rigorous advances in vehicle simulation
                modeling, it appears that the agencies have inexplicably abandoned this
                approach, expressly disregarding the EPA benchmarked engines, ALPHA
                modeling, and all its enhancements since the last rulemaking.''
                 The hallmarks ICCT lists regarding the ALPHA modeling are equally
                applicable to Autonomie.\574\ Autonomie is also based on systematic
                modeling of technologies and their synergies when combined as packages.
                The U.S. Department of Energy created Autonomie, and over the past two
                decades, helped to develop and mature the processes and inputs used to
                represent real-world vehicles using continuous feedback from the tool's
                worldwide user base of vehicle manufacturers, suppliers, government
                agencies, and other organizations. Moreover, using Autonomie brings the
                agencies closer to the NAS Committee's stated goal of ``full system
                simulation modeling for every important technology pathway and for
                every vehicle class.'' \575\ While the NAS Committee originally thought
                that full vehicle simulation modeling would not be feasible for the
                thousands of vehicles in the analysis fleets because the technologies
                present on the vehicles might differ from the configurations used in
                the simulation modeling,\576\ Argonne has developed a process to
                simulate explicitly every important technology pathway for every
                vehicle class. Moreover, although separate from the Autonomie model
                itself, the Autonomie modeling for this rulemaking incorporated other
                NAS committee recommendations regarding full vehicle simulation inputs
                and input assumptions, including using engine-model-generated maps
                derived from a validated baseline map in which all parameters except
                the new technology of interest are held constant.\577\
                ---------------------------------------------------------------------------
                 \574\ See Theo LeSieg, Ten Apples Up On Top! (1961), at 4-32.
                 \575\ 2015 NAS Report at 358.
                 \576\ 2015 NAS Report at 359.
                 \577\ NAS Recommendation 2.1.
                ---------------------------------------------------------------------------
                 As discussed further below and in VI.C.1 Engine Paths, this is one
                reason why the IAV maps were used instead of the EPA maps, and the
                agencies instead referenced EPA's engine maps to corroborate the
                Autonomie effectiveness results. The IAV maps are engine-model-
                generated maps derived from a validated baseline map in which all
                parameters except the new technology of interest are held constant.
                While EPA's engine maps benchmarking specific vehicles' engines
                incorporate multiple technologies, for example including improvements
                in engine friction and reduction in accessory parasitic loads,
                comparisons presented in Section VI.C.1 showed that engine maps
                developed by IAV, while not exactly the same, are representative of
                EPA's engine benchmarking data.
                 In addition, both ALPHA and Autonomie have been used to support
                analyses that have been published in technical journal articles and
                conference proceedings, but those analyses differ fundamentally because
                of the nature of the tools. ALPHA was developed as a tool to be used by
                EPA's in-house experts.\578\ As EPA stated in the ALPHA model peer
                review,\579\ ``ALPHA is not intended to be a commercial product or
                supported for wide external usage as a development tool.'' \580\
                Accordingly, EPA experts have published several peer-reviewed journal
                articles using ALPHA and have presented the results of those papers at
                conference proceedings.\581\
                ---------------------------------------------------------------------------
                 \578\ ALPHA Peer Review, at 4-1.
                 \579\ ICCT's comments intimate that ALPHA has been peer reviewed
                at many stages of the modeling; although EPA has published several
                peer-reviewed technical papers, the ALPHA model itself has been
                subject to one peer review. See Peer Review of ALPHA Full Vehicle
                Simulation Model, available at https://nepis.epa.gov/Exe/ZyPdf.cgi?Dockey=P100PUKT.pdf.
                 \580\ ALPHA Peer Review, at 4-2.
                 \581\ See, e.g., Dekraker, P., Kargul, J., Moskalik, A., Newman,
                K. et al., ``Fleet-Level Modeling of Real World Factors Influencing
                Greenhouse Gas Emission Simulation in ALPHA,'' SAE Int. J. Fuels
                Lubr. 10(1):2017, doi:10.4271/2017-01-0899.
                ---------------------------------------------------------------------------
                 To explore ICCT's comments on the importance of peer review
                further, it is important to take the actual substantive content of the
                ALPHA peer review into account.\582\ One reviewer raised significant
                questions over the availability of ALPHA documentation, stating
                ``[t]here is an overall lack of detail on key technical features that
                are new in the model,'' and ``[w]e were not able to find any
                information on how the model handles component weight changes.''
                Reviewers also raised questions related to model readiness, stating
                ``[a]ccording to the documentation review, ALPHA's stop/start modeling
                appears to be very simplistic.'' Moreover, when running ALPHA
                simulations, the reviewer noted the results ``strongly suggest that the
                model has errors in the underlying equations or coding with respect to
                all of the load reductions.'' Also, one reviewer said the following of
                ALPHA: ``A specific simulation runtime is significantly high, more than
                10 mins. without providing any indication to the user progress made so
                far. A fairly more complicated model such as Autonomie available even
                with enhanced capabilities is significantly faster today.'' \583\
                ---------------------------------------------------------------------------
                 \582\ EPA. ``Peer Review of ALPHA Full Vehicle Simulation
                Model.'' EPA-420-R-16-013. October 2016. https://nepis.epa.gov/Exe/ZyPdf.cgi?Dockey=P100PUKT.pdf. Last accessed Nov 18, 2019.
                 \583\ Peer Review of ALPHA Full Vehicle Simulation Model, at C-
                4, available at https://nepis.epa.gov/Exe/ZyPdf.cgi?Dockey=P100PUKT.pdf.
                ---------------------------------------------------------------------------
                 The peer reviewer's assessment of Autonomie as a more complicated
                model with enhanced capabilities is not surprising, given Autonomie's
                history of development. Autonomie is a commercial tool with more than
                275 worldwide organizational users, including vehicle manufacturers,
                suppliers, government agencies, and nonprofit organizations having
                licensed and used Autonomie. Both Autonomie's creators and user base
                unaffiliated with Argonne have published over 100 papers, including
                peer-reviewed papers in journals, related to Autonomie validation and
                other studies.584 585 One could even argue that the tool has
                been continuously peer reviewed by these thousands of experts over the
                past two decades.
                ---------------------------------------------------------------------------
                 \584\ At least 15 peer-reviewed papers authored by ANL experts
                have been referenced throughout this Section, and others can be
                found at SAE International's website, https://www.sae.org/, using
                the search bar for ``Autonomie.''
                 \585\ See, e.g., Haupt, T., Henley, G., Card, A., Mazzola, M. et
                al., ``Near Automatic Translation of Autonomie-Based Power Train
                Architectures for Multi-Physics Simulations Using High Performance
                Computing,'' SAE Int. J. Commer. Veh. 10(2):483-488, 2017, https://doi.org/10.4271/2017-01-0267; Samadani, E., Lo, J., Fowler, M.,
                Fraser, R. et al., ``Impact of Temperature on the A123 Li-Ion
                Battery Performance and Hybrid Electric Vehicle Range,'' SAE
                Technical Paper 2013-01-1521, 2013, https://doi.org/10.4271/2013-01-1521.
                ---------------------------------------------------------------------------
                 In fact, in responding to a peer review comment on the ALPHA
                model's underlying equations and coding with respect to road load
                reductions, EPA noted that Autonomie had been used as a reference
                system simulation tool to validate ALPHA model results.\586\
                ---------------------------------------------------------------------------
                 \586\ Peer Review of ALPHA Full Vehicle Simulation Model, at 4-
                14 and 4-15, available at https://nepis.epa.gov/Exe/ZyPdf.cgi?Dockey=P100PUKT.pdf.
                ---------------------------------------------------------------------------
                 Outside of formal peer-reviewed studies, Autonomie has been used by
                organizations like ICCT to support policy documents, position briefs,
                and white papers assessing the potential of future efficiency
                technologies to meet potential regulatory requirements,\587\
                [[Page 24342]]
                just as the agencies did in this rulemaking.
                ---------------------------------------------------------------------------
                 \587\ See, e.g., Oscar Delgado and Nic Lutsey, Advanced Tractor-
                Trailer Efficiency Technology Potential in the 2020-2030 Timeframe
                (April 2015), available at https://theicct.org/sites/default/files/publications/ICCT_ATTEST_20150420.pdf; Ben Sharpe, Cost-
                Effectiveness of Engine Technologies for a Potential Heavy-Duty
                Vehicle Fuel Efficiency Regulation in India (June 2015), available
                at https://theicct.org/sites/default/files/publications/ICCT_position-brief_HDVenginetech-India_jun2015.pdf; Ben Sharpe and
                Oscar Delgado, Engines and tires as technology areas for efficiency
                improvements for trucks and buses in India (working paper published
                March 2016), available at https://theicct.org/sites/default/files/publications/ICCT_HDV-engines-tires_India_20160314.pdf.
                ---------------------------------------------------------------------------
                 Similarly to ICCT, UCS stated that in contrast to Autonomie, ALPHA
                had been thoroughly peer-reviewed and is constantly being updated to
                reflect the latest technology developments based on work performed by
                the National Vehicle and Fuel Emissions Laboratory.\588\ UCS also
                stated that because EPA has direct control over the model and its
                interface to OMEGA, EPA can better ensure that the inputs into OMEGA
                reflect the most up-to-date data, unlike the Autonomie work, which
                effectively has to be ``locked in'' before it can be deployed in the
                CAFE model. UCS also stated that ALPHA is based on the GEM model (used
                to simulate compliance with heavy-duty vehicle regulations) which was
                been updated with feedback from heavy-duty vehicle manufacturers and
                suppliers, and in fact, ``NHTSA has such confidence in the GEM model
                that they accept its simulation-based results as compliance with the
                heavy-duty fuel economy regulations.''
                ---------------------------------------------------------------------------
                 \588\ NHTSA-2018-0067-12039 (UCS).
                ---------------------------------------------------------------------------
                 Again, the agencies believe that it is important to note that
                Autonomie not only meets, but also exceeds, UCS' listed metrics.
                Autonomie's models, sub-models, and controls are constantly being
                updated to reflect the latest technology developments based on work
                performed by Argonne National Laboratory's Advanced Mobility Technology
                Laboratory (AMTL) (formerly Advanced Powertrain Research Facility, or
                ARPF).589 590 The Autonomie validation has included nine
                validation studies with accompanying reports for software, six
                validation studies and reports for powertrains, nine validation studies
                and reports for advanced components, ten validation studies and reports
                for advanced controls, and overall model validation using test data
                from over 50 vehicles.\591\
                ---------------------------------------------------------------------------
                 \589\ See NPRM PRIA. The agencies cited a succinctly-summarized
                presentation of Autonomie vehicle validation procedures based on
                AMTL test data in the NPRM ANL modeling documentation and PRIA
                docket for stakeholders to review at NHTSA-2018-0067-1972 and NHTSA-
                2018-0067-0007.
                 \590\ Jeong, J., Kim, N., Stutenberg, K., Rousseau, A.,
                ``Analysis and Model Validation of the Toyota Prius Prime,'' SAE
                2019-01-0369, SAE World Congress, Detroit, April 2019; Kim, N,
                Jeong, J., Rousseau, A. & Lohse-Busch, H. ``Control Analysis and
                Thermal Model Development of PHEV,'' SAE 2015-01-1157, SAE World
                Congress, Detroit, April 15; Kim, N., Rousseau, A. & Lohse-Busch, H.
                ``Advanced Automatic Transmission Model Validation Using Dynamometer
                Test Data,'' SAE 2014-01-1778, SAE World Congress, Detroit, Apr.
                14.; Lee, D. Rousseau, A. & Rask, E. ``Development and Validation of
                the Ford Focus BEV Vehicle Model,'' 2014-01-1809, SAE World
                Congress, Detroit, Apr. 14; Kim, N., Kim, N., Rousseau, A., & Duoba,
                M. ``Validating Volt PHEV Model with Dynamometer Test Data using
                Autonomie,'' SAE 2013-01-1458, SAE World Congress, Detroit, Apr.
                13.; Kim, N., Rousseau, A., & Rask, E. ``Autonomie Model Validation
                with Test Data for 2010 Toyota Prius,'' SAE 2012-01-1040, SAE World
                Congress, Detroit, Apr. 12; Karbowski, D., Rousseau, A, Pagerit, S.,
                & Sharer, P. ``Plug-in Vehicle Control Strategy--From Global
                Optimization to Real Time Application,'' 22th International Electric
                Vehicle Symposium (EVS22), Yokohama, (October 2006).
                 \591\ Rousseau, A. Moawad, A. Kim, Namdoo. ``Vehicle System
                Simulation to Support NHTSA CAFE standards for the Draft Tar.''
                https://www.nhtsa.gov/sites/nhtsa.dot.gov/files/anl-nhtsa-workshop-vehicle-system-simulation.pdf. Last accessed Nov 20, 2019.
                ---------------------------------------------------------------------------
                 In fact, using Autonomie, which has validated data based on test
                data from over 50 vehicles, alleviates other stakeholder concerns about
                the level of model validation in past analyses. For example, Global
                Automakers expressed concerns about whether the effectiveness values
                used in past EPA analysis, generated from ALPHA full-vehicle model
                simulations, were properly validated, stating that ``[a]lthough EPA
                claims that the LPM was calibrated based on thorough testing and
                modeling with the ALPHA model, the materials provided with the Proposed
                and Final Determination only cover 18 percent of the projected vehicle
                fleet with regards to specific combinations of powertrain technology
                presented by EPA in the MY 2025 OMEGA pathway. It is unclear how EPA
                calibrated the LPM for the remaining 82 percent of the projected
                vehicles. EPA's failure to publicly share the data for such a large
                percentage of vehicles raises questions about the quality of data.''
                \592\ While simple modeled parameters like single dimensional linear
                systems, such as engine dynamometer torque measurements can be
                validated through other models,\593\ full vehicle systems are complex
                multi-dimensional non-linear systems that need to be developed with
                multiple data sets, and validated with other fully independent data
                sets. Autonomie's models and sub-models have undergone extensive
                validation that has proven the models' agreement with empirical data
                and the principles of physics.
                ---------------------------------------------------------------------------
                 \592\ Docket ID EPA-HQ-OAR-2015-0827-9728. Global later repeated
                that ``only 18% of all vehicle data used as inputs to the ALPHA
                modeling was made available in the EPA's public sources. Additional
                data had to be specifically requested subsequent to the publication
                of the Draft TAR and Proposed Determination. This lack of publicly
                available data highlights transparency concerns, which Global
                Automakers has raised on several previous occasions.''
                 \593\ Section 89.307 Dynamometer calibration.
                ---------------------------------------------------------------------------
                 In addition, the agencies disagree with UCS' comment that EPA's
                direct control over its effectiveness modeling and interface to OMEGA
                results in a more up-to-date analysis. Argonne's participation in
                developing inputs for the rulemaking analysis allowed the agencies
                access to vehicle benchmarking data from more vehicles than if the
                agencies were limited by their own resources, and access to the Argonne
                staff's extensive experience based on direct coordination with vehicle
                manufacturers, suppliers, and researchers that all actively use
                Autonomie for their own work. In addition to Autonomie's continuous
                updates to incorporate the latest fuel-economy-improving technologies,
                discussed throughout this section, the data supplied to and generated
                by Autonomie for use in the CAFE model was continuously updated during
                the analysis process. This is just one part of the iterative quality
                assurance (QA) and quality check (QC) process that the agencies
                developed when Argonne's large-scale simulation modeling based in
                Autonomie was first used for the Draft TAR.
                 In addition to Argonne's team constantly updating Autonomie,
                Argonne's use of high performance computing (HPC) allowed for constant
                update of the analysis during the rulemaking process. Argonne's HPC
                platform allows a full set of simulations--over 750,000 modeled
                vehicles that incorporate over 50 different fuel-economy-improving
                technologies--to be simulated in one week. Subsets of the simulations
                can be re-run should issues come up during QA/QC in a day or less.
                Tools like the internet and high performance computers have allowed the
                agencies to evaluate technology effectiveness with up-to-date inputs
                without the proximity of the computers and the people running them
                working as a detriment the analysis.
                 Finally, GEM, ALPHA, and Autonomie were all developed in the MATLAB
                computational environment as forward-looking physics-based vehicle
                models. Just as ALPHA has roots in GEM, created in 2010 to accompany
                the agencies' heavy-duty vehicle CO2 emissions and fuel
                consumption standards, Autonomie has its origins in the software PSAT,
                developed over 20 years ago. While this information is useful, as
                implied by UCS' comment, the origin of the software was less important
                than the capabilities the software could provide for today's analysis.
                NHTSA's acceptance of GEM
                [[Page 24343]]
                results for compliance with heavy-duty fuel economy regulations had no
                bearing on the decision to use Autonomie to assess the effectiveness of
                light-duty fuel economy and CO2 improving technologies. GEM
                was developed to serve as the compliance model for heavy-duty
                vehicles,\594\ and GEM serves that limited scope very well.
                ---------------------------------------------------------------------------
                 \594\ Newman, K., Dekraker, P., Zhang, H., Sanchez, J. et al.,
                ``Development of Greenhouse Gas Emissions Model (GEM) for Heavy- and
                Medium-Duty Vehicle Compliance,'' SAE Int. J. Commer. Veh.
                8(2):2015, doi:10.4271/2015-01-2771.
                ---------------------------------------------------------------------------
                 UCS did comment that full vehicle simulation could significantly
                improve the estimates of technology effectiveness, but thought it
                critical that the process be as open and transparent as possible. UCS
                pointed to ALPHA results published in peer-reviewed journals as an
                example of how transparency has provided the ALPHA modeling effort with
                significant and valuable feedback, and contrasted what they
                characterized as Autonomie's ``black box'' approach, which they stated
                ``does not lend itself to similar dialog, nor does it make it easy to
                assess the validity of the results.'' Specifically, UCS stated that it
                is ``impossible to verify, replicate, or alter the work done by
                Autonomie due to the expensive nature of the tools used and lack of
                open source or peer-reviewed output.'' In contrast, UCS stated that
                EPA's ALPHA model has been thoroughly peer reviewed, and is readily
                ``downloadable, editable, and accessible to anyone with a MATLAB
                license.''
                 The agencies responses on the merits of how ALPHA and Autonomie
                were peer-reviewed are discussed above. Regarding UCS' comment that it
                is impossible to verify, replicate, or alter the work done by
                Autonomie, the agencies disagree. All inputs, assumptions, model
                documentation--including of component models and individual control
                algorithms--and outputs for the NPRM Autonomie modeling were submitted
                to the docket for review.\595\ Commenters were able to provide a robust
                analysis of Autonomie's technology effectiveness inputs, input
                assumptions, and outputs, as shown by their comments on specific
                vehicle technology effectiveness assumptions, discussed throughout this
                section and in the individual technology sections below.
                ---------------------------------------------------------------------------
                 \595\ NHTSA-2018-0067-1855. ANL Autonomie Compact Car Vehicle
                Class Results. Aug 21, 2018. NHTSA-2018-0067-1856. ANL Autonomie
                Performance Compact Car Vehicle Class Results. Aug 21, 2018. NHTSA-
                2018-0067-1494. ANL Autonomie Midsize Car Vehicle Class Results. Aug
                21, 2018. NHTSA-2018-0067-1487. ANL Autonomie Performance Pick-Up
                Truck Vehicle Class Results. Aug 21, 2018. NHTSA-2018-0067-1663. ANL
                Autonomie Performance Midsize Car Vehicle Class Results. Aug 21,
                2018. NHTSA-2018-0067-1486. ANL Autonomie Small SUV Vehicle Class
                Results. Aug 21, 2018 NHTSA-2018-0067-1662. ANL Autonomie
                Performance Midsize SUV Vehicle Class Results. Aug 21, 2018. NHTSA-
                2018-0067-1661. ANL Autonomie Pickup Truck Vehicle Class Results.
                Aug 21, 2018. NHTSA-2018-0067-1485. ANL Autonomie Small Performance
                SUV Vehicle Class Results. Aug 21, 2018 NHTSA-2018-0067-1492. ANL
                Autonomie Midsize SUV Vehicle Class Results. Aug. 21, 2018. NHTSA-
                2018-0067-0005. ANL Autonomie Model Assumptions Summary. Aug 21,
                2018. NHTSA-2018-0067-0003. ANL Autonomie Summary of Main Component
                Assumptions. Aug 21, 2018. NHTSA-2018-0067-0007. Islam, E. S,
                Moawad, A., Kim, N, Rousseau, A. ``A Detailed Vehicle Simulation
                Process To Support CAFE Standards 04262018--Report'' ANL Autonomie
                Documentation. Aug 21, 2018. NHTSA-2018-0067-0004. ANL Autonomie
                Data Dictionary. Aug 21, 2018. NHTSA-2018-0067-1692. ANL BatPac
                Model 12 55. Aug 21, 2018. NHTSA-2018-0067-12299. Preliminary
                Regulatory Impact Analysis (July 2018). Posted July 2018 and updated
                August 23 and October 16, 2018.
                ---------------------------------------------------------------------------
                 The agencies also disagree with UCS' assessment of Autonomie as
                ``expensive.'' While Autonomie is a commercial product, the biggest
                financial barrier to entry for both ALPHA and Autonomie is the same: A
                MathWorks license.596 597 Regardless, Argonne has made the
                version of Autonomie used for this final rule analysis available upon
                request, including the individual runs used to generate each technology
                effectiveness estimate.\598\
                ---------------------------------------------------------------------------
                 \596\ Autonomie. Frequently Asked Questions. ``Which version of
                matlab can I use?'' https://www.autonomie.net/faq.html#faq2. Last
                accessed Nov. 19, 2019.
                 \597\ EPA ALPHA v2.2 Technology Walk Samples. ``Running this
                version of ALPHA requires Matlab/Simulink with StateFlow 2016b.''
                https://www.epa.gov/regulations-emissions-vehicles-and-engines/advanced-light-duty-powertrain-and-hybrid-analysis-alpha.
                 \598\ Argonne Nationally Laboratory. Autonomie License
                Information. https://www.autonomie.net/asp/LicenseRequest.aspx. Last
                accessed Nov, 18, 2019.
                ---------------------------------------------------------------------------
                 Next, ICCT supplanted its statement that the agencies
                ``inexplicably'' abandoned ALPHA, commenting that the agencies'
                explanation and justification for relying on Autonomie rather than
                ALPHA failed to discuss ALPHA in detail, and the agencies did not
                compare and contrast the two models. ICCT continued, ``the EPA cannot
                select its modeling tool arbitrarily, yet it appeared that the EPA has
                whimsically shifted from an extremely well-vetted, up-to-date,
                industry-grade modeling tool to a less-vetted, academic-grade framework
                with outdated inputs without even attempt to scrutinize the change.''
                ICCT also stated that the agencies are legally obligated to acknowledge
                and explain when they change position, and ``cannot simply ignore that
                EPA previously concluded that the ALPHA modeling accurately projected
                real-world effects of technologies and technology packages.''
                 The agencies disagree that a more in-depth discussion of ALPHA was
                required in the NPRM. In acknowledging the transition to using
                Autonomie for effectiveness modeling and the CAFE model for analysis of
                regulatory alternatives,\599\ the agencies described several analytical
                needs that using a single analysis from the CAFE model--with inputs
                from the Autonomie tool--addressed. These included that Autonomie
                produced realistic estimates of fuel economy levels and CO2
                emission rates through consideration of real-world constraints, such as
                the estimation and consideration of performance, utility, and
                drivability metrics (e.g., towing capability, shift busyness, frequency
                of engine on/off transitions).\600\ That EPA previously concluded the
                ALPHA modeling accurately projected real-world effects of technologies
                and technology packages has no bearing on Autonomie's ability to
                fulfill the analytical needs that the agencies articulated in the NPRM,
                including that Autonomie also accurately projects real-world effects of
                technologies and technology packages.
                ---------------------------------------------------------------------------
                 \599\ 83 FR 43000 (Aug. 24, 2018).
                 \600\ 83 FR 43001 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 The agencies also disagree with ICCT's characterization of ALPHA as
                ``an extremely well-vetted, up-to-date, industry-grade modeling tool''
                and Autonomie as a ``less-vetted, academic-grade framework with
                outdated inputs.'' Again, Autonomie has been used by government
                agencies, vehicle manufacturers (and by agencies and manufacturers
                together in the collaborative government-industry partnership U.S.
                DRIVE program), suppliers, and other organizations because of its
                ability to simulate many powertrain configurations, component
                technologies, and vehicle-level controls over numerous drive cycles.
                Characterizing ALPHA as an ``industry-grade modeling tool'' contravenes
                EPA's own description of its tool--an in-house vehicle simulation model
                used by EPA, not intended to be a commercial product.\601\
                ---------------------------------------------------------------------------
                 \601\ See, e.g., Overview of ALPHA Model, https://www.epa.gov/regulations-emissions-vehicles-and-engines/advanced-light-duty-powertrain-and-hybrid-analysis-alpha; ALPHA Effectiveness Modeling:
                Current and Future Light-Duty Vehicle & Powertrain Technologies
                (Jan. 20, 2016), available at https://www.epa.gov/sites/production/files/2016-10/documents/alpha-model-sae-govt-ind-mtg-2016-01-20.pdf
                (``ALPHA is not a commercial product (e.g. there are no user
                manuals, tech support hotlines, graphical user interfaces, or full
                libraries of components).''). See also Peer Review of ALPHA Full
                Vehicle Simulation Model, available at https://nepis.epa.gov/Exe/ZyPdf.cgi?Dockey=P100PUKT.pdf. While ALPHA peer reviewers found the
                model to be a ``fairly simple transparent model . . . [t]he model
                execution requires an expert MatLab/Simulink user since no user-
                friendly interface currently exists.'' Indeed, EPA noted in response
                to this comment that ``[a]s with any internal tool, EPA does not
                have the need for a ``user-friendly interface'' like one that would
                normally accompany a commercial product which is available for
                purchase and fully supported for wide external usage.''
                ---------------------------------------------------------------------------
                [[Page 24344]]
                 That characterization also contravenes documentation from the
                automotive industry indicating that manufacturers consider ALPHA to
                generate overly optimistic effectiveness values, to be unrepresentative
                of real-world constraints, and a difficult tool to
                use.602 603 The Alliance commented to the MTE
                reconsideration that ``[p]revious comments from the Alliance and
                individual manufacturers to the MTE docket have highlighted multiple
                concerns with EPA's ALPHA model. Many of these concerns remain
                unresolved.'' \604\ Furthermore, the Alliance commented that ALPHA
                ``has not been documented with any instructions making it difficult for
                users outside of EPA to run and interpret the model.'' \605\ Global
                Automakers further stated that the ``lack of publicly available data
                [related to inputs used in the ALPHA modeling] highlights transparency
                concerns, which Global Automakers has raised on several previous
                occasions.'' \606\ In fact, both the Alliance of Automobile
                Manufacturers and Global Automakers, the two trade organizations that
                represent the automotive industry, concluded that Autonomie should be
                used to generate effectiveness inputs for the CAFE model.\607\
                ---------------------------------------------------------------------------
                 \602\ See EPA-HQ-OAR-2015-0827-10125, at 7. As part of their
                assessment that known technologies could not meet the original MY
                2022-2025 standards, Toyota noted that the ALPHA conversion of
                Toyota's MY 2015 to MY 2025 performance ``appears to yield overly
                optimistic results because the powertrain efficiency curves
                represent best-case targets and not the average vehicle, the imposed
                performance constraints are unmarketable, and the generated credits
                are out of sync with product cadence and design cycles.'' See also
                NHTSA-2018-0067-12431, at 7. More recently, Toyota stated in their
                comments to the NPRM that ``Toyota's position [on the efficacy of
                the OMEGA and LPM models] has been clearly represented by comments
                previously submitted by the Alliance of Automobile Manufacturers,
                Global Automakers, and Novation Analytics. Those comments identify
                the LPM and OMEGA models as sources of inaccuracy in EPA technology
                evaluations and provide suggested improvements. Neither model is
                transparent, intuitive, or user friendly.''
                 \603\ EPA-HQ-OAR-2015-0827-9194.
                 \604\ EPA-HQ-OAR-2015-0827-9194, at 33.
                 \605\ EPA-HQ-OAR-2015-0827-9194.
                 \606\ EPA-HQ-OAR-2015-0827-9728.
                 \607\ EPA-HQ-OAR-2015-0827-9163 at 5. (``EPA should abandon the
                lumped-parameter model and instead use NHTSA's Autonomie and Volpe
                models to support the Revised Final Determination.''). See also EPA-
                HQ-OAR-2015-0827-9728 at 15 (stating the EPA's engine mapping and
                tear down analyses ``should be integrated into the Autonomie model,
                which then feeds into the Volpe modeling process.''); EPA-HQ-OAR-
                2015-0827-9194 at 33.
                ---------------------------------------------------------------------------
                 In addition, Autonomie contains up-to-date sub-models to represent
                the latest electrification and advanced transmission and advanced
                engine technologies. As summarized by the Alliance, ``Autonomie was
                developed from the start to address the complex task of combining 2
                power sources in a hybrid powertrain.'' \608\ Autonomie has
                continuously improved over the years by adopting new technologies into
                its modeling framework. Even a small sampling of SAE papers shows how
                Autonomie has been validated to simulate the latest fuel-economy-
                improving technologies like hybrid vehicles and PHEVs.\609\
                ---------------------------------------------------------------------------
                 \608\ Alliance, Docket ID NHTSA-2018-0067-12073 at 135.
                 \609\ Jeong, J., Kim, N., Stutenberg, K., Rousseau, A.,
                ``Analysis and Model Validation of the Toyota Prius Prime,'' SAE
                2019-01-0369, SAE World Congress, Detroit, April 2019; Kim, N,
                Jeong, J. Rousseau, A. & Lohse-Busch, H. ``Control Analysis and
                Thermal Model Development of PHEV,'' SAE 2015-01-1157.
                ---------------------------------------------------------------------------
                 Moreover, Autonomie effectively considers other real-world
                constraints faced by the automotive industry. Vehicle manufacturers and
                suppliers spend significant time and effort to ensure technologies are
                incorporated into vehicles in ways that will balance consumer
                acceptance for attributes such as driving quality,\610\ noise-
                vibration-harshness (NVH), and meeting other regulatory mandates, like
                EPA's and CARB's On-Board Diagnostics (OBD) requirements,\611\ and
                EPA's and CARB's criteria exhaust emissions standards.\612\ The
                implementation of new fuel economy improving technologies have at times
                raised consumer acceptance issues.\613\ As discussed earlier, there are
                diminishing returns for modeling every vehicle attribute and tradeoff,
                as each takes time and incurs cost; however, Autonomie sub-models are
                designed to account for a number of the key attributes and tradeoffs,
                so the resulting effectiveness estimates reflect these real world
                constraints.
                ---------------------------------------------------------------------------
                 \610\ An example of a design requirement is accommodating the
                ``lag'' in torque delivery due to the spooling of a turbine in a
                turbocharged downsized engine. This affects real-world vehicle
                performance, as well as the vehicle's ability to shift during normal
                driving and test cycles.
                 \611\ EPA adopted and incorporated by reference current OBD
                regulations by the California ARB, effective for MY 2017, that cover
                all vehicles except those in the heavier fraction of the heavy-duty
                vehicle class.
                 \612\ Tier 3 emission standards for light-duty vehicles were
                proposed in March 2013 78 FR 29815 (May 21, 2013) and signed into
                law on March 3, 2014 79 FR 23413 (June 27, 2014). The Tier 3
                standards--closely aligned with California LEV III standards--are
                phased-in over the period from MY2017 through MY2025. The regulation
                also tightens sulfur limits for gasoline.
                 \613\ Atiyeh, C. ``What you need to know about Ford's PowerShift
                Transmission Problems'' Car and Driver. July 11, 2019. https://www.caranddriver.com/news/a27438193/ford-powershift-transmission-problems/.
                ---------------------------------------------------------------------------
                 Furthermore, aside from the fact that Autonomie represents the
                structural state-of-the-art in full-vehicle modeling and simulation,
                Autonomie can be populated with any inputs that could be populated in
                the ALPHA model.\614\ The agencies chose to use specific inputs for
                this rulemaking because, as discussed further in Sections VI.C below,
                they best represent the technologies that manufacturers could
                incorporate in the rulemaking timeframe, in a way that balanced
                important concerns like consumer acceptance. Some other examples of how
                Autonomie inputs have been updated with the latest vehicle technology
                data specifically for this analysis include test data incorporated from
                both Argonne and NHTSA-sponsored vehicle benchmarking, including an
                updated automatic transmission skip-shifting feature,\615\ additional
                application of cylinder deactivation for turbocharged downsized
                engines, and as discussed above, new modeling and simulation that
                includes variable compression ratio and Miller Cycle engines.
                ---------------------------------------------------------------------------
                 \614\ For example, Autonomie used the HCR1 and HCR2 engine maps
                used as inputs to ALHPA in the Draft TAR and Proposed Determination.
                 \615\ NHTSA Benchmarking, ``Laboratory Testing of a 2017 Ford F-
                150 3.5 V6 EcoBoost with a 10-speed transmission.'' DOT HS 812 520.
                ---------------------------------------------------------------------------
                 Finally, ICCT commented that the agencies must conduct a systematic
                comparison of the Autonomie modeling system and ALPHA modeling in
                several respects, including the differences in technical inputs and
                resulting efficiency estimates, to explain how the choice of model
                altered the regulatory technology penetration and compliance cost
                estimations, and the differences in modeling methodologies, including
                regarding the relative level of experience of the teams conducting the
                effectiveness modeling, to demonstrate that the choice to use Autonomie
                was not ``due to convenience and easier access by the NHTSA research
                team, rather than for any technical improvement.'' ICCT stated that
                without performing this comparison, ``it otherwise appears that the
                agencies switched from a better-vetted model and system of inputs with
                more recent input data to a less-vetted model and system of inputs as a
                way to bury many dozens of changes without transparency or expert
                assessment (as illustrated in the
                [[Page 24345]]
                above errors and invalidated data on individual technologies).'' Each
                issue is discussed below in turn.
                 First, regarding technical inputs, technology pathways, and
                resulting outputs, ICCT stated that the agencies must compare (1)
                whether the models have been routinely strengthened by incorporating
                cutting edge 2020-2025 automotive technologies to ensure they reflect
                the available improvements; (2) every efficiency technology in the 2016
                Draft TAR and original EPA TSD and Proposed and Final Determination
                analysis against the NPRM; (3) all the major technology package
                pathways (i.e., all combinations with high uptake in the Adopted and
                Augural 2025 standards) in the current NPRM versus the 2016 Draft TAR
                and the 2016 TSD and original Final Determination analysis; (4) each of
                the major 2025 technology package synergies; (5) the modeling work of
                EPA's, Ricardo's, and Argonne's 2014-2018 model year engine
                benchmarking and modeling of top engine and transmission models; and
                ``defend why they appear to have chosen to dismiss the superior and
                better vetted technology modeling approach.''
                 ICCT stated that the agencies must make these comparisons because,
                ``[o]therwise, it seems obvious that the agencies have subjectively
                decided to use the modeling that increases the modeled cost, providing
                further evidence of a high degree of bias without an objective
                accounting of the methodological differences and the sensitivity of the
                results to their new decision.'' Moreover, ICCT stated that ``[b]ecause
                ALPHA is the dominant, preferred, and better-vetted modeling and was
                used in the original Proposed and Final Determination, the agencies are
                responsible for assessing and describing how the use of the ALPHA
                modeling would result in a different regulatory result for their
                analysis of the 2017-2025 adopted [CO2] and Augural CAFE
                standards.''
                 The agencies do not believe that it is necessary to conduct a
                retrospective comparison of ALPHA/LPM and Autonomie effectiveness for
                every technology in the Draft TAR and Proposed Determination to the
                NPRM and final rule analyses, between the two models for technologies
                and packages used in the NPRM and final rule analysis, or to explain
                where and why Autonomie provided different results from ALPHA and the
                LPM, to assess and describe how the use of the ALPHA modeling would
                result in a different regulatory result of CAFE and CO2
                standards, per ICCT's request. While it is anticipated that different
                values will be produced using different tools in an analysis, it is not
                appropriate to select the tool for use based on preferred results. The
                selection of an analysis tool should be based on an evaluation of the
                tool's capabilities and appropriateness for the analysis task. The
                analysis tool should support the full extent of the analysis and
                support the level of input and output resolution required. To compare
                the output of the two models for the purpose of selecting a tool for
                the analysis would likely be biased and disingenuous to the purpose of
                the analysis. In this case, Autonomie was selected for this analysis
                for the reasons discussed throughout this section, and accordingly the
                agencies believe that it was reasonable to consider effectiveness
                estimates developed with Autonomie.
                 That said, comparison of how the tools behave is discussed here to
                further support the agencies' decision process. To demonstrate, in
                addition to everything discussed previously in this section,
                differences in how each model handles powertrain systems modeling with
                specific examples are discussed below as a reference, and differences
                between the agencies' approaches to effectiveness modeling for specific
                technologies is discussed in Section VI.C where appropriate. While the
                improved approach to estimating technology effectiveness estimates
                certainly impacted the regulatory technology penetration, compliance
                cost estimates, and ``major 2025 technology packages and synergies,''
                how technologies are applied in the compliance modeling and the
                associated costs of the technologies is equally as important to
                consider when examining factors that might impact the regulatory
                analysis; that consideration goes beyond the scope of simply
                considering which full vehicle simulation model better performs the
                functions required of this analysis.
                 The agencies have discussed updates to the technologies considered
                in the Autonomie modeling throughout this section, in addition to
                Autonomie's models and sub-models that control advanced technologies
                like hybrid and electrified powertrains. Autonomie's explicit models,
                sub-models, and controls for hybrid and electric vehicles have been
                continuously validated over the past several years,\616\ as Autonomie
                was developed from the beginning to address the complex task of
                combining two power sources in a hybrid powertrain.
                ---------------------------------------------------------------------------
                 \616\ Karbowski, D., Kwon, J., Kim, N., & Rousseau, A.,
                ``Instantaneously Optimized Controller for a Multimode Hybrid
                Electric Vehicle,'' SAE paper 2010-01-0816, SAE World Congress,
                Detroit, April 2010; Sharer, P., Rousseau, A., Karbowski, D., &
                Pagerit, S. ``Plug-in Hybrid Electric Vehicle Control Strategy--
                Comparison between EV and Charge-Depleting Options,'' SAE paper
                2008-01-0460, SAE World Congress, Detroit (April 2008); and
                Rousseau, A., Shidore, N., Carlson, R., & Karbowski, D. ``Impact of
                Battery Characteristics on PHEV Fuel Economy,'' AABC08; Jeong, J.,
                Kim, N., Stutenberg, K., Rousseau, A., ``Analysis and Model
                Validation of the Toyota Prius Prime,'' SAE 2019-01-0369, SAE World
                Congress, Detroit, April 2019; Kim, N, Jeong, J. Rousseau, A. &
                Lohse-Busch, H. ``Control Analysis and Thermal Model Development of
                PHEV,'' SAE 2015-01-1157, SAE World Congress, Detroit, April 15;
                Lee, D. Rousseau, A. & Rask, E. ``Development and Validation of the
                Ford Focus BEV Vehicle Model,'' 2014-01-1809, SAE World Congress,
                Detroit, Apr. 14; Kim, N., Kim, N., Rousseau, A., & Duoba, M.
                ``Validating Volt PHEV Model with Dynamometer Test Data using
                Autonomie,'' SAE 2013-01-1458, SAE World Congress, Detroit, Apr.
                13.; Kim, N., Rousseau, A., & Rask, E. ``Autonomie Model Validation
                with Test Data for 2010 Toyota Prius,'' SAE 2012-01-1040, SAE World
                Congress, Detroit, Apr. 12; Karbowski, D., Rousseau, A, Pagerit, S.,
                & Sharer, P. ``Plug-in Vehicle Control Strategy--From Global
                Optimization to Real Time Application,'' 22th International Electric
                Vehicle Symposium (EVS22), Yokohama, (October 2006).
                ---------------------------------------------------------------------------
                 Also regarding the inputs to both models, as highlighted in Section
                VI.C.3.a), and discussed above, inputs and assumptions for the ALPHA
                modeling used for the EPA Draft TAR and Proposed Determination analysis
                were projected from benchmarking testing. While it is straightforward
                to measure engine fuel consumption and create an engine fuel map, it is
                extremely challenging to identify the specific technologies and levels
                of technologies present on a benchmarking engine. Attributing changes
                in the overall engine fuel consumption to the individual engine
                technologies that make up the complete engine involves significant
                uncertainty.
                 The fixed-point model approach used by the ALPHA model does not
                develop an effectiveness function and assigns a single value to a
                technology. The single value is derived from benchmark testing, which
                often does not isolate the effect of a single technology from the
                effects of other technologies on the tested vehicle. To isolate a
                single technology's effect for use in fixed point modeling properly,
                the agencies would need to benchmark multiple versions of a single
                vehicle, carefully controlling changes to the vehicles' fuel efficiency
                technologies. This process would need to be repeated for a large
                portion of the vehicle fleet and would require significant funding and
                thousands of lab hours to complete. Without this level of data, fixed-
                point effectiveness estimates tend to be too high, as they are unable
                to account for synergetic effects of multiple technologies.
                Specifically, when EPA benchmarks vehicles like the 2018 Toyota Camry,
                the resulting fuel map captures the benefits of many
                [[Page 24346]]
                technologies associated with that engine. This data can be helpful when
                developing controls and validating component operations in modeling,
                but it is inaccurate to conclude that fuel consumption is directly
                related to individual engine technologies, such as lubrication and
                friction reduction, and geometric improvements in efficiency.
                 Contrasted, the NPRM and final rule Autonomie analyses selected
                specific base engine maps and applied technologies incrementally, both
                individually and in known combinations, to better isolate the impacts
                of the technologies. As discussed above, this also implemented NAS
                Recommendation 2.1, to use engine-model-generated maps in the full
                vehicle simulations derived from a validated baseline map in which all
                parameters except the new technology of interest are held
                constant.\617\ While the different methods are valid for different
                purposes, the method selected for the analysis presented today was more
                useful for measuring the incremental effectiveness increments as
                opposed to the absolute values of technology effectiveness, e.g., that
                could be measured by benchmarking a technology package.
                ---------------------------------------------------------------------------
                 \617\ 2015 NAS Report at p. 82.
                ---------------------------------------------------------------------------
                 To provide an example of another difference in behavior between the
                simulation tools, a comparison between ALPHA and Autonomie
                transmissions shifting behavior was conducted. The comparison
                highlighted the differences in how each simulation tool approaches
                transmission shift logic. The ALPHA simulation tool used ALPHAShift.
                ALPHAShift is an optimization algorithm that uses numerous vehicle
                characteristics to find a best shifting strategy. The primary inputs
                for the algorithm includes the fuel consumption (or cost) map for the
                vehicle engine.\618\ Although a public version of ALPHA is available
                for evaluation, the ALPHAShift algorithm used by the tool is hard coded
                with fixed values.619 620 This is an issue, because despite
                peer reviewed documentation on how to tune the algorithm,\621\ no
                documentation of how the algorithm logic works is available for review.
                This is confounding for the use of the software, particularly when the
                observed behavior of the model departs from expected behavior. Figure
                VI-6 below shows simulated gear shift (left) versus actual gear shift
                (right), demonstrating an unexpected shift to neutral before shifting
                to the requested gear.
                ---------------------------------------------------------------------------
                 \618\ Newman, K., Kargul, J., and Barba, D., ``Development and
                Testing of an Automatic Transmission Shift Schedule Algorithm for
                Vehicle Simulation,'' SAE Int. J. Engines 8(3):2015, doi:10.4271/
                2015-01-1142.
                 \619\ Aymeric, R. Islam, E. S. ``Analysis of EPA's ALPHA Shift
                Model--ALPHAShift.'' ANL. March 9, 2020.
                 \620\ ALPHA v2.2 Technology Walk Samples. EPA. January 2017.
                https://www.epa.gov/sites/production/files/2017-01/alpha-20170112.zip. Last Accessed March 9, 2020.
                 \621\ Newman, K., Kargul, J., and Barba, D., ``Development and
                Testing of an Automatic Transmission Shift Schedule Algorithm for
                Vehicle Simulation,'' SAE Int. J. Engines 8(3):2015, doi:10.4271/
                2015-01-1142.
                ---------------------------------------------------------------------------
                 By contrast, and discussed further in VI.C.2 Transmission Paths,
                Autonomie uses a fully documented algorithm to develop a best shifting
                strategy for each unique vehicle configuration. The algorithm develops
                shifting strategies unique to each individual vehicle based on gear
                ratio, final drive ratio, engine BSFC and other vehicle
                characteristics. This is one example of model behavior, in addition to
                the availability of more transparency on this behavior for greater
                stakeholder review, that led the agencies to determine it was
                reasonable and appropriate to use Autonomie for this analysis.
                ---------------------------------------------------------------------------
                 \622\ ALPHA v2.2 Technology Walk Samples. Jan. 12, 2017. https://www.epa.gov/sites/production/files/2017-01/alpha-20170112.zip. Last
                accessed Dec 9, 2019.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.114
                 Regarding the technical expertise of the team conducting the
                ---------------------------------------------------------------------------
                effectiveness modeling, ICCT commented:
                 [T]he agencies should also disclose how much commercial business
                is conducted by the Ricardo, IAV, and Argonne Autonomie teams that
                underpin the modeling of EPA and NHTSA, respectively, including how
                much related research they have done for auto industry clients over
                the past ten years. We mention this because we strongly suspect that
                Ricardo, upon which EPA built its ALPHA model, has done at least an
                order of magnitude (in number of projects, person-hours, and budget)
                more work with and for the automotive industry than the IAV and
                Autonomie teams have in direct work for
                [[Page 24347]]
                automotive industry clients. A conventional government procurement
                effort that competitively vets potential research expert teams would
                presumably have selected for such automotive industry credentials
                and experience, yet it appears that the agencies are wholly
                deferring to Autonomie's less rigorous research-grade modeling
                framework and data due to convenience and easier access by the NHTSA
                research team, rather than for any technical improvement, and this
                is to the detriment of showing clear understanding of real-world
                automotive engineering developments (as demonstrated by many
                erroneous technology combination results throughout these comments).
                 First, NHTSA follows Federal Acquisition Regulation (FAR) to award
                contracts and Interagency Agreements (IAAs),\623\ and any awarded
                contracts and IAAs must follow the FAR requirements. Importantly, FAR
                3.101-1 includes key aspects of conduct and ethics that NHTSA must
                follow in awarding a contract or IAA:
                ---------------------------------------------------------------------------
                 \623\ Federal Acquisition Regulation (FAR). https://www.acquisition.gov/.
                 Government business shall be conducted in a manner above
                reproach and, except as authorized by statute or regulation, with
                complete impartiality and with preferential treatment for none.
                Transactions relating to the expenditure of public funds require the
                highest degree of public trust and an impeccable standard of
                conduct. The general rule is to avoid strictly any conflict of
                interest or even the appearance of a conflict of interest in
                Government-contractor relationships. While many Federal laws and
                regulations place restrictions on the actions of Government
                personnel, their official conduct must, in addition, be such that
                they would have no reluctance to make a full public disclosure of
                their actions.\624\
                ---------------------------------------------------------------------------
                 \624\ FAR 3.101-1.
                 While some factors are more relevant than others in considering
                whether to award a contract or enter into an IAA, the amount of work
                that an organization has performed, characterized by projects, person-
                hours, and budget, is only one of a multitude of factors that is
                considered (if it is even considered at all--an agency might not
                request this information and an organization might decline to provide
                it because of contractual clauses or to protect commercial business
                interests) when assessing whether an organization meets the agency's
                needs for a specific task. Other factors, such as the federal budget,
                also set boundaries for the scope of work that can be performed under
                any competitive government procurement effort.
                 As discussed throughout this section, the team at Argonne National
                Laboratory behind Autonomie has developed and refined a state-of-the-
                art tool that is used by the automotive industry, government agencies,
                and research or other nongovernmental institutions around the world.
                The tool has been and continues to be validated to production vehicles,
                and updated to include models, sub-models, and controls representing
                the state-of-the-art in fuel economy improving technology. To the
                extent that ICCT believes that ``research done for auto industry
                clients,'' ``work with and for the automotive industry,'' and
                ``automotive industry credentials and experience,'' are metrics upon
                which to base this type of important decision, the agencies point ICCT
                to the statements from the automotive industry, above, recommending
                Autonomie be used for technology effectiveness modeling.
                 ICCT concluded that ``[w]hile the agencies are in their process of
                conducting a proper vetting of their NPRM's foundational Autonomie-
                based modeling, we recommend that they rely on what appears to be the
                superior and better vetted technology modeling approach with more
                thorough and state-of-the-art advanced powertrain systems modeling and
                engine maps from the EPA ALPHA modeling.''
                 The agencies properly vetted the Autonomie modeling and decided
                that Autonomie represented a reasonable and appropriate tool to provide
                technology effectiveness estimates for this rulemaking. To the extent
                that commenters' concerns were more about the effectiveness results
                than the tools used to model technology effectiveness, modeling updates
                detailed in the Section VI.B.3.c), below, address those comments. While
                some commenters may still be dissatisfied with Autonomie's technology
                effectiveness estimates, the agencies believe that the refinement of
                inputs and input assumptions, and associated explanation of why those
                refinements are appropriate and reasonable, have appropriately
                addressed comments on these issues. Importantly, none of these
                refinements have led either agency to reconsider using Autonomie for
                this rulemaking analysis.
                 Additional discussion of the agencies' decision to rely on one set
                of modeling tools for this rulemaking is located in Section VI.A of
                this preamble.
                c) Technology Effectiveness Values Implementation in the CAFE Model
                 While the Autonomie model produces a large amount of information
                about each simulation run--for a single technology combination, in a
                single technology class--the CAFE model only uses two elements of that
                information: Battery costs and fuel consumption on the city and highway
                cycles. The agencies combine the fuel economy information from the two
                cycles to produce a composite fuel economy for each vehicle, on each
                fuel. Plug-in hybrids, being the only dual-fuel vehicles in the
                Autonomie simulation, require efficiency estimates of operation on both
                gasoline and electricity--as well as an estimate of the utility factor,
                or the number of miles driven on each fuel. The fuel economy
                information for each technology combination, for each technology class,
                is converted into a single number for use in the CAFE model.
                 As described in greater detail below, each Autonomie simulation
                record represents a unique combination of technologies, and the
                agencies create a technology ``key'' or technology state vector that
                describes all the technology content associated with a record. The 2-
                cycle fuel economy of each combination is converted into fuel
                consumption (gallons per mile) and then normalized relative to the
                starting point for the simulations. In each technology class, the
                combination with the lowest technology content is the VVT (only)
                engine, with a 5-speed transmission, no electrification, and no body-
                level improvements (mass reduction, aerodynamic improvements, or low
                rolling resistance tires). This is the reference point (for each
                technology class) for all the effectiveness estimates in the CAFE
                model. The improvement factors that the model uses are a given
                combination's fuel consumption improvement relative to the reference
                vehicle in its technology class.
                 For the majority of the technologies analyzed within the CAFE
                Model, the fuel economy improvements were derived from the database of
                Autonomie's detailed full-vehicle modeling and simulation results. In
                addition to the technologies found in the Autonomie simulation
                database, the CAFE modeling system also incorporated a handful of
                technologies that were required for CAFE modeling, but were not
                explicitly simulated in Autonomie. The total effectiveness of these
                technologies either could not be captured on the 2-cycle test, or there
                was no robust data that could be used as an input to the full-vehicle
                modeling and simulation, like with emerging technologies such as
                advanced cylinder deactivation (ADEAC). These additional technologies
                are discussed further in Sections VI.B.3 Technology Effectiveness and
                individual technologies sections. For calculating fuel economy
                improvements attributable to these additional technologies, the model
                used defined fuel consumption improvement factors that are constant
                [[Page 24348]]
                across all technology combinations in the database and scale
                multiplicatively when applied together. The Autonomie-simulated and
                additional technologies were then externally combined, forming a single
                dataset of simulation results (referred to as the vehicle simulation
                database, or simply, database), which may then be utilized by the CAFE
                modeling system.
                 To incorporate the results of the combined database of Autonomie-
                simulated and additional technologies, while still preserving the basic
                structure of the CAFE Model's technology subsystem, it was necessary to
                translate the points in this database into corresponding locations
                defined by the technology pathways. By recognizing that most of the
                pathways are unrelated, and are only logically linked to designate the
                direction in which technologies are allowed to progress, it is possible
                to condense the paths into a smaller number of groups based on the
                specific technology. In addition, to allow for technologies present on
                the Basic Engine and Dynamic Road Load (DLR--i.e., MASS, AERO, and
                ROLL) paths to be evaluated and applied in any given combination, a
                unique group was established for each of these technologies.
                 As such, the following technology groups are defined within the
                modeling system: Engine cam configuration (CONFIG), VVT engine
                technology (VVT), VVL engine technology (VVL), SGDI engine technology
                (SGDI), DEAC engine technology (DEAC), non-basic engine technologies
                (ADVENG), transmission technologies (TRANS), electrification and
                hybridization (ELEC), low rolling resistance tires (ROLL), aerodynamic
                improvements (AERO), mass reduction levels (MR), EFR engine technology
                (EFR), electric accessory improvement technologies (ELECACC), LDB
                technology (LDB), and SAX technology (SAX). The combination of
                technologies along each of these groups forms a unique technology state
                vector and defines a unique technology combination that corresponds to
                a single point in the database for each technology class evaluated
                within the modeling system.
                 As an example, a technology state vector describing a vehicle with
                a SOHC engine, variable valve timing (only), a 6-speed automatic
                transmission, a belt-integrated starter generator, rolling resistance
                (level 1), aerodynamic improvements (level 2), mass reduction (level
                1), electric power steering, and low drag brakes, would be specified as
                ``SOHC; VVT; AT6; BISG; ROLL10; AERO20; MR1; EPS; LDB.'' \625\ By
                assigning each unique technology combination a state vector such as the
                one in the example, the CAFE Model can then assign each vehicle in the
                analysis fleet an initial state that corresponds to a point in the
                database.
                ---------------------------------------------------------------------------
                 \625\ In the example technology state vector, the series of
                semicolons between VVT and AT6 correspond to the engine technologies
                which are not included as part of the combination, while the gap
                between MR1 and EPS corresponds to EFR and the omitted technology
                after LDB is SAX. The extra semicolons for omitted technologies are
                preserved in this example for clarity and emphasis, and will not be
                included in future examples.
                ---------------------------------------------------------------------------
                 Once a vehicle is assigned (or mapped) to an appropriate technology
                state vector (from one of approximately three million unique
                combinations, which are defined in the vehicle simulation database as
                CONFIG; VVT; VVL; SGDI; DEAC; ADVENG; TRANS; ELEC; ROLL; AERO; MR; EFR;
                ELECACC; LDB; SAX), adding a new technology to the vehicle simply
                represents progress from a previous state vector to a new state vector.
                The previous state vector simply refers to the technologies that are
                currently in use on a vehicle. The new state vector, however, is
                computed within the modeling system by adding a new technology to the
                combination of technologies represented by the previous state vector,
                while simultaneously removing any other technologies that are
                superseded by the newly added one.
                 For example, consider the vehicle with the state vector described
                as: SOHC; VVT; AT6; BISG; ROLL10; AERO20; MR1; EPS; LDB. Assume the
                system is evaluating PHEV20 as a candidate technology for application
                on this vehicle. The new state vector for this vehicle is computed by
                removing SOHC, VVT, AT6, and BISG technologies from the previous state
                vector,\626\ while also adding PHEV20, resulting in the following:
                PHEV20; ROLL10; AERO20; MR1; EPS; LDB.
                ---------------------------------------------------------------------------
                 \626\ For more discussion of how the CAFE Model handles
                technology supersession, see Section VI.A.7.
                ---------------------------------------------------------------------------
                 From here, it is relatively simple to obtain a fuel economy
                improvement factor for any new combination of technologies and apply
                that factor to the fuel economy of a vehicle in the analysis fleet. The
                formula for calculating a vehicle's fuel economy after application of
                each successive technology represented within the database is defined,
                simply put, as the difference between the fuel economy improvement
                factor associated with the technology state vector before application
                of a candidate technology, and after the application of a candidate
                technology.\627\ This is applied to the original compliance fuel
                economy value for a discrete vehicle in the MY 2017 analysis fleet, as
                discussed previously in Section VI.B.3 Technology Effectiveness.
                ---------------------------------------------------------------------------
                 \627\ For more discussion of how the CAFE Model calculates a
                vehicle's fuel economy where the vehicle switches from one type of
                fuel to another, for example, from gasoline operation to diesel
                operation or from gasoline operation to plug-in hybrid/electric
                vehicle operation, see Section VI.A CAFE Model.
                ---------------------------------------------------------------------------
                 The fuel economy improvement factor is defined in a way that
                captures the incremental improvement of moving between points in the
                database, where each point is defined uniquely as a combination of up
                to 15 distinct technologies describing, as mentioned above, the
                engine's cam configuration, multiple distinct combinations of engine
                technologies, transmission, electrification type, and various vehicle
                body level technologies.
                 Unlike the preceding versions of the modeling system, the current
                version of the CAFE Model relies entirely on the vehicle simulation
                database for calculating fuel economy improvements resulting from all
                technologies available to the system. The fuel economy improvements are
                derived from the factors defined for each unique technology combination
                or state vector. Each time the improvement factor for a new state
                vector is added to a vehicle's existing fuel economy, the factor
                associated with the old technology combination is entirely removed. In
                that sense, application of technologies obtained from the Autonomie
                database is ``self-correcting'' within the model. As such, special-case
                adjustments defined by the previous version of the model are not
                applicable to this one.
                 Meszler Engineering Services, commenting on behalf of Natural
                Resources Defense Council, commented that ``[w]ith very limited
                exception, technology is not included in the NPRM CAFE model if it was
                not included in the simulation modeling that underlies the Argonne
                database,'' citing the ``add-on'' technologies and technologies with
                fixed effectiveness values.\628\ Meszler continued, ``[t]his same
                limitation controls the coupling of technologies, and by extension the
                definition of the CAFE model technology pathways. If a combination of
                technologies were not modeled during the development of the Argonne
                database, that package (or combination) of technologies is not
                available for adoption in the CAFE model. Both of these design
                constraints serve to limit the slate of technologies available to
                respond to fuel economy
                [[Page 24349]]
                standards. The slate of available technologies is basically constrained
                to those included in NHTSA's research activity. If a technology or
                technology combination was not in the NHTSA research planning process,
                it is not available in the model.'' Finally, Meszler stated that
                ``because of the constrained model architecture and the reliance on the
                Argonne database for impact estimates, independently expanding the
                model to include additional technologies or technology combinations is
                not trivial.''
                ---------------------------------------------------------------------------
                 \628\ NHTSA-2018-0067-11723, at 4-5.
                ---------------------------------------------------------------------------
                 We agree that expanding the database to include new technologies is
                not trivial. However, it is possible. The set of available technologies
                is part of the model code, and the code is made public upon each
                release of the model. Many commenters made modifications to the model
                code, conducted additional tests of their own, and presented their
                results to the agencies in the form of public comments before the end
                of the public comment period. A user could add the new technology,
                identify the associated engineering restrictions that determine
                combinations for which that technology should not be considered, and
                add the relevant rows (representing possible technology combinations
                that include the new technology) in the database (which exists locally
                on every computer that runs the model). An enterprising user could also
                take an existing technology along a given path and replace the
                efficiency values with new values--presumably from their own full
                vehicle simulations for each technology combination that contains the
                technology in question. Given the length of time and computing power
                required to simulate vehicle fuel economy on the test cycle for every
                possible combination that could be considered by the CAFE model, using
                a pre-defined database that represents a large ensemble of simulated
                technology combinations is preferable to the alternative of fully
                integrating a vehicle simulation model that would be required to run in
                real-time during the compliance simulation to evaluate the
                effectiveness of every combination considered (not just applied) by the
                model.
                4. Technology Costs
                 In the proposal, the agencies estimated present and future costs
                for fuel-saving technologies, taking into consideration the type of
                vehicle, or type of engine if technology costs vary by application.
                These cost estimates are based on three main inputs. First, the
                agencies estimated direct manufacturing costs (DMCs), or the component
                and labor costs of producing and assembling the physical parts and
                systems, with estimated costs assuming high volume production. DMCs
                generally do not include the indirect costs of tools, capital
                equipment, financing costs, engineering, sales, administrative support
                or return on investment. Second, the agencies accounted for these
                indirect costs via a scalar markup of direct manufacturing costs (the
                retail price equivalent, or RPE). Finally, costs for technologies may
                change over time as industry streamlines design and manufacturing
                processes. The agencies therefore estimated potential cost improvements
                with learning effects (LE). The retail cost of equipment in any future
                year is estimated to be equal to the product of the DMC, RPE, and LE.
                Considering the retail cost of equipment, instead of merely direct
                manufacturing costs, is important to account for the real-world price
                effects of a technology, as well as market realities. Absent a
                government mandate, motor vehicle manufacturers will not undertake
                expensive development and production efforts to implement technologies
                without realistic prospects of consumers being willing to pay enough
                for such technology to allow for the manufacturers to recover their
                investment.
                a) Direct Manufacturing Costs
                 Direct manufacturing costs (DMCs) are the component costs of the
                physical parts and systems that make up a complete vehicle. The
                analysis used agency-sponsored tear-down studies of vehicles and parts
                to estimate the DMCs of individual technologies, in addition to
                independent tear-down studies, other publications, and confidential
                business information. In the simplest cases, the agency-sponsored
                studies produced results that confirmed third-party industry estimates,
                and aligned with confidential information provided by manufacturers and
                suppliers. In cases with a large difference between the tear-down study
                results and credible independent sources, study assumptions were
                scrutinized, and sometimes the analysis was revised or updated
                accordingly.
                 Due to the variety of technologies and their applications, and the
                cost and time required to conduct detailed tear-down analyses, the
                agencies did not sponsor teardown studies for every technology. In
                addition, many fuel-saving technologies were considered that are pre-
                production, or sold in very small pilot volumes. For those
                technologies, a tear-down study could not be conducted to assess costs
                because the product is not yet in the marketplace for evaluation. In
                these cases, the agencies relied upon third-party estimates and
                confidential information from suppliers and manufacturers were relied
                upon; however, there are some common pitfalls with relying on
                confidential business information to estimate costs. The agencies and
                the source may have had incongruent or incompatible definitions of
                ``baseline.'' The source may have provided DMCs at a date many years in
                the future, and assumed very high production volumes, important caveats
                to consider for agency analysis. In addition, a source, under no
                contractual obligation to the agencies, may provide incomplete and/or
                misleading information. In other cases, intellectual property
                considerations and strategic business partnerships may have contributed
                to a manufacturer's cost information and could be difficult to account
                for in the model as not all manufacturer's may have access to
                proprietary technologies at stated costs. The agencies carefully
                evaluated new information in light of these common pitfalls, especially
                regarding emerging technologies.
                 Specifically, the analysis used third-party, forward-looking
                information for advanced cylinder deactivation and variable compression
                ratio engines. While these cost estimates may be preliminary (as is the
                case with many emerging technologies prior to commercialization), the
                agencies consider them to be reasonable estimates of the likely costs
                of these technologies.
                 While costs for fuel-saving technologies reflect the best estimates
                available today, technology cost estimates will likely change in the
                future as technologies are deployed and as production is expanded. For
                emerging technologies, the best information available at the time of
                the analysis was utilized, and cost assumptions will continue to be
                updated for any future analysis. Below, discussion of each category of
                technologies (e.g., engines, transmissions, electrification) summarizes
                comments on corresponding direct cost estimates, and reviews estimates
                the agencies have applied for today's analysis.
                Indirect Costs
                 As discussed above, direct costs represent the cost associated with
                acquiring raw materials, fabricating parts, and assembling vehicles
                with the various technologies manufacturers are expected to use to meet
                future CAFE and CO2 standards. They include materials,
                labor, and variable energy costs required to produce and assemble the
                vehicle. However, they do not
                [[Page 24350]]
                include overhead costs required to develop and produce the vehicle,
                costs incurred by manufacturers or dealers to sell vehicles, or the
                profit manufacturers and dealers make from their investments. All of
                these items contribute to the price consumers ultimately pay for the
                vehicle. These components of retail prices are illustrated in Table VI-
                23 below.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.115
                 In addition to direct manufacturing costs, the agencies estimated
                and considered indirect manufacturing costs. To estimate indirect
                costs, direct manufacturing costs are multiplied by a factor to
                represent the average price for fuel-saving technologies at retail.
                 In the Draft TAR and preceding CAFE and safety rulemaking analyses,
                NHTSA relied on a factor, referred to as the retail price equivalent
                (RPE), to account for indirect manufacturing costs. The RPE accounts
                for indirect costs like engineering, sales, and administrative support,
                as well as other overhead costs, business expenses, warranty costs, and
                return on capital considerations. In the Draft TAR (and subsequent
                Determination) as well as the 2012 rulemaking analysis, EPA applied an
                ``Indirect Cost Multiplier'' (ICM) approach that it first applied in
                the 2010 rulemaking regarding standards for MYs 2012-2016, which also
                accounted for indirect manufacturing costs, albeit in a different way
                than the RPE approach.
                 Some commenters recommended the agencies rely on the ICM approach
                for the current rulemaking, citing EPA's prior peer review and use of
                this approach.\629\ Others supported the agencies' reliance on the RPE
                approach, citing the National Research Council's observations in 2015
                that the ICM approach lacks an empirical basis.\630\ The agencies have
                carefully considered these comments, and conclude that while the ICM
                approach has conceptual merit, its application requires a range of
                specific estimates, and data to support such estimates is scant and, in
                some cases, nonexistent. The agencies have, therefore, applied the RPE
                approach for this final rule, as in the NPRM analysis and other
                rulemaking analyses. The following sections discuss both approaches in
                detail to explain why the RPE approach was chosen for this final rule.
                ---------------------------------------------------------------------------
                 \629\ See, e.g., ICCT, NHTSA-2018-0067-11741, Attachment 3, at
                I-83. See also CFA, NHTSA-2018-0067-12005, Attachment B, at p.189.
                 \630\ See, e.g., Alliance, NHTSA-2018-0067-12073, at 143. See
                also National Research Council, ``Cost, Effectiveness, and
                Deployment of Fuel Economy Technologies for Light-Duty Vehicles,''
                2015, available at https://www.nap.edu/catalog/21744/cost-effectiveness-and-deployment-of-fuel-economy-technologies-for-lightduty-vehicles (``. . . the empirical basis for such multipliers
                is still lacking, and, since their application depends on expert
                judgment, it is not possible for to determine whether the Agencies'
                ICMs are accurate or not'').
                ---------------------------------------------------------------------------
                (1) Retail Price Equivalent
                 Historically, the method most commonly used to estimate indirect
                costs of producing a motor vehicle has been the retail price equivalent
                (RPE). The RPE markup factor is based on an examination of historical
                financial data contained in 10-K reports filed by manufacturers with
                the Securities and Exchange Commission (SEC). It represents the ratio
                between the retail price of motor vehicles and the direct costs of all
                activities that manufacturers engage in, including the design,
                development, manufacturing, assembly,
                [[Page 24351]]
                and sales of new vehicles, refreshed vehicle designs, and modifications
                to meet safety or fuel economy standards.
                 Figure VI-7 indicates that for more than three decades, the retail
                price of motor vehicles has been, on average, roughly 50 percent above
                the direct cost expenditures of manufacturers. This ratio has been
                remarkably consistent, averaging roughly 1.5 with minor variations from
                year to year over this period. At no point has the RPE markup exceeded
                1.6 or fallen below 1.4.\631\ During this time frame, the average
                annual increase in real direct costs was 2.5 percent, and the average
                annual increase in real indirect costs was also 2.5 percent. Figure VI-
                7 illustrates the historical relationship between retail prices and
                direct manufacturing costs.\632\
                ---------------------------------------------------------------------------
                 \631\ Based on data from 1972-1997 and 2007. Data were not
                available for intervening years, but results for 2007 seem to
                indicate no significant change in the historical trend.
                 \632\ Rogozhin, A., Gallaher, M., & McManus, W., 2009,
                Automobile Industry Retail Price Equivalent and Indirect Cost
                Multipliers. Report by RTI International to Office of Transportation
                Air Quality. U.S. Environmental Protection Agency, RTI Project
                Number 0211577.002.004, February, Research Triangle Park, N.C.
                Spinney, B.C., Faigin, B., Bowie, N., & St. Kratzke, 1999, Advanced
                Air Bag Systems Cost, Weight, and Lead Time analysis Summary Report,
                Contract NO. DTNH22-96-0-12003, Task Orders--001, 003, and 005.
                Washington, DC, U.S. Department of Transportation.
                ---------------------------------------------------------------------------
                 An RPE of 1.5 does not imply that manufacturers automatically mark
                up each vehicle by exactly 50 percent. Rather, it means that, over
                time, the competitive marketplace has resulted in pricing structures
                that average out to this relationship across the entire industry.
                Prices for any individual model may be marked up at a higher or lower
                rate depending on market demand. The consumer who buys a popular
                vehicle may, in effect, subsidize the installation of a new technology
                in a less marketable vehicle. But, on average, over time and across the
                vehicle fleet, the retail price paid by consumers has risen by about
                $1.50 for each dollar of direct costs incurred by manufacturers.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.116
                 It is also important to note that direct costs associated with any
                specific technology will change over time as some combination of
                learning and resource price changes occurs. Resource costs, such as the
                price of steel, can fluctuate over time and can experience real long-
                term trends in either direction, depending on supply and demand.
                However, the normal learning process generally reduces direct
                production costs as manufacturers refine production techniques and seek
                out less costly parts and materials for increasing production volumes.
                By contrast, this learning process does not generally influence
                indirect costs. The implied RPE for any given technology would thus be
                expected to grow over time as direct costs decline relative to indirect
                costs. The RPE for any given year is based on direct costs of
                technologies at different stages in their learning cycles, and which
                may have different implied RPEs than they did in previous years. The
                RPE averages 1.5 across the lifetime of technologies of all ages, with
                a lower average in earlier years of a technology's life, and, because
                of learning effects on direct costs, a higher average in later years.
                 The RPE has been used in all NHTSA safety and most previous CAFE
                rulemakings to estimate costs. The National Academy of Sciences
                recommends RPEs of 1.5 for suppliers and 2.0 for in-house production be
                used to estimate total costs. The Alliance of Automobile Manufacturers
                also advocates these values as appropriate markup factors for
                estimating costs of technology changes. An RPE of 2.0 has also been
                adopted by a coalition of environmental and research groups (NESCCAF,
                ICCT, Southwest Research Institute, and TIAX-LLC) in a report on
                reducing heavy truck emissions, and 2.0 is recommended by the U.S.
                Department of Energy for estimating the cost of hybrid-electric and
                automotive fuel cell costs ((see Vyas et al. (2000) in Table VI-24,
                below).
                 Table VI-24 below lists other estimates of the RPE. Note that all
                RPE estimates vary between 1.4 and 2.0, with most in the 1.4 to 1.7
                range.
                [[Page 24352]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.117
                 The RPE has thus enjoyed widespread use and acceptance by a variety
                of governmental, academic, and industry organizations. The RPE has been
                the most commonly used basis for indirect cost markups in regulatory
                analyses. However, as noted above, the RPE is an aggregate measure
                across all technologies applied by manufacturers and is not technology
                specific. A more detailed examination of these technologies is possible
                through an alternative measure, the indirect cost multiplier, which was
                developed to focus more specifically on technologies used to meet CAFE
                and CO2 standards.
                ---------------------------------------------------------------------------
                 \633\ Duleep, K.G. ``2008 Analysis of Technology Cost and Retail
                Price.'' Presentation to Committee on Assessment of Technologies for
                Improving Light Duty Vehicle Fuel Economy, January 25, Detroit, MI.;
                Jack Faucett Associates, September 4, 1985. Update of EPA's Motor
                Vehicle Emission Control Equipment Retail Price Equivalent (RPE)
                Calculation Formula. Chevy Chase, MD--Jack Faucett Associates;
                McKinsey & Company, October 2003. Preface to the Auto Sector Cases.
                New Horizons--Multinational Company Investment in Developing
                Economies, San Francisco, CA.; NRC (National Research Council),
                2002. Effectiveness and Impact of Corporate Average Fuel Economy
                Standards, Washington, DC--The National Academies Press; NRC, 2011.
                Assessment of Fuel Economy Technologies for Light Duty Vehicles.
                Washington, DC--The National Academies Press; Sierra Research, Inc.,
                November 21, 2007, Study of Industry-Average Mark-Up Factors used to
                Estimate Changes in Retail Price Equivalent (RPE) for Automotive
                Fuel Economy and Emissions Control Systems, Sacramento, CA--Sierra
                Research, Inc.; Vyas, A. Santini, D., & Cuenca, R. 2000. Comparison
                of Indirect Cost Multipliers for Vehicle Manufacturing. Center for
                Transportation Research, Argonne National Laboratory, April.
                Argonne, Ill.
                ---------------------------------------------------------------------------
                (2) Indirect Cost Multiplier
                 A second approach to accounting for indirect costs is the indirect
                cost multiplier (ICM). ICMs specifically evaluate the components of
                indirect costs likely to be affected by vehicle modifications
                associated with environmental regulation. EPA developed the ICM concept
                to enable the application of markups more specific to each technology.
                For example, the indirect cost implications of using tires with better
                rolling resistance would not be the same as those for developing an
                entire new hybrid vehicle technology, which would require far more R&D,
                capital investment, and management oversight. With more than 80
                different technologies available to incrementally achieve fuel economy
                improvements,\634\ a wide range of indirect cost effects might be
                expected. ICMs attempt to isolate only those indirect costs that would
                have to change to develop a specific technology. Thus, for example, if
                a company were to hire additional staff to sell vehicles equipped with
                fuel economy improving technology, or to search the technology
                requirements of new CO2 or CAFE standards, the cost of these
                staff would be included in ICMs. However, if these functions were
                accomplished by existing staff, they would not be included. For
                example, if an executive who normally devoted 10 percent of his time to
                fuel economy standards compliance were to devote 50 percent of his time
                in response to new more stringent requirements, his salary would not be
                included in ICMs because he would be paid the same salary regardless of
                whether he devoted his time to addressing CAFE requirements, developing
                new performance technologies, or improving the company's market share.
                ICMs thus do not account for the diverted resources required for
                manufacturers to meet these standards, but rather for the net change in
                costs manufacturers might experience because of hiring additional
                personal or acquiring additional assets or services.
                ---------------------------------------------------------------------------
                 \634\ There are roughly 40 different basic unique technologies,
                but variations among these technologies roughly double the possible
                number of different technology applications.
                ---------------------------------------------------------------------------
                 For past rulemakings EPA developed both short-term and long-term
                ICMs. Long-term ICMs are lower than short-term ICMs. This decline
                reflects the belief that many indirect costs will decline over time.
                For example, research is initially required to develop a new technology
                and apply it throughout the vehicle fleet, but a lower level of
                research will be required to improve, maintain, or adapt that new
                technology to subsequent vehicle designs.
                 While the RPE was derived from data in financial statements
                (reflecting real-world operating and financial results), no similar
                data sources were available to estimate ICMs. ICMs are based on the
                RPE, broken into its components, as shown in Table VI-25. Adjustment
                factors were then developed for those components, based on the
                complexity and time frame of low-, medium-, and high-complexity
                technologies. The adjustment factors were developed from two panels of
                engineers with background in the automobile industry. Initially, a
                group of engineers met and developed an estimate of ICMs for three
                different technologies. This ``consensus'' panel examined one low
                complexity technology, one medium complexity technology, and one high
                complexity technology, with the initial intent of using these
                technologies to represent ICM factors for all technologies falling in
                those categories. At a later date, a second panel was convened to
                examine three more technologies (one low, one medium, and one high
                complexity), using a modified Delphi approach to estimate indirect cost
                effects. The results from the second panel identified the same pattern
                as those of the original report--the indirect cost multipliers increase
                with the
                [[Page 24353]]
                complexity of the technology and decrease over time. The values derived
                in process are higher than those in the RPE/IC Report by values ranging
                from 0.09 (that is, the multiplier increased from 1.20 to 1.29) to 0.19
                (the multiplier increased from 1.45 to 1.64). This variation may be due
                to differences in the technologies used in each panel. The results are
                shown in Figure VI-8, together with the historical average RPE.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.118
                 In subsequent CAFE and CO2 analyses for MYs 2011, as
                well as for the 2012-2016 rulemaking, a simple average of the two
                resulting ICMs in the low and medium technology complexity categories
                was applied to direct costs for all unexamined technologies in each
                specific category. For high complexity technologies, the lower
                consensus-based estimate was used for high complexity technologies
                currently being produced, while the higher modified Delphi-based
                estimate was used for more advanced technologies, such as plug-in
                hybrid or electric vehicles, which had little or no current market
                penetration. Note that ICMs originally did not include profit or
                ``return on capital,'' a fundamental difference from the RPE. However,
                prior to the 2012-2016 CAFE analysis, ICMs were modified to include
                provision for return on capital.
                (3) Application of ICMs in the 2017-2025 Analysis
                 For the model year 2017-2025 rulemaking analysis, NHTSA and EPA
                revisited technologies evaluated by EPA staff and reconsidered their
                method of application. The agencies were concerned that averaging
                consensus and modified Delphi ICMs might not be the most accurate way
                to develop an estimate for the larger group of unexamined technologies.
                Specifically, there was concern that some technologies might not be
                representative of the larger groups they were chosen to represent.
                Further, the agencies were concerned that the values developed under
                the consensus method were not subject to the same analytical discipline
                as those developed from the modified Delphi method. As a result, the
                agencies relied primarily on the modified Delphi-based technologies to
                establish their revised distributions. Thus, for the MY 2017-2025
                analysis, the agencies used the following basis for estimating ICMs:
                 All low complexity technologies were estimated to equal
                the ICM of the modified Delphi-based low technology-passive aerodynamic
                improvements.
                 All medium complexity technologies were estimated to equal
                the ICM of the modified Delphi-based medium technology-engine turbo
                downsizing.
                 Strong hybrids and non-battery plug-in hybrid electric
                vehicles (PHEVs) were estimated to equal the ICM of the high complexity
                consensus-based high technology-hybrid electric vehicle.
                 PHEVs with battery packs and full electric vehicles were
                estimated to equal the ICM of the high complexity modified Delphi-based
                high technology-plug-in hybrid electric vehicle.
                 In addition to shifting the proxy basis for each technology group,
                the agencies reexamined each technology's complexity designation in
                light of the examined technologies that would serve as the basis for
                each group. The resulting designations together with the associated
                proxy technologies are shown in Table VI-25.
                [[Page 24354]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.119
                 Many basic technologies noted in Table VI-25 have variations
                sharing the same complexity designation and ICM estimate. Table VI-26
                lists each technology used in the CAFE model together with their ICM
                category and the year through which the short-term ICM would be
                applied. Note that the number behind each ICM category designation
                refers to the source of the ICM estimate, with 1 indicating the
                consensus panel and 2 indicating the modified Delphi panel.
                BILLING CODE 4910-59-P
                [[Page 24355]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.120
                [[Page 24356]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.121
                [[Page 24357]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.122
                [[Page 24358]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.123
                BILLING CODE 4910-59-C
                 An additional adjustment was made to ICMs to account for the fact
                that they were derived from the RPE analysis for a specific year
                (2007). The agencies believed it would be more appropriate to base ICMs
                on the expected long-term average RPE rather than that of one specific
                year. To account for this, ICMs were normalized to an average RPE
                multiplier level of 1.5.
                 Table VI-27 lists values of ICMs by technology category used in the
                previous MY 2017-2025 rulemaking. As noted previously, the Low 1 and
                Medium 1 categories, which were derived using the initial consensus
                panel, are not used. Short-term values applied to CAFE technologies
                thus range from 1.24 for Low complexity technologies, 1.39 for Medium
                complexity technologies, 1.56 for High1 complexity technologies, and
                1.77 for High2 complexity technologies. When long-term ICMs are applied
                in the year following that noted in the far-right column of Table VI-
                27, these values will drop to 1.19 for Low, 1.29 for Medium, 1.35 for
                High1 and 1.50 for High2 complexity technologies.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.124
                 Note that ICMs for warranty costs are listed separately in Table
                VI-27. This was done because warranty costs are treated differently
                than other indirect costs. In some previous analyses (prior to MY 2017-
                2025), learning was applied directly to total costs. However, the
                agencies believe learning curves are more appropriately applied only to
                direct costs, with indirect costs established up front based on the ICM
                and held constant while direct costs are reduced by learning.
                Warranties are an exception to this because warranty costs involve
                future replacement of defective parts, and the cost of these parts
                would reflect the effect of learning. Warranty costs were thus treated
                as being subject to learning along with direct costs.\635\
                ---------------------------------------------------------------------------
                 \635\ Note that warranty costs also involve labor costs for
                installation. This is typically done at dealerships, and it is
                unlikely labor costs would be subject to learning curves that affect
                motor vehicle parts or assembly costs. However, the portion of these
                costs that is due to labor versus that due to parts is unknown, so
                for this analysis, learning is applied to the full warranty cost.
                ---------------------------------------------------------------------------
                 The effect of learning on direct costs, together with the eventual
                substitution of lower long-term ICMs, causes the effective markup from
                ICMs to differ from the initial ICM on a yearly basis. An example of
                how this occurs is provided in Table VI-28.\636\ This table, which was
                originally developed for the MY 2017-2025 analysis, traces the effect
                of learning on direct costs and its implications for both total costs
                and the ICM-based markup. Direct costs are assigned a value
                (proportion) of 1 to facilitate analysis on the same basis as ICMs (in
                an ICM markup factor, the proportion of direct costs is represented by
                1 while the proportion of indirect costs is represented by the fraction
                of 1 to the right of the decimal.) Table VI-28 examines the effects of
                these factors on turbocharged downsized engines, one of the more
                prevalent CAFE technologies.
                ---------------------------------------------------------------------------
                 \636\ Table VI-22 illustrates the learning process from the base
                year consistent with the direct cost estimate obtained by the
                agencies. It is a mature technology well into the flat portion of
                the learning curve. Note that costs were actually applied in this
                rulemaking example beginning with MY 2017.
                ---------------------------------------------------------------------------
                [[Page 24359]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.125
                 The second column of Table VI-28 lists the learning schedule
                applied to turbocharged downsized engines. Turbocharged downsized
                engines are a mature technology, so the learning schedule captures the
                relatively flat portion of the learning curve occurring after larger
                decreases have already reduced direct costs. The cost basis for
                turbocharged downsized engines in the analysis was effective in 2012,
                so this is the base year for this calculation when direct costs are set
                to 1. The third column shows the progressive decline in direct costs as
                the learning schedule in column 2 is applied to direct costs. Column 4
                contains the value of all indirect costs except warranty. Turbocharged
                downsized engines are a medium-complexity technology, so this value is
                taken from the Medium2 row of Table VI-27. The initial value in 2012 is
                the short-term value, which is used through 2018. During this time,
                these indirect costs are not affected by learning, and they remain
                constant. Beginning in 2019, the long-term ICM from Table VI-27 is
                applied.
                 The fifth column contains warranty costs. As previously mentioned,
                these costs are considered to be affected by learning like direct
                costs, so they decline steadily until the long-term ICM is applied in
                2019, at which point they drop noticeably before continuing their
                gradual decline. In the sixth column, direct and indirect costs are
                totaled. Results indicate a decline in total costs of roughly 30
                percent during this 14-year period. The last column shows the effective
                ICM-based markup, which is derived by dividing total costs by direct
                costs. Over this period, the ICM-based markup rose from the initial
                short-term ICM level of 1.39 to 1.45 in 2018. It then declined to 1.35
                in 2019 when the long-term ICM was applied to the 2019 direct cost.
                Over the remaining years, it gradually rises back up to 1.41 as
                learning continues to degrade direct costs.
                 There are thus two somewhat offsetting processes affecting total
                costs derived from ICMs. The first is the learning curve, which reduces
                direct costs, which raises the effective ICM-based markup. As noted
                previously, learning reflects learned efficiencies in assembly methods
                as well as reduced parts and materials costs. The second is the
                application of a long-term ICM, which reduces the effective ICM-based
                markup. This represents the reduced burden needed to maintain new
                technologies once they are fully developed. In this case, the two
                processes largely offset one another and produce an average real ICM
                over this 14-year period that roughly equals the original short-term
                ICM.
                 Figure VI-9 illustrates this process for each of the 4 technologies
                used to represent the universe of fuel economy and CO2
                improving technologies. As with the turbocharged engines, aerodynamic
                improvements and mild hybrid vehicles show a gradual increase in the
                effective ICM-based markup through the point where the long-term ICM is
                applied. At that time, the ICM-based markup makes an abrupt decline
                before beginning a gradual rise. The decline due to application of
                long-term ICMs is particularly pronounced in the case of the mild
                hybrid--even more so than for the advanced hybrid. The advanced hybrid
                ICM behaves somewhat differently because it is shown through its
                developing stages when more radical learning is applied, but only every
                few years. This produces a significant step-up in ICM levels concurrent
                with each learning
                [[Page 24360]]
                application, followed by a sharp decline when the long-term ICM is
                applied. After that, it begins a gradual rise as more moderate learning
                is applied to reflect its shift to a mature technology. Note that as
                with the turbocharged downsized engine example above, for the
                aerodynamic improvements and mild hybrid technologies, the offsetting
                processes of learning and long-term ICMs result in an average ICM over
                the full time frame that is roughly equal to the initial short-term
                ICM. However, the advanced hybrid ICM rose to a level significantly
                higher than the initial ICM. This is a direct function of the rapid
                learning schedule applied in the early years to this developing
                technology. Brand new technologies might thus be expected to have
                effective lifetime ICM markups exceeding their initial ICMs, while more
                mature technologies are more likely to experience ICMs over their
                remaining life span that more closely approximate their initial ICMs.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.126
                 ICMs for these 4 technologies would drive the indirect cost markup
                rate for the analysis. However, the effect on total costs is also a
                function of the relative incidence of each of the 50+ technologies
                shown in Table VI-26 which are assumed to have ICMs similar to one of
                these 4 technologies. The net effect on costs of these ICMs is also
                influenced by the learning curve appropriate to each technology,
                creating numerous different and unique ICM paths. The average ICM
                applied by the model is also a function of each technology's direct
                cost and because ICMs are applied to direct costs, the measured
                indirect cost is proportionately higher for any given ICM when direct
                costs are higher. The average ICM applied to the fleet for any given
                model year is calculated as follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.127
                where:
                D = direct cost of each technology
                A = application rate for each technology
                ICM = average ICM applied to each technology
                and n = 1, 2 . . . . 88
                 The CAFE model predicts technology application rates assuming
                manufacturers will apply technologies
                [[Page 24361]]
                to meet standards in a logical fashion based on estimated costs and
                benefits. The application rates will thus be different for each model
                year and for each alternative scenario examined. For the MY 2017-2025
                FRIA, to illustrate the effects of ICMs on total technology costs,
                NHTSA calculated the weighted average ICM across all technologies for
                the preferred alternative.\637\ This was done separately for each
                vehicle type and then aggregated based on predicted sales of each
                vehicle type used in the model. Results are shown in Table VI-29.
                ---------------------------------------------------------------------------
                 \637\ For each alternative, this rulemaking examined numerous
                scenarios based on different assumptions, and these assumptions
                could influence the relative frequency of selection of different
                technologies, which in turn could affect the average ICM. The
                scenario examined here assumed a 3 percent discount rate, a 1-year
                payback period, real world application of expected civil penalties,
                and reflects expected voluntary over-compliance by manufacturers.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.128
                 The ICM-based markups in Table VI-29 were derived in a manner
                consistent with the way the RPE is measured, that is, they reflect
                combined influences of direct cost learning and changes in indirect
                cost requirements weighted by both the incidence of each technology's
                adaptation and the relative direct cost of each technology. The results
                indicate generally higher ICMs for passenger cars than for light
                trucks. This is a function of the technologies estimated to be adopted
                for each respective vehicle type, especially in later years when
                hybrids and electric vehicles become more prevalent in the passenger
                car fleet. The influence of these advanced vehicles is driven primarily
                by their direct costs, which greatly outweigh the costs of other
                technologies. This results in the application of much more weight to
                their higher ICMs. This is most notable in MYs 2024 and 2025 for
                passenger cars, when electric vehicles begin to enter the fleet. The
                average ICM increased 0.013 in 2024 primarily because of these
                vehicles. It immediately dropped 0.017 in 2025 because both an
                additional application of steep (20 percent) learning is applied to the
                direct cost of these vehicles (which reduces their relative weight),
                and the long-term ICM becomes effective in that year (which decreases
                the absolute ICM factor). Both influences occur one year after these
                vehicles begin to enter the fleet because of CAFE requirements.
                 ICMs also change over time, again, reflecting the different mix of
                technologies present during earlier years but that are often replaced
                with more expensive technologies in later years. Across all model
                years, the wide-ranging application of diverse technologies required to
                meet CAFE and CO2 standards produced an average ICM-based
                markup (or RPE equivalent) of approximately 1.34, applying only 67
                percent of the indirect costs found in the RPE and implying total costs
                11 percent below those predicted by the RPE-based calculation.
                (4) Uncertainty
                 As noted above, the RPE and ICM assign different markups over
                direct manufacturing costs, and thus imply different total cost
                estimates for CAFE and CO2 technologies. While there is a
                level of uncertainty associated with both markups, this uncertainty
                stems from different issues. The RPE is derived from financial
                statements and is thus grounded in historical data. Although
                compilation of this data is subject to some level of interpretation,
                the two independent researchers who derived RPE estimates from these
                financial reports each reached essentially identical conclusions,
                placing the RPE at roughly 1.5. All other estimates of the RPE fall
                between 1.4 and 2.0, and most are between 1.4 and 1.7. There is thus a
                reasonable level of consistency among researchers that RPEs are 1.4 or
                greater. In addition, the RPE is a measure of the cumulative effects of
                all operations manufacturers undertake in the course of producing their
                vehicles, and is thus not specific to individual technologies, nor of
                CAFE or CO2 technologies in particular. Because this
                provides only a single aggregate measure, using the RPE multiplier
                results in the application of a common incremental markup to all
                technologies. This assures the aggregate cost effect across all
                technologies is consistent with empirical data, but it does not allow
                for indirect cost discrimination among different technologies or over
                time. Because it is applied across all changes, this implies the markup
                for some technologies is likely to be understated, and for others it is
                likely to be overstated.
                 By contrast, the ICM process derives markups specific to several
                CAFE and CO2 technologies, but these markups
                [[Page 24362]]
                have no basis in empirical data. They are based on informed judgment of
                a panel of engineers with auto industry experience regarding cost
                effects of a small sample (roughly 8 percent) of the 50+ technologies
                applied to achieve compliance with CAFE and CO2 standards.
                Uncertainty regarding ICMs is thus based both on the accuracy of the
                initial assessments of the panel on the examined technologies and on
                the assumption that these 4 technologies are representative of the
                remaining technologies that were not examined. Both agencies attempted
                to categorize these technologies in the most representative way
                possible. However, while this represented the best judgment of EPA and
                NHTSA's engineering staffs at that time, the actual effect on indirect
                costs remains uncertain for most technologies. As with RPEs, this means
                that even if ICMs were accurate for the specific technologies examined,
                indirect cost will be understated for some technologies and overstated
                for others.
                 There was considerable uncertainty demonstrated in the ICM panel's
                assessments, as illustrated by the range of estimates among the 14
                modified Delphi panel members surrounding the central values reported
                by the panel. These ranges are shown in Table VI-30 and Figure VI-10,
                Figure VI-11, and Figure VI-12 below. For the low complexity
                technology, passive aerodynamic improvements, panel responses ranged
                from a low of basically no indirect costs (1.001 short term and 1.0
                long term), to a high of roughly a 40 percent markup (1.434 and 1.421).
                For the medium complexity technology, turbo charged and downsized
                engines, responses ranged from a low estimate implying almost no
                indirect cost (1.018 and 1.011), to a high estimate implying that
                indirect costs for this technology would roughly equal the average RPE
                (1.5) for all technologies (1.527 and 1.445). For the high complexity
                technology, plug-in hybrid electric vehicles, responses ranged from a
                low estimate that these vehicles would require significantly less
                indirect cost than the average RPE (1.367 and 1.121) to a high estimate
                implying they would require more indirect costs than the average RPE
                (2.153 and 1.691). There was considerable diversity of opinion among
                the panel members.\638\ This is apparent in Figure VI-10, Figure VI-11,
                and Figure VI-12, which show the 14 panel members' final estimates for
                short-term ICMs as scatter plots.
                ---------------------------------------------------------------------------
                 \638\ Sample confidence intervals, which mitigate the effect of
                outlying opinions, indicate a less extreme but still significant
                range of ICMs. Applying mean ICMs helps mitigate these potential
                differences, but there is clearly a significant level of uncertainty
                regarding indirect costs. A t-distribution is used to estimate
                confidence intervals because of the small sample size (14 panel
                members).
                [GRAPHIC] [TIFF OMITTED] TR30AP20.129
                [[Page 24363]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.130
                [GRAPHIC] [TIFF OMITTED] TR30AP20.131
                [[Page 24364]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.132
                 Although these results were based on modified Delphi panel
                techniques, it is apparent the goal of the Delphi process, an eventual
                consensus or convergence of opinion among panel experts, was not
                achieved. Given this lack of consensus and the divergence of ICM-based
                results from the only available empirical measure (the RPE), there is
                considerable uncertainty that current ICM estimates provide a realistic
                basis of estimating indirect costs. ICMs have not been validated
                through a direct accounting of actual indirect costs for individual
                technologies, and they produce results that conflict with the only
                available empirical evidence of indirect cost markups. Further, they
                are intended to represent indirect costs specifically associated with
                the most comprehensive redesign effort ever undertaken by the auto
                industry, with virtually every make/model requiring ground-up design
                modifications to comply. This includes entirely new vehicle design
                concepts, extensive material substitution, and complete drivetrain
                redesigns, all of which require significant research efforts and
                assembly plant redesign. Under these circumstances, one might expect
                indirect costs to equal or possibly increase above the historical
                average, but not to decrease, as implied by estimated ICMs. For
                regulations, such as the CAFE and CO2 emission standards
                under consideration, that drive changes to nearly every vehicle system,
                the overall average indirect costs should align with the RPE value.
                Applying RPE to the cost for each technology assures that alignment.
                 In the 2015 NAS study, the Committee stated a conceptual agreement
                with the ICM method because ICM takes into account design challenges
                and the activities required to implement each technology. However,
                although endorsing ICMs as a concept, the NAS Committee stated ``the
                empirical basis for such multipliers is still lacking, and, since their
                application depends on expert judgment, it is not possible to determine
                whether the Agencies' ICMs are accurate or not.'' \639\ NAS also stated
                ``the specific values for the ICMs are critical because they may affect
                the overall estimates of costs and benefits for the overall standards
                and the cost effectiveness of the individual technologies.'' \640\ The
                Committee encouraged continued research into ICMs given the lack of
                empirical data for them to evaluate ICMs used by the agencies in past
                analyses. On balance, and considering the relative merits of both
                approaches for realistically estimating indirect costs, the agencies
                consider the RPE method to be a more reliable basis for estimating
                indirect costs.
                ---------------------------------------------------------------------------
                 \639\ National Research Council of the National Academies
                (2015). Cost, Effectiveness, and Deployment of Fuel Economy
                Technologies for Light-Duty Vehicles. https://www.nap.edu/resource/21744/deps_166210.pdf.
                 \640\ Ibid.
                ---------------------------------------------------------------------------
                (5) Using RPE To Evaluate Indirect Costs in This Analysis
                 To ensure overall indirect costs in the analysis align with the
                historical RPE value, the primary analysis has been developed based on
                applying the RPE value of 1.5 to each technology. As noted previously,
                the RPE is the ratio of aggregate retail prices to aggregate direct
                manufacturing costs. The ratio already reflects the mixture of learned
                costs of technologies at various stages of maturity. Therefore, the RPE
                is applied directly to the learned direct cost for each technology in
                each year. This was previously done in the MY 2017-2025 FRIA for the
                preferred alternative for that rulemaking, used in the above analysis
                of average ICMs. Results are shown in Table VI-31.
                 Recognizing there is uncertainty in any estimate of indirect costs,
                a sensitivity analyses of indirect costs has also been conducted by
                applying a lower RPE value as a proxy for the ICM approach. This value
                was derived from a direct comparison of incremental technology costs
                determined in the MY 2017-2025 FRIA.\641\ This analysis is summarized
                in Table VI-31 below. From this table, total costs were estimated to be
                roughly 18 percent lower using ICMs compared to the RPE. As previously
                mentioned, there are two different reasons for these differences. The
                first is the direct effect of applying a higher retail markup. The
                second is an indirect effect resulting from the influence these
                differing markups have on the order of the selection of technologies in
                the CAFE model, which can change as different direct cost levels
                interact with altered retail markups, shifting their relative overall
                effectiveness.
                ---------------------------------------------------------------------------
                 \641\ See Table 5-9a in Final Regulatory Impact Analysis,
                Corporate Average Fuel Economy for MY 2017-MY 2025 Passenger Cars
                and Light Trucks.
                ---------------------------------------------------------------------------
                 The relative effects of ICMs may vary somewhat by scenario, but in
                this case, the application of ICMs produces total
                [[Page 24365]]
                technology cost estimates roughly 18 percent lower than those that
                would result from applying a single RPE factor to all technologies, or,
                conversely, the RPE produces estimates that averaged 21 percent higher
                than the ICM. Under the CAFE model construct, which will apply an
                alternate RPE to the same base technology profile to represent ICMs,
                this implies an RPE equivalent of 1.24 would produce similar net
                impacts [1.5/(1 + x) = 1.21, x = 0.24]. This value is applied for the
                ICM proxy estimate. Additional values were also examined over a range
                of 1.1-2.0. The results, as well as the reference case using the 1.5
                RPE, are summarized in Table VI-32.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.133
                [GRAPHIC] [TIFF OMITTED] TR30AP20.134
                 Several responders submitted comments on the issue of indirect
                costs. The International Council on Clean Transportation (ICCT) stated
                that ``The agencies abandoned their previously-used indirect cost
                multiplier method for estimating total costs, which was vetted with
                peer review, and more complexly handled differing technologies with
                different supply chain and manufacturing aspects. The agencies have, at
                this point, opted to use a simplistic retail price equivalent method,
                which crudely assumes all technologies have a 50 percent markup from
                the direct manufacturing technology cost. We recommend the agencies
                revert back to the previously-used and better substantiated ICM
                approach.'' \642\
                ---------------------------------------------------------------------------
                 \642\ NHTSA-2018-0067-11741.
                ---------------------------------------------------------------------------
                 A private commenter, Thomas Stephens, noted that ``In Section II.
                Technical Foundation for NPRM Analysis, under 1. Data Sources and
                Processes for Developing Individual Technology Assumptions, the
                agencies state that indirect costs are estimated using a Retail Price
                Equivalent (RPE) factor. Concerns with RPE factors and the difficulty
                of accounting for differences in indirect costs of different
                technologies when using this approach were identified by the EPA
                (Rogozhin et al., Using indirect cost multipliers to estimate the total
                cost of adding new technology in the automobile industry, International
                Journal of Production Economics 124, 360-368, 2010), which suggested
                using indirect cost (IC) multipliers instead of RPE factors. The EPA
                developed and updated IC multipliers for relevant vehicle technologies
                with automotive industry input and review. The agencies should consider
                using these IC multipliers to estimate indirect manufacturing costs
                instead of RPE factors.'' \643\
                ---------------------------------------------------------------------------
                 \643\ NHTSA-2018-0067-12067.
                ---------------------------------------------------------------------------
                 By contrast, the Alliance of Automobile Manufacturers (The
                Alliance) ``supports the use of retail
                [[Page 24366]]
                price equivalents in the compliance cost modeling to estimate the
                indirect costs associated with the additional added technology required
                to meet a given future standard. The alternative indirect cost
                multiplier (``ICM'') approach is not sufficiently developed for use in
                rulemaking. As noted by the National Research Council, the indirect
                cost multipliers previously developed by EPA have not been validated
                with empirical data.\644\ Furthermore, in reference to the memorandum
                documenting the development of ICMs previously used by EPA, Exponent
                Failure Analysis Associates found that,
                ---------------------------------------------------------------------------
                 \644\ Cost, Effectiveness, and Development of Fuel Economy
                Technologies for Light-Duty Vehicles, pages 248-49, National
                research Council, the National Academies Press (2015).
                ---------------------------------------------------------------------------
                Past Toyota Comments on Atkinson-Cycle Benefits Have Addressed Only
                Those Derived From Variable Valve Timing
                 In response to these comments the agencies continue to find the RPE
                approach preferable to the ICM approach, at least at this stage in the
                development ICM estimates, for the reasons discussed both above and
                previously in the NPRM. The agencies note that the concerns are not
                with the concept of ICMs, but rather with the judgment-based values
                suggested for use as ICMs, which have not been validated, and which
                conflict with the empirically derived RPE value. The agencies will
                continue to monitor any developments in ICM methodologies as part of
                future rulemakings.
                c) Stranded Capital Costs
                 Past analyses accounted for costs associated with stranded capital
                when fuel economy standards caused a technology to be replaced before
                its costs were fully amortized. The idea behind stranded capital is
                that manufacturers amortize research, development, and tooling expenses
                over many years, especially for engines and transmissions. The
                traditional production life-cycles for transmissions and engines have
                been a decade or longer. If a manufacturer launches or updates a
                product with fuel-saving technology, and then later replaces that
                technology with an unrelated or different fuel-saving technology before
                the equipment and research and development investments have been fully
                paid off, there will be unrecouped, or stranded, capital costs.
                Quantifying stranded capital costs accounts for such lost investments.
                 In the Draft TAR and NPRM analyses, only a few technologies for a
                few manufacturers were projected to have stranded capital costs. As
                more technologies are included in this analysis, and as the CAFE model
                has been expanded to account for platform and engine sharing and
                updated with redesign and refresh cycles, accounting for stranded
                capital has become increasingly complex. Separately, manufacturers may
                be shifting their investment strategies in ways that may alter how
                stranded capital calculations were traditionally considered. For
                example, some suppliers sell similar transmissions to multiple
                manufacturers. Such arrangements allow manufacturers to share in
                capital expenditures, or amortize expenses more quickly.
                 Manufacturers share parts on vehicles around the globe, achieving
                greater scale and greatly affecting tooling strategies and costs. Given
                these trends in the industry and their uncertain effect on capital
                amortization, and given the difficulty of handling this uncertainty in
                the CAFE model, this analysis does not account for stranded capital.
                The agencies' analysis continues to rely on the CAFE model's explicit
                year-by-year accounting for estimated refresh and redesign cycles, and
                shared vehicle platforms and engines, to moderate the cadence of
                technology adoption and thereby limit the implied occurrence of
                stranded capital and the need to account for it explicitly. The
                agencies will monitor these trends to assess the role of stranded
                capital moving forward.
                d) Cost Learning
                 Manufacturers make improvements to production processes over time,
                which often result in lower costs. ``Cost learning'' reflects the
                effect of experience and volume on the cost of production, which
                generally results in better utilization of resources, leading to higher
                and more efficient production. As manufacturers gain experience through
                production, they refine production techniques, raw material and
                component sources, and assembly methods to maximize efficiency and
                reduce production costs. Typically, a representation of this cost
                learning, or learning curves, reflect initial learning rates that are
                relatively high, followed by slower learning as additional improvements
                are made and production efficiency peaks. This eventually produces an
                asymptotic shape to the learning curve, as small percent decreases are
                applied to gradually declining cost levels. These learning curve
                estimates are applied to various technologies that are used to meet
                CAFE standards.
                 For the NPRM and this final rule, the agencies estimated cost
                learning by considering methods established by T.P. Wright \645\ and
                later expanded upon by J.R. Crawford. Wright, examining aircraft
                production, found that every doubling of cumulative production of
                airplanes resulted in decreasing labor hours at a fixed percentage.
                This fixed percentage is commonly referred to as the progress rate or
                progress ratio, where a lower rate implies faster learning as
                cumulative production increases. J.R. Crawford expanded upon Wright's
                learning curve theory to develop a single unit cost model,\646\ that
                estimates the cost of the nth unit produced given the following
                information is known: (1) Cost to produce the first unit; (2)
                cumulative production of n units; and (3) the progress ratio.
                ---------------------------------------------------------------------------
                 \645\ Wright, T.P., Factors Affecting the Cost of Airplanes.
                Journal of Aeronautical Sciences, Vol. 3 (1936), pp. 124-125.
                Available at http://www.uvm.edu/pdodds/research/papers/others/1936/wright1936a.pdf.
                 \646\ Crawford, J.R., Learning Curve, Ship Curve, Ratios,
                Related Data, Burbank, California-Lockheed Aircraft Corporation
                (1944).
                ---------------------------------------------------------------------------
                 As pictured in Figure VI-13, Wright's learning curve shows the
                first unit is produced at a cost of $1,000. Initially cost per unit
                falls rapidly for each successive unit produced. However, as production
                continues, cost falls more gradually at a decreasing rate. For each
                doubling of cumulative production at any level, cost per unit declines
                20 percent, so that 80 percent of cost is retained. The CAFE model uses
                the basic approach by Wright, where cost reduction is estimated by
                applying a fixed percentage to the projected cumulative production of a
                given fuel economy technology.
                [[Page 24367]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.135
                 The analysis accounts for learning effects with model year-based
                cost learning forecasts for each technology that reduce direct
                manufacturing costs over time. The agencies evaluated the historical
                use of technologies, and reviewed industry forecasts to estimate future
                volumes for the purpose of developing the model year-based technology
                cost learning curves.
                 The following section discusses the agencies' development of model
                year-based cost learning forecasts, including how the approach has
                evolved from the 2012 rulemaking for MY 2017-2025 vehicles, and how the
                progress ratios were developed for different technologies considered in
                the analysis. Finally, the agencies discuss how these learning effects
                are applied in the CAFE Model.
                (1) Time Versus Volume-Based Learning
                 For the 2012 joint CAFE/CO2 rulemaking, the agencies
                developed learning curves as a function of vehicle model year.\647\
                Although the concept of this methodology is derived from Wright's
                cumulative production volume-based learning curve, its application for
                CAFE and CO2 technologies was more of a function of time.
                More than a dozen learning curve schedules were developed, varying
                between fast and slow learning, and assigned to each technology
                corresponding to its level of complexity and maturity. The schedules
                were applied to the base year of direct manufacturing cost and
                incorporate a percentage of cost reduction by model year declining at a
                decreasing rate through the technology's production life. Some newer
                technologies experience 20 percent cost reductions for introductory
                model years, while mature or less complex technologies experience 0-3
                percent cost reductions over a few years.
                ---------------------------------------------------------------------------
                 \647\ CAFE 2012 Final Rule, NHTSA DOT, 77 FR 62624.
                ---------------------------------------------------------------------------
                 In their 2015 report to Congress, the National Academy of Sciences
                (NAS) recommended the agencies should ``continue to conduct and review
                empirical evidence for the cost reductions that occur in the automobile
                industry with volume, especially for large-volume technologies that
                will be relied on to meet the CAFE/GHG standards.'' \648\
                ---------------------------------------------------------------------------
                 \648\ Cost, Effectiveness, and Deployment of Fuel Economy
                Technologies for Light-Duty Vehicles, National Research Council of
                the National Academies (2015), available at https://www.nap.edu/resource/21744/deps_166210.pdf.
                ---------------------------------------------------------------------------
                 In response, the agencies have incorporated statically projected
                cumulative volume production data of fuel economy improving
                technologies, representing an improvement over the previously used
                time-based method. Dynamic projections of cumulative production are not
                feasible with current CAFE model capabilities, so one set of projected
                cumulative production data for most vehicle technologies was developed
                for the purpose of determining cost impact. For many technologies
                produced and/or sold in the U.S., historical cumulative production data
                was obtained to establish a starting point for learning schedules.
                Groups of similar technologies or technologies of similar complexity
                may share identical learning schedules.
                 The slope of the learning curve, which determines the rate at which
                cost reductions occur, has been estimated using research from an
                extensive literature review and automotive cost tear-down reports (see
                below). The slope of the learning curve is derived from the progress
                ratio of manufacturing automotive and other mobile source technologies.
                (2) Deriving the Progress Ratio Used in This Analysis
                 Learning curves vary among different types of manufactured
                products. Progress ratios can range from 70 to 100 percent, where 100
                percent indicates no learning can be achieved.\649\ Learning effects
                tend to be greatest in operations where workers often touch the
                product, while effects are less substantial in operations consisting of
                more automated processes. As automotive manufacturing plant processes
                become increasingly automated, a progress ratio towards the higher end
                would seem more suitable. The agencies incorporated findings from
                automotive cost-teardown studies with EPA's literature review of
                learning-related studies to estimate a progress ratio used to determine
                learning schedules of fuel economy improving technologies.
                ---------------------------------------------------------------------------
                 \649\ Martin, J., ``What is a Learning Curve?'' Management and
                Accounting Web, University of South Florida, available at: https://www.maaw.info/LearningCurveSummary.htm.
                ---------------------------------------------------------------------------
                 EPA's literature review examined and summarized 20 studies related
                to learning in manufacturing industries and mobile source
                manufacturing.\650\
                [[Page 24368]]
                The studies focused on many industries, including motor vehicles,
                ships, aviation, semiconductors, and environmental energy. Based on
                several criteria, EPA selected five studies providing quantitative
                analysis from the mobile source sector (progress ratio estimates from
                each study are summarized in Table VI-33, below). Further, those
                studies expand on Wright's Learning Curve function by using cumulative
                output as a predictor variable, and unit cost as the response variable.
                As a result, EPA determined a best estimate of 84 percent as the
                progress ratio in mobile source industries. However, of those five
                studies, EPA at the time placed less weight on the Epple et al. (1991)
                study, because of a disruption in learning due to incomplete knowledge
                transfer from the first shift to introduction of a second shift at a
                North American truck plant. While learning may have decelerated
                immediately after adding a second shift, the agencies note that unit
                costs continued to fall as the organization gained experience operating
                with both shifts. The agencies now recognize that disruptions are an
                essential part of the learning process and should not, in and of
                themselves, be discredited. For this reason, the analysis uses a re-
                estimated average progress ratio of 85 percent from those five studies
                (equally-weighted).
                ---------------------------------------------------------------------------
                 \650\ Cost Reduction through Learning in Manufacturing
                Industries and in the Manufacture of Mobile Sources, United States
                Environmental Protection Agency (2015). Prepared by ICF
                International and available at https://19january2017snapshot.epa.gov/sites/production/files/2016-11/documents/420r16018.pdf.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.136
                 In addition to EPA's literature review, this progress ratio
                estimate was informed based on NHTSA's findings from automotive cost-
                teardown studies. NHTSA routinely performs evaluations of costs of
                previously issued Federal Motor Vehicle Safety Standards (FMVSS) for
                new motor vehicles and equipment. NHTSA's engages contractors to
                perform detailed engineering ``tear-down'' analyses for representative
                samples of vehicles, to estimate how much specific FMVSS add to the
                weight and retail price of a vehicle. As part of the effort, cost and
                production volume are examined for automotive safety technologies. In
                particular, the agency estimated costs from multiple cost tear-down
                studies for technologies with actual production data from the Cost and
                weight added by the Federal Motor Vehicle Safety Standards for MY 1968-
                2012 passenger cars and LTVs (2017).\656\
                ---------------------------------------------------------------------------
                 \651\ Argote, L., Epple, D., Rao, R. D., & Murphy, K., The
                acquisition and depreciation of knowledge in a manufacturing
                organization--Turnover and plant productivity, Working paper,
                Graduate School of Industrial Administration, Carnegie Mellon
                University (1997).
                 \652\ Benkard, C. L., Learning and Forgetting--The Dynamics of
                Aircraft Production, The American Economic Review, Vol. 90(4), pp.
                1034-54 (2000).
                 \653\ Epple, D., Argote, L., & Devadas, R., Organizational
                Learning Curves--A Method for Investigating Intra-Plant Transfer of
                Knowledge Acquired through Learning by Doing, Organization Science,
                Vol. 2(1), pp. 58-70 (1991).
                 \654\ Epple, D., Argote, L., & Murphy, K., An Empirical
                Investigation of the Microstructure of Knowledge Acquisition and
                Transfer through Learning by Doing, Operations Research, Vol. 44(1),
                pp. 77-86 (1996).
                 \655\ Levitt, S. D., List, J. A., & Syverson, C., Toward an
                Understanding of Learning by Doing--Evidence from an Automobile
                Assembly Plant, Journal of Political Economy, Vol. 121 (4), pp. 643-
                81 (2013).
                 \656\ Simons, J. F., Cost and weight added by the Federal Motor
                Vehicle Safety Standards for MY 1968-2012 Passenger Cars and LTVs
                (Report No. DOT HS 812 354). Washington, DC--National Highway
                Traffic Safety Administration (November 2017), at pp. 30-33.
                ---------------------------------------------------------------------------
                 NHTSA chose five vehicle safety technologies with sufficient data
                to estimate progress ratios of each, because these technologies are
                large-volume technologies and are used by almost all vehicle
                manufacturers. Table VI-34 below includes these five technologies and
                yields an average progress rate of 92 percent:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.137
                [[Page 24369]]
                 For a final progress ratio used in the CAFE model, the five
                progress rates from EPA's literature review and five progress rates
                from NHTSA's evaluation of automotive safety technologies results were
                averaged. This resulted in an average progress rate of approximately 89
                percent. Equal weight was placed on progress ratios from all 10
                sources. More specifically, equal weight was placed on the Epple et al.
                (1991) study, because disruptions have more recently been recognized as
                an essential part in the learning process, especially in an effort to
                increase the rate of output. Further discussion of how the progress
                ratios were derived for this analysis is located in FRIA Section 9.
                 ICCT commented that the choice to use safety technology as a model
                for fuel efficiency led to lower learning rates in the NPRM analysis
                compared to prior analyses.\657\ ICCT stated that safety technologies
                were chosen for the NPRM because they are used by almost every
                manufacturer, in contrast to fuel efficiency technologies, where not
                every manufacturer will use them, particularly when they are first
                introduced. ICCT stated that to show the impact of changing learning
                rates, the agencies should run a sensitivity analysis using the
                learning rates in the TAR, as well as EPA's learning rates in its Final
                Determination. ICCT concluded that ``[w]ithout doing so and without
                conducting a peer review of the change in approach, it appears clear
                the agencies have decided to switch to a new costing method that
                affects all future costs, but without any significant research
                justification, vetting, or review.''
                ---------------------------------------------------------------------------
                 \657\ NHTSA-2018-0067-11741.
                ---------------------------------------------------------------------------
                 The agencies' selection of a progress rate of 0.89 is based on an
                average of findings across research and literature reviews conducted by
                NHTSA and EPA. The EPA cited rates were derived from five studies
                selected from a sample of 20 transportation modal learning studies that
                were examined by an EPA contractor, ICF International.\658\ One of
                these 5 studies (Benkard (2000) examines learning in the commercial
                aircraft industry, which the author notes has many unique features that
                influence marginal costs. It also has the lowest progress rate. The
                agencies note that EPA regulates all mobile sources, and while the
                inclusion of non-passenger vehicle studies in their report was
                justified, it may have biased the estimate of learning attributable to
                the motor vehicle industry. Notably, nearly all of the other studies
                included in the ICF International study found progress rates higher
                than the 0.84 rate selected by the authors at that time. In reviewing
                the ICF study, NHTSA found many other studies not included in the
                report, including many specific to the motor vehicle and environmental
                technology industries. Over 90 percent of those studies indicated
                higher progress ratios than ICF recommended.\659\ The agencies' current
                approach includes a broader and more representative sample of these
                studies rather than the narrow sample selected by ICF.
                ---------------------------------------------------------------------------
                 \658\ Cost Reduction through Learning in Manufacturing
                Industries and in the Manufacture of Mobile Sources. United States
                Environmental Protection Agency. Prepared by ICF International and
                available at: https://19january2017snapshot.epa.gov/sites/production/files/2016-11/documents/420r16018.pdf.
                 \659\ See, for example, progress ratios of multiple technologies
                referenced in The Carbon Productivity Challenge: Curbing Climate
                Change and Sustaining Economic Growth, McKinsey Climate Change
                Special Initiative, McKinsey Global Institute, June 2008 (quoting
                from UC Berkeley Energy Resource Group, Navigant Consulting) and
                Technology Innovation for Climate Mitigation and its Relation to
                Government Policies, Edward S. Rubin, Carnegie Mellon University,
                Presentation to the UNFCCC Workshop on Climate Change Mitigation,
                Bonn, Germany, June 19, 2004.
                ---------------------------------------------------------------------------
                 The agencies do not agree that safety technologies are adopted by
                all manufacturers at an early stage. Most safety technologies are
                initially offered as options or standard equipment on only a small
                segment of the vehicle fleet, typically luxury vehicles. After a number
                of years, these technologies may be adopted on less expensive vehicles,
                and eventually they will become required equipment on all vehicles, but
                the production process is gradual, as it is with fuel efficiency
                technologies. FMVSS are necessarily established as performance
                standards--and automakers are free to develop or choose from existing
                technologies to achieve such performance requirements--much like
                automakers can develop or choose from a number of established fuel
                efficiency technologies to achieve fuel economy requirements. Further,
                the derivation of progress ratios is based on the concept of a doubling
                of cumulative production, not time. Therefore, even if production
                continues at a different pace, it should not disqualify non-fuel
                efficiency studies. Moreover, the derivation of the progress ratio used
                in the TAR and Final Determination document were not confined to fuel
                efficiency technologies. In fact, as noted above, they even included at
                least one entirely unrelated study of the aircraft industry.
                 Finally, the agencies note that the previous learning schedules
                used in the TAR and EPA's Final Determination were only developed
                through 2025, whereas this final rule projects learning through 2050.
                The previous learning schedules are thus not directly compatible with
                the analysis conducted in this Final Rule, making a sensitivity
                analysis problematic.
                (3) Obtaining Appropriate Baseline Years for Direct Manufacturing Costs
                To Create Learning Curves
                 Direct manufacturing costs for each fuel economy improving
                technology were obtained from various sources, as discussed above. To
                establish a consistent basis for direct manufacturing costs in the
                rulemaking analysis, each technology cost is adjusted to MY 2018
                dollars. For each technology, the DMC is associated with a specific
                model year, and sometimes a specific production volume, or cumulative
                production volume. The base model year is established as the MY in
                which direct manufacturing costs were assessed (with learning factor of
                1.00). With the aforementioned data on cumulative production volume for
                each technology and the assumption of a 0.89 progress ratio for all
                automotive technologies, the agencies can solve for an implied cost for
                the first unit produced. For some technologies, the agencies used
                modestly different progress ratios to match detailed cost projections
                if available from another source (for instance, batteries for plug-in
                hybrids and battery electric vehicles).
                 This approach produced reasonable estimates for technologies
                already in production, and some additional steps were required to set
                appropriate learning rates for technologies not yet in production.
                Specifically, for technologies not yet in production in MY 2017 (the
                baseline analysis fleet), the cumulative production volume in MY 2017
                is zero, because manufacturers have not yet produced the technologies.
                For pre-production cost estimates in the NPRM, the agencies often
                relied on confidential business information sources to predict future
                costs. Many sources for pre-production cost estimates include
                significant learning effects, often providing cost estimates assuming
                high volume production, and often for a timeframe late in the first
                production generation or early in the second generation of the
                technology. Rapid doubling and re-doubling of a low cumulative volume
                base with Wright's learning curves can provide unrealistic cost
                estimates. In addition, direct manufacturing cost projections can vary
                depending on the initial production volume assumed. Accordingly, the
                agencies carefully examined direct costs with learning, and made
                adjustments to the starting point for those technologies on the
                learning curve to better align
                [[Page 24370]]
                with the assumptions used for the initial direct cost estimate.
                (4) Cost Learning as Applied in the CAFE Model
                 For the NPRM analysis, the agencies updated the manner in which
                learning effects apply to costs. In the Draft TAR analysis, the
                agencies had applied learning curves only to the incremental direct
                manufacturing costs or costs over the previous technology on the
                technology tree. In practice, two things were observed: (1) If the
                incremental direct manufacturing costs were positive, technologies
                could not become less expensive than their predecessors on the
                technology tree, and (2) absolute costs over baseline technology
                depended on the learning curves of root technologies on the technology
                tree. For the NPRM and final rule analysis, the agencies applied
                learning effects to the incremental cost over the null technology state
                on the applicable technology tree. After this step, the agencies
                calculated year-by-year incremental costs over preceding technologies
                on the tech tree to create the CAFE model inputs. As discussed below,
                for the final rule, the agencies revised the CAFE model to replace
                incremental cost estimates with absolute estimates, each specified
                relative to the null technology state on the applicable technology
                tree. This change facilitated quality assurance and is expected to make
                cost inputs more transparently relatable to detailed model output.
                Likewise, this change made it easier to apply learning curves in the
                course of developing inputs to the CAFE model.
                 The agencies grouped certain technologies, such as advanced
                engines, advanced transmissions, and non-battery electric components
                and assigned them to the same learning schedule. While these grouped
                technologies differ in operating characteristics and design, the
                agencies chose to group them based on their complexity, technology
                integration, and economies of scale across manufacturers. The low
                volume of certain advanced technologies, such as hybrid and electric
                technologies, poses a significant issue for suppliers and prevents them
                from producing components needed for advanced transmissions and other
                technologies at more efficient high scale production. The technology
                groupings were carried over from the NPRM analysis for the final rule
                analysis.\660\ Like the NPRM, this final rule analysis uses the same
                groupings that considers market availability, complexity of technology
                integration, and production volume of the technologies that can be
                implemented by manufacturers and suppliers. For example, technologies
                like ADEAC and VCR are grouped together; these technologies were not in
                production or were only in limited introduction in MY 2017, and are
                planned to be introduced in limited production by a few manufacturers.
                The details of these technologies are discussed in Section VI.C.
                ---------------------------------------------------------------------------
                 \660\ See PRIA Chapter 6 for technology groupings.
                ---------------------------------------------------------------------------
                 In addition, for the final rule, as discussed in Section VI.A.4
                Compliance Simulation, the agencies expanded model inputs to extend the
                explicit simulation of technology application through MY 2050, in
                response to comments on the NPRM. Accordingly, the agencies updated the
                learning curves for each technology group to cover MYs through 2050.
                For MYs 2017-2032, the agencies expect incremental improvements in all
                technologies, particularly in electrification technologies because of
                increased production volumes, labor efficiency, improved manufacturing
                methods, specialization, network building, and other factors. While
                these and other factors contribute to continual cost learning, the
                agencies believe that many fuel economy improving technologies
                considered in this rule will approach a flat learning level by the
                early 2030s. Specifically, older and less complex internal combustion
                engine technologies and transmissions will reach a flat learning curve
                sooner when compared to electrification technologies, which have more
                opportunity for improvement. For batteries and non-battery
                electrification components, the agencies estimated a steeper learning
                curve that will gradually flatten after MY 2040. For a more detailed
                discussion of the electrification learning curves used for the final
                rule analysis, see Section VI.C.3.e) Electrification Costs. The
                following Table VI-35 and Table VI-36 show the learning curve schedules
                for CAFE model technologies for MYs 2017-2033 and MYs 2034-2050.
                BILLING CODE 4910-59-P
                [[Page 24371]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.138
                [[Page 24372]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.139
                [[Page 24373]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.140
                [[Page 24374]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.141
                BILLING CODE 4910-59-C
                 Each technology in the CAFE Model is assigned a learning schedule
                developed from the methodology explained previously. For example, the
                [[Page 24375]]
                following chart shows learning rates for several technologies
                applicable to midsize sedans, demonstrating that while the agencies
                estimate that such learning effects have already been almost entirely
                realized for engine turbocharging (a technology that has been in
                production for many years), the agencies estimate that significant
                opportunities to reduce the cost of the greatest levels of mass
                reduction (e.g., MR5) remain, and even greater opportunities remain to
                reduce the cost of batteries for HEVs, PHEVs, BEVs. In fact, for
                certain advanced technologies, the agencies determined that the results
                predicted by the standard learning curves progress ratio was not
                realistic, based on unusual market price and production relationships.
                For these technologies, the agencies developed specific learning
                estimates that may diverge from the 0.89 progress rate. As shown in
                Figure VI-14, these technologies include: Turbocharging and downsizing
                level 1 (TURBO1), variable turbo geometry electric (VTGE), aerodynamic
                drag reduction by 15 percent (AERO15), mass reduction level 5 (MR5), 20
                percent improvement in low-rolling resistance tire technology over the
                baseline, and battery integrated starter/generator (BISG).
                [GRAPHIC] [TIFF OMITTED] TR30AP20.142
                (5) Potential Future Approaches to Considering Cost Learning in the
                CAFE Model
                 As discussed above, cost inputs to the CAFE model incorporate
                estimates of volume-based learning. As an alternative approach, the
                agencies have considered modifications to the CAFE model that would
                calculate degrees of volume-based learning dynamically, responding to
                the model's application of affected technologies. While it is intuitive
                that the degree of cost reduction achieved through experience producing
                a given technology should depend on the actual accumulated experience
                (i.e., volume) producing that technology, such dynamic implementation
                in the CAFE model is thus far infeasible. Insufficient data have been
                available regarding manufacturers' historical application of specific
                technology. Further, insofar as the agencies' estimates of underlying
                direct manufacturing costs already make some assumptions about volume
                and scale, insufficient information is currently available to determine
                how to dynamically adjust these underlying costs. It should be noted
                that if learning responds dynamically to volume, and volume responds
                dynamically to learning, an internally consistent model solution would
                likely require iteration of the CAFE model to seek a stable solution
                within the model's representation of multiyear planning. As discussed
                below, the CAFE model now supports iteration to balance vehicle
                [[Page 24376]]
                cost and fuel economy changes with corresponding changes in sales
                volumes, but, this iteration is not yet implemented in a manner that
                would necessarily support the balance of learning effects on a
                multiyear basis. The agencies invited comment on the issue, seeking
                data and methods that would provide the basis for a practicable
                approach to doing so. Having reviewed comments on cost learning
                effects, the agencies conclude it remains infeasible to calculate
                degrees of volume-based learning in a manner that responds dynamically
                to modeled technology application. The agencies will continue to
                examine this issue for future development.
                e) Cost Accounting
                 The CAFE model applied for the NPRM analysis used an incremental
                approach to specifying technology cost estimates, such that the cost
                for any given technology was specified as an incremental value,
                relative to the technology immediately preceding on the relevant
                technology pathway. For example, the cost of a 7-speed transmission was
                specified as an amount beyond the cost of a 6-speed transmission. This
                approach necessitated careful dynamic accounting for the progressive
                application of the technology as the model worked on a step-by-step
                basis to ``build'' a technology solution. As discussed in the
                corresponding model documentation, the model included complex logic to
                ``back out'' some of these costs carefully when, for example, replacing
                a conventional powertrain with a hybrid-electric system.\661\
                ---------------------------------------------------------------------------
                 \661\ The CAFE Model is available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system with documentation and all inputs and outputs supporting
                today's notice.
                ---------------------------------------------------------------------------
                 To facilitate specification of detailed model inputs and review of
                detailed model outputs, today's CAFE model replaces incremental cost
                inputs with absolute cost inputs, such that the estimated cost of each
                technology is specified relative to a common reference point for the
                relevant technology pathway. For example, the cost of the above-
                mentioned 7-speed transmission is specified relative to a 4-speed
                transmission, as is the cost of every other transmission technology.
                This change in the structure of cost inputs does not, by itself, change
                model results, but it does make the connection between these inputs and
                corresponding outputs more transparent. Model documentation
                accompanying today's analysis presents details of the updated structure
                for model cost inputs.
                5. Other Inputs to the Agencies' Analysis
                 CAFE Model input files described above defining the analysis fleet
                and the fuel-saving technologies to be included in the analysis span
                more than a million records, but deal with a relatively discrete range
                of subjects (e.g., what vehicles are in the fleet, what are the key
                characteristics of those vehicles, what fuel-saving technologies are
                expected to be available, and how might adding those technologies
                impact vehicles' fuel economy levels and costs). The CAFE Model makes
                use of a considerably wider range of other types of inputs, and most of
                these are contained in other model input files. The nature and function
                of many of these inputs remains unchanged relative to the model and
                input files applied for the analysis documented in the proposal that
                preceded today's notice. The CAFE Model documentation accompanying
                today's notice lists and describes all model inputs, and explains how
                inputs are used by the model. Many commenters addressed not only the
                model's function and design, but also specific inputs. Most input
                values are discussed either above (e.g., the preceding subsection
                addresses specific inputs regarding technology costs) or below, in
                subsections discussing specific economic, energy, safety, and
                environmental factors. The remainder of this subsection provides an
                overview of the scope of different model input files. The overview is
                organized based on CAFE Model file types, as in the model
                documentation.
                a) Market Data File
                 The ``Market Data'' file contains the detailed description--
                discussed above--of the vehicle models and model configurations each
                manufacturer produces for sale in the U.S. The file also contains a
                range of other inputs that, though not specific to individual vehicle
                models, may be specific to individual manufacturers. The file contains
                a set of specific worksheets, as follows:
                 ``Manufacturers'' worksheet: Lists specific manufacturers,
                indicates whether manufacturers are expected to prefer paying CAFE
                fines to applying technologies that would not be cost-effective,
                indicates what ``payback period'' defines buyers' willingness to pay
                for fuel economy improvements, enumerates CAFE and CO2
                credits banked from model years prior to those represented explicitly,
                and indicates how sales ``multipliers'' are to be applied when
                simulating compliance with CO2 standards.
                 ``Credits and Adjustments'' worksheet: Enumerates estimates--
                specific to each manufacturer and fleet--of expected CO2 and
                CAFE adjustments reflecting improved AC efficiency, reduced AC
                refrigerant leakage, improvements to ``off cycle'' efficiency, and
                production of flexible fuel vehicles (FFVs). The model applies AC
                refrigerant leakage adjustments only to CO2 levels, and
                applies FFV adjustments only to CAFE levels.
                 ``Vehicles'' worksheet: Lists vehicle models and model
                configurations each manufacturer produces for sale in the U.S.;
                identifies shared vehicle platforms; indicates which engine and
                transmission is present in each vehicle model configuration; specifies
                each vehicle model configuration's fuel economy level, production
                volume, and average price; specifies several engineering
                characteristics (e.g., curb weight, footprint, and fuel tank volume);
                assigns each vehicle model configuration to a regulatory class,
                technology class, engine class, and safety class; specifies schedules
                on which specific vehicle models are expected to be redesigned and
                freshened; specifies how much U.S. labor is involved in producing each
                vehicle model/configuration; and indicates whether specific
                technologies are already present on specific vehicle model
                configurations, or, due to engineering or product planning
                considerations, should be skipped.
                 ``Engines'' worksheet: Identifies specific engines used by each
                manufacturer and for each engine, lists a unique code (referenced by
                the engine code specified for each vehicle model configuration and
                identifies the fuel(s) with which the engine is compatible, the
                valvetrain design (e.g., DOHC), the engine's displacement, cylinder
                configuration and count, and the engine's aspiration type (e.g.,
                naturally aspirated, turbocharged). The worksheet also indicates
                whether specific technologies are already present on specific engines,
                or, due to engineering or product planning considerations, should be
                skipped.
                 ``Transmissions'' worksheet: Similar to the Engines worksheet,
                identifies specific transmissions used by each manufacturer and for
                each transmission, lists a unique code (referenced by the transmission
                code specified for each vehicle model configuration and identifies the
                type (e.g., automatic or CVT) and number of forward gears. Also
                indicates whether specific technologies are already present or, due to
                engineering or product planning considerations, should be skipped.
                [[Page 24377]]
                b) Technologies File
                 The Technologies file identifies about six dozen technologies to be
                included in the analysis, indicates when and how widely each technology
                can be applied to specific types of vehicles, provides most of the
                inputs involved in estimating what costs will be incurred, and provides
                some of the inputs involved in estimating impacts on vehicle fuel
                consumption and weight. The file contains the following types of
                worksheets:
                 ``Parameters'' worksheet: Not to be confused with the
                ``Parameters'' file discussed below, this worksheet in the Technologies
                file indicates, for each technology class, the share of the vehicle's
                curb weight represented by the ``glider'' (the vehicle without the
                powertrain).
                 ``Technologies'' worksheet: For each named technology, specifies
                the share of the entire fleet to which the technology may be
                additionally applied in each model year.
                 Technology Class worksheets: In a separate worksheet for each of
                the 10 technology classes discussed above (and an additional 2--not
                used for this analysis--for heavy-duty pickup trucks and vans),
                identifies whether and how soon the technology is expected to be
                available for wide commercialization, specifies the percentage of miles
                a vehicle is expected to travel on a secondary fuel (if applicable, as
                for plug-in hybrid electric vehicles), indicates a vehicle's expected
                electric power and all-electric range (if applicable), specifies
                expected impacts on vehicle weight, specifies estimates of costs in
                each model year (and factors by which electric battery costs are
                expected to be reduced in each model year), specifies any estimates of
                maintenance and repair cost impacts, and specifies any estimates of
                consumers' willingness to pay for the technology.
                 Engine Type worksheets: In a separate worksheet for each of 28
                initial engine types identified by cylinder count, number of cylinder
                banks, and configuration (DOHC, unless identified as OHV or SOHC),
                specifies estimates of costs in each model year, as well as any
                estimates of impacts on maintenance and repair costs.
                c) Parameters File
                 The ``Parameters'' file contains inputs spanning a range of
                considerations, such as economic and labor utilization impacts, vehicle
                fleet characteristics, fuel prices, scrappage and safety model
                coefficients, fuel properties, and emission rates. The file contains a
                set of specific worksheets, as follows:
                 Economic Values worksheet: Specifies a variety of inputs, including
                social and consumer discount rates to be applied, the ``base year'' to
                which to discount social benefits and costs (i.e., the reference years
                for present value analysis), discount rates to be applied to the social
                cost of CO2 emissions, the elasticity of highway travel with
                respect to per-mile fuel costs (also referred to as the rebound
                effect), the gap between test (for certification) and on-road (aka real
                world) fuel economy, the fixed amount of time involved in each refuel
                event, the share of the tank refueled during an average refueling
                event, the value of travel time (in dollars per hour per vehicle), the
                estimated average number of miles between mid-trip EV recharging events
                (separately for 200 and 300-mile EVs), the rate (in miles of capacity
                per hour of charging) at which EV batteries are recharged during such
                events, the values (in dollars per vehicle-mile) of congestion and
                noise costs, costs of vehicle ownership and operation (e.g., sales
                tax), economic costs of oil imports, estimates of future macroeconomic
                measures (e.g., GDP), and rates of growth in overall highway travel
                (separately for low, reference, and high oil prices).
                 Vehicle Age Data worksheet: Specifies nominal average survival
                rates and annual mileage accumulation for cars, vans and SUVs, and
                pickup trucks. These inputs are used only for displaying estimates of
                avoided fuel savings and CO2 emissions while the model is
                operating. Calculations reported in model output files reflect, among
                other things, application of the scrappage model.
                 Fuel Prices worksheet: Separately for gasoline, E85, diesel,
                electricity, hydrogen, and CNG, specifies historical and estimated
                future fuel prices (and average rates of taxation). Includes values
                reflecting low, reference, and high estimates of oil prices.
                 Scrappage Model Values worksheet: Specifies coefficients applied by
                the scrappage model, which the CAFE Model uses to estimate rates at
                which vehicles will be scrapped (removed from service) during the
                period covered by the analysis.
                 Historic Fleet Data worksheet: For model years not simulated
                explicitly (here, model years through 2016), and separately for cars,
                vans and SUVs, and pickup trucks, specifies the initial size (i.e.,
                number new vehicles produced for sale in the U.S.) of the fleet, the
                number still in service in the indicated calendar year (here, 2016),
                the relative shares of different fuel types, and the average fuel
                economy achieved by vehicles with different fuel types, and the
                averages of horsepower, curb weight, fuel capacity, and price (when
                new).
                 Safety Values worksheet: Specifies coefficients used to estimate
                the extent to which changes in vehicle mass impact highway safety. Also
                specifies statistical value of highway fatalities, the share of
                incremental risk (of any additional driving) internalized by drivers,
                rates relating the cost of damages from non-fatal losses to the cost of
                fatalities, and rates relating the occurrence of non-fatal injuries to
                the occurrence of fatalities.
                 Fatality Rates worksheet: Separately for each model year from 1975-
                2050, and separately for each vehicle age (through 39 years) specifies
                the estimated nominal number of fatalities incurred per billion miles
                of travel by which to offset fatalities.
                 Credit Trading Values worksheet: Specifies whether various
                provisions related to compliance credits are to be simulated (currently
                limited to credit carry-forward and transfers), and specifies the
                maximum number of years credits may be carried forward to future model
                years. Also specifies statutory (for CAFE only) limits on the quantity
                of credit that may be transferred between fleets, and specifies amounts
                of lifetime mileage accumulation to be assumed when adjusting the value
                of transferred credits. Also accommodates a setting indicating the
                maximum number of model years to consider when using expiring credits.
                 Employment Values worksheet: Specifies the estimated average
                revenue OEMs and suppliers earn per employee, the retail price
                equivalent factor applied in developing technology costs, the average
                quantity of annual labor (in hours) per employee, a multiplier to apply
                to U.S. final assembly labor utilization in order to obtain estimated
                direct automotive manufacturing labor, and a multiplier to be applied
                to all labor hours.
                 Fuel Properties worksheet: Separately for gasoline, E85, diesel,
                electricity, hydrogen, and CNG, specifies energy density, mass density,
                carbon content, and tailpipe SO2 emissions (grams per unit
                of energy).
                 Fuel Import Assumptions worksheet: Separately for gasoline, E85,
                diesel, electricity, hydrogen, and CNG, specifies the extent to which
                (a) changes in fuel consumption lead to changes in net imports of
                finished fuel, (b) changes in fuel consumption lead to changes in
                domestic refining output, (c) changes in domestic refining output lead
                to changes in domestic crude oil production, and (d) changes in
                domestic refining output lead to changes in net imports of crude oil.
                [[Page 24378]]
                 Emissions Health Impacts worksheet: Separately for NOX,
                SO2 and PM2.5 emissions, separately for upstream
                and vehicular emissions, and for each of calendar years 2016, 2020,
                2025, and 2030, specifies estimates of various health impacts, such as
                premature deaths, acute bronchitis, and respiratory hospital
                admissions.
                 Carbon Dioxide Emission Costs worksheet: For each calendar year
                through 2080, specifies low, average, and high estimates of the social
                cost of CO2 emissions, in dollars per metric ton.
                Accommodates analogous estimates for CH4 and N2O.
                 Criteria Pollutant Emission Costs worksheet: Separately for
                NOX, SO2 and PM2.5 emissions,
                separately for upstream and vehicular emissions, and for each of
                calendar years 2016, 2020, 2025, and 2030, specifies social costs on a
                per-ton basis.
                 Upstream Emissions (UE) worksheets: Separately for gasoline, E85,
                diesel, electricity, hydrogen, and CNG, and separately for calendar
                years 2017, 2020, 2025, 2030, 2035, 2040, 2045, and 2050, and
                separately for various upstream processes (e.g., petroleum refining),
                specifies emission factors (in grams per million BTU) for each included
                criteria pollutant (e.g., NOX) and toxic air contaminant
                (e.g., benzene).
                 Tailpipe Emissions (TE) worksheets: Separately for gasoline and
                diesel, for each of model years 1975-2050, for each vehicle vintage
                through age 39, specifies vehicle tailpipe emission factors (in grams
                per mile) for CO, VOC, NOX, PM2.5,
                CH4, N2O, acetaldehyde, acrolein, benzene,
                butadiene, formaldehyde, and diesel PM10.
                d) Scenarios File
                 The CAFE Model represents each regulatory alternative as a discrete
                scenario, identifying the first-listed scenario as the baseline
                relative to which impacts are to be calculated. Each scenario is
                described in a worksheet in the Scenarios input file, with standards
                and related provisions specified separately for each regulatory class
                (passenger car or light truck) and each model year. Inputs specify the
                standards' functional forms and defining coefficients in each model
                year. Multiplicative factors and additive offsets are used to convert
                fuel economy targets to CO2 targets, the two being directly
                mathematically related by a linear transformation. Additional inputs
                specify minimum CAFE standards for domestic passenger car fleets,
                determine whether upstream emissions from electricity and hydrogen are
                to be included in CO2 compliance calculations, specify the
                governing rates for CAFE civil penalties, specify estimates of the
                value of CAFE and CO2 credits (for CAFE Model operating
                modes applying these values), specify how flexible fuel vehicles (FFVs)
                and PHEVs are to be accounted for in CAFE compliance calculations,
                specific caps on adjustments reflecting improvements to off-cycle and
                AC efficiency and emissions, specify any estimated amounts of average
                Federal tax credits earned by HEVs, PHEVs, BEVs, and FCVs. The
                worksheets also accommodate some other inputs, such those as involved
                in analyzing standards for heavy-duty pickups and vans, not used in
                today's analysis.
                e) ``Run Time'' Settings
                 In addition to inputs contained in the above-mentioned files, the
                CAFE Model makes use of some settings selected when operating the
                model. These include which standards (CAFE or CO2) are to be
                evaluated; what model years the analysis is to span; when technology
                application is to begin; what ``effective cost'' mode is to be used
                when selecting among technologies; whether use of compliance credits is
                to be simulated and, if so, until what model year; whether dynamic
                economic models are to be exercised and, if so, how many sales model
                iterations are to be undertaken and using what price elasticity;
                whether low, average, or high estimates are to be applied for fuel
                prices, the social cost of carbon, and fatality rates; by how much to
                scale benefits to consumers; and whether to report an implicit
                opportunity cost.
                f) Simulation Inputs
                 As mentioned above, the CAFE Model makes use of databases of
                estimates of fuel consumption impacts and, as applicable, battery costs
                for different combinations of fuel saving technologies. For today's
                analysis, the agencies developed these databases using a large set of
                full vehicle and accompanying battery cost model simulations developed
                by Argonne National Laboratory. To be used as files provided separately
                from the model and loaded every time the model is executed, these
                databases are prohibitively large, spanning more than a million records
                and more than half a gigabyte. To conserve space and speed model
                operation, the agencies have integrated the databases into the CAFE
                Model executable file. When the model is run, however, the databases
                are extracted and placed in an accessible location on the user's disk
                drive. The databases, each of which is in the form of a simple (if
                large) text file, are as follows:
                 ``FE1_Adjustments.csv:'' This is the main database of fuel
                consumption estimates. Each record contains such estimates for a
                specific indexed (using a multidimensional ``key'') combination of
                technologies for each of the technology classes in the Market Data and
                Technologies files. Each estimate is specified as a percentage of the
                ``base'' technology combination for the indicated technology class.
                 ``FE2_Adjustments.csv:'' Specific to PHEVs, this is a database of
                fuel consumption estimates applicable to operation on electricity,
                specified in the same manner as those in the main database.
                 ``Battery_Costs.csv:'' Specific to technology combinations
                involving vehicle electrification (including 12V stop-start systems),
                this is a database of estimates of corresponding base costs (before
                learning effects) for batteries in these systems.
                g) On Road Fuel Economy and CO2 Emissions Gap
                 Rather than rely on the compliance values of fuel economy for
                either historical vehicles or vehicles that go through the full
                compliance simulation, the model applies an ``on-road gap'' to
                represent the expected difference between fuel economy on the
                laboratory test cycle and fuel economy under real-world operation. In
                other words, all of the reported physical impacts analysis (including
                emissions impacts) are based on actual real world fuel consumption and
                emissions, not on values based on 2-cycle fuel economy ratings and
                CO2 emission rates, nor on regulatory incentives such as
                sales multipliers that treat a single vehicle as two vehicles, or that
                set aside emissions resulting from generation of electricity to power
                electric vehicles. This was a topic of interest in the recent peer
                review of the CAFE model. While the model currently allows the user to
                specify an on-road gap that varies by fuel type (gasoline, E85, diesel,
                electricity, hydrogen, and CNG), it does not vary over time, by vehicle
                age, or by technology combination. It is possible that the ``gap''
                between laboratory fuel economy and real-world fuel economy has changed
                over time, that fuel economy changes as a vehicle ages, or that
                specific combinations of fuel-saving technologies have a larger
                discrepancy between laboratory and real-world fuel economy than others.
                For today's analysis, and considering data EPA collects from
                manufacturers regarding vehicles' fuel economy and CO2 as
                tested for both fuel economy and emissions compliance and for vehicle
                fuel economy and emissions labeling
                [[Page 24379]]
                (labeling making use of procedures spanning a wider range of real-world
                vehicle operating conditions), the agencies have determined that the
                future gap is, at this time, best estimated using the same values
                applied for the analysis documented in the NPRM. The agencies will
                continue to assess such test data and any other available data
                regarding real-world fuel economy and emissions and, as warranted, will
                revise methods and inputs representing the gap between laboratory and
                real-world fuel economy and CO2 emissions in future
                rulemakings. The sensitivity analysis summarized in the FRIA
                accompanying the final rule includes cases representing narrower and
                wider gaps.
                C. The Model Applies Technologies Based on a Least-Cost Technology
                Pathway to Compliance, Given the Framework Above
                 The CAFE model, discussed in detail above, is designed to simulate
                compliance with a given set of CAFE or tailpipe CO2
                emissions standards for each manufacturer that sells vehicles in the
                United States. For the final rule analysis, the model began with a
                representation of the MY 2017 vehicle model offerings for each
                manufacturer that included the specific engines and transmissions on
                each model variant, observed sales volumes, and all fuel economy
                improving technology that is already present on those vehicles. From
                there the model added technology, in response to the standards being
                considered, in a way that minimized the cost of compliance and
                reflected many real-world constraints faced by automobile
                manufacturers. The model addressed fleet year-by-year compliance,
                taking into consideration vehicle refresh and redesign schedules and
                shared platforms, engines, and transmissions among vehicles.
                 The agencies evaluated a wide array of technologies manufacturers
                could use to improve the fuel economy of new vehicles, in both the
                immediate future and during the timeframe of this rulemaking, to meet
                the fuel economy and CO2 standards. The agencies evaluated
                costs for these technologies, and looked at how costs may change over
                time. The agencies also considered how fuel-saving technologies may be
                used on many types of vehicles (ranging from small cars to trucks) and
                how the technologies may perform in improving fuel economy and
                CO2 emissions in combination with other technologies. With
                cost and effectiveness estimates for technologies, the agencies
                forecast how manufacturers may respond to potential standards and can
                estimate the associated costs and benefits related to technology and
                equipment changes. This assists the assessment of technological
                feasibility and is a building block for the consideration of economic
                practicability of the standards.
                 The agencies described in the NPRM that the characterization of
                current and anticipated fuel-saving technologies relied on portions of
                the analysis presented in the Draft TAR, in addition to new information
                that had been gathered and developed since conducting that analysis,
                and the significant, substantive input that was received during the
                Draft TAR comment period.\662\ The Draft TAR considered many
                technologies previously assessed in the 2012 final rule; \663\ in some
                cases, manufacturers have nearly universally adopted a technology in
                today's new vehicle fleet (for example, electric power steering), but
                in other cases, manufacturers only occasionally use a technology in
                today's new vehicle fleet (like turbocharged engines). For a few
                technologies considered in the 2012 rulemaking, manufacturers began
                implementing the technologies but have since largely pivoted to other
                technologies due to consumer acceptance issues (for instance,
                drivability and performance feel issues associated with some dual
                clutch transmissions without a torque converter) or limited commercial
                success.
                ---------------------------------------------------------------------------
                 \662\ 83 FR 43021-22 (Aug. 24, 2018).
                 \663\ 77 FR 62624 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 In some cases, EPA and NHTSA presented different analytical
                approaches in the Draft TAR. However, for the NPRM and final rule
                analysis, the agencies harmonized their analytical approach to use one
                set of effectiveness values (developed with one tool), one set of cost
                assumptions, and one set of assumptions about the limitations of some
                technologies. To develop these assumptions, the agencies evaluated many
                sources of data, in addition to many stakeholder comments received on
                the Draft TAR. The preferred approach was to harmonize on sources and
                methodologies that were data-driven and reproducible for independent
                verification, produced using tools utilized by OEMs, suppliers, and
                academic institutions, and using tools that could support both CAFE and
                CO2 analysis. As the agencies noted in the NPRM, a single
                set of assumptions also facilitated and focused public comment by
                reducing burden on stakeholders who sought to review all of the
                supporting documentation surrounding the analysis.
                 The agencies also identified a preference to use values developed
                from careful review of commercialized technologies; however, in some
                cases for technologies that are new, and are not yet for sale in any
                vehicle, the analysis relied on information from other sources,
                including CBI and third-party research reports and publications. The
                agencies strived to keep the technology analysis as current as possible
                in light of the ongoing technology development and implementation in
                the automotive industry. Additional emerging technologies added for the
                final rule analysis are described in further detail, below.
                 The agencies' process to develop effectiveness assumptions is
                described in detail in Section VI.B.3 Technology Effectiveness, and
                summarized here. The NPRM and final rule analysis modeled combinations
                of more than 50 fuel economy-improving technologies across 10 vehicle
                types (an increase from five vehicle types in NHTSA's Draft TAR
                analysis). Only 10 vehicle technology classes were used because large
                portions of the production volume in the analysis fleet have similar
                specifications, especially in highly competitive segments. For
                instance, many mid-sized sedans, small SUVs, and large SUVs coalesce
                around similar specifications, respectively. Baseline simulations have
                been aligned around these modal specifications. Parametrically
                combining these technologies generated more than 100,000 unique
                combinations per vehicle class. Multiplying the unique technology
                combinations by the 10 technology classes resulted in the simulation of
                more than one million individual full-vehicle system models. Modeling
                was also conducted to determine appropriate levels of engine downsizing
                required to maintain baseline vehicle performance when advanced mass
                reduction technology or advanced engine technology were applied.
                Performance neutrality is discussed in detail in VI.B.3.
                 Some baseline vehicle assumptions used in the simulation modeling
                were updated since the Draft TAR based on public comments, and further
                assessment of the NPRM and final rule analysis fleets. The agencies
                updated assumptions about curb weight, as well as technology properties
                like baseline rolling resistance, aerodynamic drag coefficients, and
                frontal areas. Many of the assumptions are aligned with published
                research from the Department of Energy and other independent
                [[Page 24380]]
                sources.\664\ Additional transmission technologies and more levels of
                aerodynamic technologies than NHTSA presented in the Draft TAR analysis
                were also added for the analysis. Having additional technologies in the
                model allowed the agencies to assign baselines and estimate fuel-
                savings opportunities with more precision.
                ---------------------------------------------------------------------------
                 \664\ See, e.g., Islam, E., A. Moawad, N. Kim, and A. Rousseau,
                2018a, An Extensive Study on Vehicle Sizing, Energy Consumption and
                Cost of Advance Vehicle Technologies, Report No. ANL/ESD-17/17,
                Argonne National Laboratory, Lemont, Ill., Oct 2018. https://www.autonomie.net/pdfs/ANL_BaSce_FY17_Report_10042018.pdf. Last
                accessed March 18, 2020; Pannone, G. ``Technical Analysis of Vehicle
                Load Reduction Potential for Advanced Clean Cars,'' April 29, 2015.
                Available at https://www.arb.ca.gov/research/apr/past/13-313.pdf.
                Last accessed December 28, 2019.
                ---------------------------------------------------------------------------
                 To develop technology cost assumptions, the agencies estimated
                present and future costs for fuel-saving technologies, taking into
                consideration the type of vehicle, or type of engine if technology
                costs vary by application. Since the 2012 final rule, many cost
                assessments, including tear down studies, were funded and completed,
                and presented as part of the Draft TAR analysis. These studies
                evaluated transmissions, engines, hybrid technologies, and mass
                reduction.\665\ The NPRM and final rule analyses use the 2016 Draft
                TAR's cost estimates for many technologies. In addition to those
                studies, the analysis also leveraged research reports from other
                organizations to assess costs.\666\ Consistent with past analyses, this
                analysis used BatPaC to provide estimates for future battery costs for
                hybrids, plug-in hybrids, and electric vehicles, taking into account
                the different battery design characteristics and taking into account
                the size of the battery for different applications.\667\ The agencies
                also updated technology costs for the NPRM to 2016 dollars, because, as
                in many cases, technology costs were estimated several years ago, and
                since then have further updated technology costs to 2018 dollars for
                the final rule.
                ---------------------------------------------------------------------------
                 \665\ FEV prepared several cost analysis studies for EPA on
                subjects ranging from advanced 8-speed transmissions to belt
                alternator starter, or Start/Stop systems. NHTSA also contracted
                with Electricore and EDAG on teardown studies evaluating mass
                reduction. The 2015 NAS report on fuel economy technologies for
                light-duty vehicles also evaluated the agencies' technology costs
                developed based on these teardown studies, and the technology costs
                used in this proposal were updated accordingly.
                 \666\ For example, the agencies relied on reports from the
                Department of Energy's Office of Energy Efficiency & Renewable
                Energy's Vehicle Technologies Office. More information on that
                office is available at https://www.energy.gov/eere/vehicles/vehicle-technologies-office. Other agency reports that were relied on for
                technology or other information are referenced throughout the NPRM
                and accompanying PRIA, and this final rule and the accompanying
                FRIA.
                 \667\ For instance, battery electric vehicles with high levels
                of mass reduction may use a smaller battery than a comparable
                vehicle with less mass reduction technology and still deliver the
                same range on a charge. See, e.g., Ward, J. & Gohlke, D. & Nealer,
                Rachael. (2017). The Importance of Powertrain Downsizing in a
                Benefit-Cost Analysis of Vehicle Lightweighting. JOM. 69.
                ---------------------------------------------------------------------------
                 Cost and effectiveness values were estimated for each technology
                included in the analysis. As mentioned above, more than 50 technologies
                were considered in the NPRM and final rule analyses, and the agencies
                evaluated many combinations of these technologies in many applications.
                In the NPRM, the agencies identified overarching potential issues in
                assessing technology effectiveness and cost, including:
                 Baseline vehicle technology level assessed as too low, or
                too high. Compliance information was extensively reviewed and
                supplemented with available literature on the vehicle models considered
                in the analysis fleet. Manufacturers could also review the baseline
                technology assignments for their vehicles, and the analysis
                incorporates feedback received from manufacturers.
                 Technology costs too low or too high. Tear down cost
                studies, CBI, literature, and the 2015 NAS study information were
                referenced to estimate technology costs. In cases where one technology
                appeared to exceed all other technologies on cost and effectiveness,
                information was acquired from additional sources to confirm or reject
                assumptions. Cost assumptions for emerging technologies were reassessed
                in cases where new information became available.
                 Technology effectiveness too high or too low in
                combination with other vehicle technologies. Technology effectiveness
                was evaluated using the Autonomie full-vehicle simulation modeling,
                taking into account the impact of other technologies on the vehicle and
                the vehicle type. Inputs and modeling for the analysis took into
                account laboratory test data for production and some pre-production
                technologies, technical publications, manufacturer and supplier CBI,
                and simulation modeling of specific technologies. Evaluating recently
                introduced production products to inform the technology effectiveness
                models of emerging technologies was preferred; however, some
                technologies that are not yet in production were considered using CBI.
                Simulation modeling used carefully chosen baseline configurations to
                provide a consistent, reasonable reference point for the incremental
                effectiveness estimates.
                 Vehicle performance not considered or applied in an
                infeasible manner. Performance criteria, including low speed
                acceleration (0-60 mph time), high speed acceleration (50-80 mph time),
                towing, and gradeability (six percent grade at 65 mph) were also
                considered. In the simulation modeling, resizing was applied to achieve
                the same performance level as the baseline for the least capable
                performance criteria but only with significant design changes. The
                analysis struck a balance by employing a frequency of engine downsizing
                that took product complexity and economies of scale into account.
                 Availability of technologies for production application
                too soon or too late. A number of technologies were evaluated that are
                not yet in production. CBI was gathered on the maturity and timing of
                these technologies and the cadence at which manufacturers could adopt
                these technologies.
                 Product complexity and design cadence constraints too low
                or too high. Product platforms, refresh and redesign cycles, shared
                engines, and shared transmissions were also considered in the analysis.
                Product complexity and the cadence of product launches were matched to
                historical values for each manufacturer.
                 Customer acceptance under estimated or over estimated.
                Resale prices for hybrid vehicles, electric vehicles, and internal
                combustion engine vehicles were evaluated to assess consumer
                willingness to pay for those technologies. The analysis accounts for
                the differential in the cost for those technologies and the amount
                consumers have actually paid for those technologies. Separately, new
                dual-clutch transmissions and manual transmissions were applied to
                vehicles already equipped with these transmission architectures.
                 The agencies sought comments on all assumptions for fuel economy
                technology costs, effectiveness, availability, and applicability to
                vehicles in the fleet.
                 Several commenters compared the technology effectiveness and cost
                estimates from prior rulemaking actions to the NPRM, some commenting
                that the NPRM analysis represented a better balance of input from all
                stakeholders regarding the potential costs and benefits of future fuel
                economy improving technologies,\668\ and some commenting that the NPRM
                analysis represented a step back from the Draft TAR and EPA's Proposed
                Determination in terms of both the analysis itself and the resulting
                conclusions about the level of technology required to meet the
                [[Page 24381]]
                augural standards.\669\ Specifically, while some commenters stated that
                the Draft TAR and subsequent EPA midterm review documents had recently
                concluded that augural standards were achievable with very low levels
                of electrification based on currently available information on
                technology effectiveness and cost,\670\ other commenters reiterated
                that conventional gasoline powertrains alone were insufficient to
                achieve post-2021 model year targets.\671\
                ---------------------------------------------------------------------------
                 \668\ See, e.g., NHTSA-2018-0067-11928.
                 \669\ See, e.g., NHTSA-2018-0067-11873.
                 \670\ See, e.g., NHTSA-2018-0067-11969.
                 \671\ See, e.g., NHTSA-2018-0067-12150.
                ---------------------------------------------------------------------------
                 Generally, the automotive industry supported the agencies' NPRM
                analysis over previous analyses. In addition to the automotive
                industry's support of the agencies' use of one modeling tool for
                analysis, discussed in Section IV, above, the industry also commented
                in support of specific technology effectiveness, cost, and adoption
                assumptions used in the updated analysis.
                 The Alliance commented in support of the NPRM modeling approach,
                and referenced important technology-specific features of the modeling
                process, including ``The acknowledgement and application of real-world
                limitations on technology application including a limit on the number
                of engine displacements available to any one manufacturer, application
                of shared platforms, engines, and transmissions, and the reality that
                improvements and redesigns of components are not only extended across
                vehicles but sometimes constrained in implementation opportunity to
                common vehicle redesign cycles; recognition of the need for
                manufacturers to follow ``technology'' pathways that retain capital and
                implementation expertise, such as specializing in one type of engine or
                transmission instead of following an unconstrained optimization that
                would cause manufacturers to leap to unrelated technologies and show
                overly optimistic costs and benefits; the application of specific
                instead of generic technology descriptions that allow for the above-
                mentioned real-world constraints; [and] the need to accommodate for
                intellectual property rights in that not all technologies will be
                available to all manufacturers.'' \672\
                ---------------------------------------------------------------------------
                 \672\ NHTSA-2018-0067-12073, at 9.
                ---------------------------------------------------------------------------
                 More specifically, the Alliance commented that the analysis
                appropriately restricted the application of some technologies, like the
                application of low rolling resistance tires on performance vehicles,
                and limited aerodynamic improvements for trucks and minivans.\673\
                Similarly, the Alliance commented in support of the decision to exclude
                HCR2 technology from the analysis, citing previous comments stating
                that ``the inexplicably high benefits ascribed to this theoretical
                combination of technologies has not been validated by physical
                testing.''
                ---------------------------------------------------------------------------
                 \673\ NHTSA-2018-0067-12073, at 134.
                ---------------------------------------------------------------------------
                 Ford commented more broadly that ``[t]he previous analyses
                performed by the Agencies too often selected technology benefits from
                the high-end of the forecasted range, and cost from the lower-end, in
                part because deference was given to supplier or other third-party
                claims over manufacturers' estimates.'' \674\ Ford noted that,
                ``[m]anufacturer estimates, while viewed as conservative by some, are
                informed by years of experience integrating new technologies into
                vehicle systems in a manner that avoids compromising other important
                attributes (NVH, utility, safety, etc.),'' continuing that ``[t]he need
                to preserve these attributes often limits the actualized benefit of a
                new technology, an effect insufficiently considered in projections from
                most non-OEM sources.'' Ford concluded, as mentioned above, that the
                NPRM analysis better balanced these considerations.
                ---------------------------------------------------------------------------
                 \674\ NHTSA-2018-0067-11928.
                ---------------------------------------------------------------------------
                 Toyota commented that the discrepancy between the automotive
                industry and prior regulatory assessments stemmed from ``agency
                modeling relying on overly optimistic assumptions about technology cost
                effectiveness and deployment rates.'' \675\ Toyota pointed to a prior
                analysis that projected compliance for Toyota's MY 2025 lineup using
                the ALPHA model as an example of how ``the agency's analysis failed to
                account for customer requirements (cost, power, weight-adding options,
                etc.) that erode optimal fuel economy, and normal business
                considerations that govern the pace of technology deployment.'' In
                contrast, Toyota stated that the ``[m]odeled technology cost,
                effectiveness, and compliance pathways in the proposed rulemaking rely
                on more recent data as well as more realistic assumptions about the
                level of technology already on the road today, the pace of technology
                deployment, and trade-offs between vehicle efficiency and customer
                requirements.''
                ---------------------------------------------------------------------------
                 \675\ NHTSA-2018-0067-12150.
                ---------------------------------------------------------------------------
                 Honda, in its feedback on the models used in the standard setting
                process, commented that ``the current version of the CAFE model is
                reasonably accurate in terms of technology efficiency, cost, and
                overall compliance considerations, and reflects a notable improvement
                over previous agency modeling efforts conducted over the past few
                years.'' \676\
                ---------------------------------------------------------------------------
                 \676\ NHTSA-2018-0067-11818.
                ---------------------------------------------------------------------------
                 FCA commented in recognition of the CAFE model improvements over
                the Draft TAR version, but noted they ``continue to believe that the
                cost and benefits used as inputs to the model are overly optimistic.''
                \677\ FCA used its updated Jeep Wrangler Unlimited and Ram 1500 pickup
                models as examples of vehicles that ``provide real life examples of the
                costs and benefits that can be achieved with fuel and weight saving
                technology;'' however, ``after all of the real world concerns such as
                emissions, drivability, OBD, and fuels are considered, the benefits
                observed remain less than those derived by the Autonomie model and used
                as inputs to the Volpe model.''
                ---------------------------------------------------------------------------
                 \677\ NHTSA-2018-0067-11943.
                ---------------------------------------------------------------------------
                 Conversely, environmental groups, consumer groups, and some States
                and localities commented that the Draft TAR and subsequent EPA analyses
                were more representative of the current state of vehicle technologies.
                These groups all generally commented, in different terms, that the NPRM
                analysis technology effectiveness was understated and technology costs
                were overstated, and additional constraints the agencies placed on the
                analysis, like excluding technologies already in production or
                constraining technology pathways, also helped lead to that result.\678\
                ---------------------------------------------------------------------------
                 \678\ NHTSA-2018-0067-11873; NHTSA-2018-0067-11984.
                ---------------------------------------------------------------------------
                 ICCT commented that the agencies ``ignored their own rigorous 2015-
                2017 technological assessment, and have adopted a series of invalid and
                unsupportable decisions which artificially constrain the availability
                and dramatically under-estimate levels of effectiveness of many
                different fuel economy improvement and GHG-reduction technologies and
                unreasonably increase modeled compliance costs.'' \679\ ICCT also
                commented that the agencies ignored, suppressed, dismissed, or
                restricted the use of work done to update technologies and technology
                cost and effectiveness assessments since the 2012 final rule for MYs
                2017-2025. ICCT stated that the ``invalid high cost result [of the
                modeled augural standards in 2025] was created by the agencies by
                making many dozens of unsupported changes in the technology
                effectiveness and availability inputs, the technology cost inputs, and
                the technology package constraints.''
                [[Page 24382]]
                ICCT stated that ``the agencies failed to capture the latest available
                information and, as a result, their assessment incorrectly and
                artificially overstates technology costs.''
                ---------------------------------------------------------------------------
                 \679\ NHTSA-2018-0067-11741 full comments.
                ---------------------------------------------------------------------------
                 CARB commented that the agencies did not present sufficient new
                evidence to change previous technical findings, specifically in regards
                to conventional vehicle technologies.\680\ CARB stated that instead of
                relying on new information, as had been asserted as justification for
                the proposal, the analysis was based on older data that did not reflect
                current technology. Accordingly, CARB pointed out that previous
                analysis by the agencies projected far less need for electrification
                than what was required in the proposal, stating that the underlying
                cause is a reduction in the assumed cumulative improvements for what
                advanced gasoline technology is able to achieve.
                ---------------------------------------------------------------------------
                 \680\ NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 A coalition of States and Cities similarly commented that ``[t]he
                Agencies' conclusions regarding the technology necessary to meet the
                2025 standards and the cost of that technology run counter to the
                evidence before the agency, diverge from prior factual findings without
                explanation and without transparency as to the source of data relied
                on, and are unsupported by any reasoned analysis. Such analysis bears
                many hallmarks of an arbitrary and capricious action.'' \681\
                ---------------------------------------------------------------------------
                 \681\ NHTSA-2018-0067-11735 (citing State Farm, 463 U.S. at 43;
                Fox Television, 556 U.S. at 515; Humane Soc. of U.S. v. Locke, 626
                F.3d 1040, 1049 (9th Cir. 2010)).
                ---------------------------------------------------------------------------
                 Roush Industries, commenting on behalf of CARB, commented that
                ``the 2018 PRIA projected average costs for technology implementation
                to achieve the existing standards to be significantly overstated and in
                conflict with the 2016 Draft TAR cost estimates generated by the
                Agencies only two years earlier.'' \682\ Roush commented that the Draft
                TAR analyses of cost and incremental fuel economy improvement necessary
                to achieve the augural standards was consistent with Roush's own
                estimates and other published data.
                ---------------------------------------------------------------------------
                 \682\ NHTSA-2018-0067-11984.
                ---------------------------------------------------------------------------
                 Similarly, H-D Systems (HDS), commenting on behalf of the
                California DOJ, commented that ``the estimates in the 2016 TAR on
                technology cost and effectiveness still represent the correct estimates
                based on the latest available data.'' \683\ HDS, in its analysis of the
                costs of technologies to meet different potential standards between the
                Draft TAR and the NPRM, noted that ``costs for most conventional (i.e.,
                non-electric) drivetrain technologies were similar in both reports in
                that costs were within +5% of the average of the costs from the two
                reports. The only exception was the cost estimate for the High CR
                second generation Atkinson cycle or HCR2 engine which was estimated to
                be much more expensive. Due to differences in nomenclature,
                transmission technology costs could not be directly compared but were
                similar at the highest efficiency level. In contrast, cost of hybrid
                technology was estimated to be much higher in the PRIA and were 200 to
                250% higher for strong hybrids. Costs of drag reduction, rolling
                resistance reduction and auxiliary system technologies were also quite
                similar but the cost of mass reduction was substantially higher in the
                PRIA by a factor of 2 to 3. Costs of engine friction reduction appear
                not to be included in the cost computation for the PRIA although the
                technology appears to be integrated into some of the engine technology
                packages analyzed in the PRIA to estimate effectiveness.''
                ---------------------------------------------------------------------------
                 \683\ NHTSA-2018-0067-11985.
                ---------------------------------------------------------------------------
                 CFA commented that ``[t]he overarching discussion of technology
                developments that introduces the NHTSA analysis is fundamentally flawed
                and infects the entire proposal,'' taking issue with the NPRM statement
                that ``some options considered in the original order for the National
                Program ha[d] not worked out as EPA/NHTSA anticipated.'' \684\ CFA
                commented that the agencies failed to note that some technology options
                have performed better than anticipated, and ``the fact that some
                technologies have done better than expected is a basis for increasing
                the standards, not in the context of a mid-term review that was
                supposed to tweak the long-term program.''
                ---------------------------------------------------------------------------
                 \684\ NHTSA-2018-0067-12005.
                ---------------------------------------------------------------------------
                 NCAT commented that the ``inflation of projected technology costs
                does not appear to be attributable primarily to the projected cost of
                any given technology, but rather to modeling constraints on the
                application of such technologies to vehicles. Many of these constraints
                appear to be arbitrary and NHTSA's departure from prior analyses in
                these respects is not adequately supported.'' \685\
                ---------------------------------------------------------------------------
                 \685\ NHTSA-2018-0067-11969.
                ---------------------------------------------------------------------------
                 Environmental groups and States also commented that the agencies
                either should reincorporate all the Draft TAR or the EPA Proposed and
                Final Determination analyses' technologies, technology effectiveness
                values, and technology costs into the analysis, and/or compare the
                final rule analysis with those prior analyses to show how the updated
                assumptions changed the results from those prior analyses.
                 For example, ICCT commented that ``[f]or the agencies to conduct a
                credible regulatory assessment they must remove all the technology
                availability constraints, re-incorporate and make available the full
                portfolio of technology options as was available in EPA's analysis for
                the original 2017 Final Determination, and include at least 15 g/mile
                CO2 for off-cycle credits by 2025, to credibly reflect the
                real-world technology developments in the auto industry.'' \686\ ICCT
                also stated that ``[t]he agencies need to identify each and every
                technology cost input used in their modeling, and provide a clear
                engineering and evidence based justification for why that cost differs
                from the costs employed in the extremely well documented and well
                justified Draft TAR and in EPA's 2016 TSD and 2017 Final Determination,
                taking into account the above discussion of significant new evidence
                developed since those prior estimates were made. Absent such disclosure
                and justification, the default assumption needs to be that the prior
                costs estimated based on the most recent data are more appropriate than
                the estimates used for the proposal.''
                ---------------------------------------------------------------------------
                 \686\ NHTSA-2018-0067-11741 full comments.
                ---------------------------------------------------------------------------
                 In addition, groups of commenters were equally split on the ability
                of technologies to meet different compliance targets. For example, the
                Alliance commented that ``the only technologies that have demonstrated
                the improvements necessary to meet the MY 2025 standards are strong
                hybrids, plug-in electric vehicles, and fuel cell electric vehicles.
                The Agencies' analysis for this Proposed Rule predict the need for
                significant growth in sales of electrified vehicles, a finding
                consistent with third-party analyses.'' \687\ In contrast, UCS
                commented that electrified powertrains ``are not especially relevant
                for the MY 2022-2025 regulations.'' \688\
                ---------------------------------------------------------------------------
                 \687\ NHTSA-2018-0067-Alliance at 15.
                 \688\ NHTSA-2018-0067-UCS at 23.
                ---------------------------------------------------------------------------
                 The agencies are aware that the prior analyses concluded that
                compliance with the augural standards could largely be met through
                advances in gasoline vehicle technologies, and with only very low
                levels of strong hybrids and electric vehicles. As the agencies stated
                in the NPRM, consistent with both agencies' statutes, the proposal was
                entirely de novo, based on an entirely new analysis reflecting the best
                and most up-to-date information available to the agencies at the time
                of this rulemaking.\689\ As discussed in Section IV, Section VI.B, and
                further below, the NPRM and final rule analyses reflect updates to
                [[Page 24383]]
                technology effectiveness estimates, technology costs, and the
                methodology for applying technologies to vehicles that the agencies
                believed better represent the state of technology and the associated
                costs compared to prior analyses, that result in pathways to compliance
                that look both similar and different to those in prior analyses.
                ---------------------------------------------------------------------------
                 \689\ 83 FR 42897.
                ---------------------------------------------------------------------------
                 That said, several of the effectiveness and cost values used in the
                NPRM and final rule analysis were directly carried over from the 2012
                rule for MYs 2017-2025, Draft TAR, and EPA Midterm Evaluation
                analyses.\690\ Several others were carried over from the 2015 NAS
                report,\691\ which the agencies heavily relied upon in past analyses
                even if specific cost or effectiveness values were not used. Different
                technology effectiveness estimates, cost estimates, or adoption
                constraints were employed where the agencies had information, from
                technical reports, manufacturers, or other stakeholders, indicating
                that a technology could or could not be feasibly adopted in the
                rulemaking timeframe, or a technology could or could not be adopted in
                the way that the agencies had previously modeled it. Notably, most
                differences in pathways to compliance are attributable to only a few
                significant differences between this rulemaking analysis and prior
                rulemaking analyses.
                ---------------------------------------------------------------------------
                 \690\ See, e.g., PRIA at 449, 451, 452, 453, 458.
                 \691\ See, e.g., PRIA at 358-360.
                ---------------------------------------------------------------------------
                 For example, as discussed in Section VI.B.3 Technology
                Effectiveness and Modeling and Section VI.C.1 Engine Paths, in the EPA
                Draft TAR and Proposed Determination analyses, effectiveness of HCR
                engine technologies and downsized turbocharged engine technologies were
                estimated using Tier 2 certification fuel. Tier 2 certified fuel has a
                higher octane rating compared to regular octane
                fuel.692 693 694 As summarized by EPA in the PD TSD, ``EPA's
                estimate of effectiveness for gasoline-fueled engines and engine
                technologies was based on Tier 2 Indolene fuel although protection for
                operation in-use on Tier 3 gasoline (87 AKI E10) was included in the
                analysis of engine technologies considered both within the Draft TAR
                and Proposed Determination. Additionally, in the technology assessment
                for this Proposed Determination, EPA has considered the required engine
                sizing and associated effectiveness adjustments when performance
                neutrality is maintained on 87AKI gasoline typical of real-world use.''
                \695\
                ---------------------------------------------------------------------------
                 \692\ Draft TAR at 5-228.
                 \693\ Tier 2 fuel has an octane rating of 93. Typical regular
                grade fuel has an octane rating of 87 ((R+M)/2 octane.
                 \694\ EPA Proposed Determination TSD at 2-209 to 2-212.
                 \695\ EPA Proposed Determination TSD at 2-210.
                ---------------------------------------------------------------------------
                 NHTSA's effectiveness analysis for the Draft TAR used some engine
                maps also developed using premium octane gasoline. However, at the time
                NHTSA stated the agency would ensure all future engine model
                development will be performed with regular grade octane gasoline.\696\
                Commenters like Ford stated the effectiveness estimates for turbo
                downsized engine packages were too high, in part because of the use of
                high octane fuel. However they also commented in appreciation of
                NHTSA's acknowledgement that any subsequent analysis would be based on
                fuel at an appropriate octane level, as they stated the impact of the
                change needed to be reflected in future analyses.\697\
                ---------------------------------------------------------------------------
                 \696\ Draft TAR at 5-504, 5-512.
                 \697\ Ford Motor Company Response to the Draft TAR September 26,
                2016 NHTSA-2016-0068-0048, at 4.
                ---------------------------------------------------------------------------
                 Engine specifications used to create the engine maps for the NPRM
                and the final rule analysis were developed using Tier 3 fuel to assure
                the engines were capable of operating on real world regular octane (87
                pump octane = (R+M/2)). The process was similar to what manufacturers
                must do to ensure engines have acceptable noise, vibration, harshness,
                drivability, performance, and will not fail prematurely when operated
                on regular octane fuel. This eliminated the need for any adjustments
                that were applied in the 2016 Draft TAR and PD TSD to account for Tier
                2 to Tier 3 fuel properties. This accounts for some of the
                effectiveness and cost differences for engine technologies between the
                Draft TAR/Proposed Determination and the NPRM/final rule. For more
                details, see Section VI.C.1 Engine Paths.
                 The agencies believe ICCT's and other commenters' assertions that
                the engine maps should reflect Tier 2 fuel and not be updated for Tier
                3 fuel would ignore these important considerations, and would provide
                engine maps that could not achieve the fuel economy improvements unless
                operated on high octane fuel. Therefore, the agencies determined that
                engine maps developed for the Draft TAR and EPA Proposed Determination
                that were based on Tier 2 fuel should not be used for the NPRM and
                final rule analyses for these technical reasons.
                 As another related example, the agencies described that prior
                analyses had relied heavily on the availability of the HCR2 (or ATK2)
                ``future'' Atkinson Cycle engine as a cost-effective pathway to
                compliance for stringent alternatives, but many engine experts
                questioned its technical feasibility and near-term commercial
                practicability.\698\ The agencies explained that EPA staff began
                theoretical development of this conceptual engine with a best-in-class
                2.0L Atkinson cycle engine and then increased the efficiency of the
                engine map further, through the theoretical application of additional
                technologies in combination, including cylinder deactivation, engine
                friction reduction, and cooled exhaust gas recirculation. While the
                potential of such an engine is interesting, nevertheless the engine
                remains entirely speculative. No production HCR2/ATK2 engine, as
                outlined in the EPA SAE paper,\699\ has ever been commercially
                produced. Furthermore, the engine map has not been validated with
                hardware, bench data, or even on a prototype level (as no such engine
                exists to test to validate the engine map).
                ---------------------------------------------------------------------------
                 \698\ 83 FR 43038.
                 \699\ Schenk, C. and Dekraker, P., ``Potential Fuel Economy
                Improvements from the Implementation of cEGR and CDA on an Atkinson
                Cycle Engine,'' SAE Technical Paper 2017-01-1016, 2017. Available at
                https://doi.org/10.4271/2017-01-1016.
                ---------------------------------------------------------------------------
                 Vehicle manufacturers also commented on EPA's effectiveness
                assumptions and estimates of HCR2/ATK2 model's future penetration
                levels in the Draft TAR, stating ``[t]he effectiveness values for the
                `futured' ATK2 package--projected at 40% penetration in 2025MY and
                includes cooled exhaust gas recirculation (CEGR) and cylinder
                deactivation (DEAC)--are too high, primarily due to overtly-optimistic
                efficiencies in the base engine map, insufficient accounting of CEGR
                and DEAC integration losses, and no accounting of the impact of 91RON
                Tier 3 test fuel,'' and that ``44% fleet-wide penetration of ATK2 in
                2025MY is unrealistic given the limited number of powertrain refresh
                cycles available before 2025MY. In addition, it is unreasonable to
                assume that OEMs already heavily invested in different high-efficiency
                powertrain pathways (e.g., turbo-downsizing) would be able to commit
                the immense resources needed to reach these high ATK2 penetration
                levels in such a short time.'' \700\
                ---------------------------------------------------------------------------
                 \700\ Ford Motor Company Response to the Draft TAR September 26,
                2016 NHTSA-2016-0068-0048, at 4.
                ---------------------------------------------------------------------------
                 Accordingly, the agencies decided to not include HCR2 technology in
                the NPRM and final rule analysis. The engine model was not used because
                no observable physical demonstration of the speculative technology
                combination model has yet been created. Further,
                [[Page 24384]]
                many questions remain about the model's practicability as specified,
                especially in high load, low engine speed operating conditions. The
                HCR2 model combines multiple technologies to provide cumulative
                estimate of benefits without consideration the practical interaction of
                technologies. This approach runs contrary to the modeling approach
                attempted in the NPRM and final rule analysis. The approach the
                agencies tried to follow restricted models to adding discrete advanced
                technologies. This approach allowed an accounting of synergetic
                effects, identified incremental benefits, and increased the precision
                of cost estimates.
                 As another example, further discussed in Section VI.B.1 Analysis
                Fleet, the agencies had traditionally taken different approaches to
                assigning baseline road load reduction technology assignments. For
                analyzing baseline levels of mass reduction in an analysis fleet, NHTSA
                had developed for the Draft TAR a regression model to summarize a
                vehicle's weight savings using a relative performance approach and
                accounting for vehicle content, using cost curves developed from
                teardown studies of a MY 2011 Honda Accord and MY 2014 Chevrolet
                Silverado pickup truck. EPA developed its own methodology that
                classified vehicles based on weight reductions from a MY 2008 vehicle,
                compared to the MY 2014 version of the same vehicle, using a cost curve
                from a tear-down study of a MY 2010 Toyota Venza. In the EPA's mass
                reduction technology costing approach, a cost reduction was applied
                when mass reduction 1 technology was applied to a system at mass
                reduction 0 technology level. NHTSA's approach, used in the NPRM and
                final rule analysis, set baseline mass reduction assignments so costs
                of implementing mass reduction technologies are fully applied as
                vehicle platforms move along the mass reduction technology path.
                 The agencies also included additional advanced powertrain
                technologies and other vehicle-level technologies in the technology
                pathways between the Draft TAR and NPRM, and between the NPRM and final
                rule. However, manufacturers and suppliers have repeatedly told the
                agencies that there are diminishing returns to increasing the
                complexity of advanced gasoline engines, including in the amount of
                fuel efficiency benefit that they can provide. For example, Toyota
                commented, in response to the EPA SAE paper benchmarking the 2018 Camry
                with the 2.5L Atkinson-cycle engine and ``futuring'' midsize exemplar
                vehicles based on the generated engine map,\701\ that although EPA's
                addition of cylinder deactivation to the hypothetical 2025 exemplar
                vehicle is technically possible and would provide some fuel economy and
                CO2 benefit, the primary function of cylinder deactivation
                is to reduce engine pumping losses which the Atkinson cycle and EGR
                already accomplish on the 2018 Camry.\702\ Toyota concluded, ``The
                overlapping and redundant measures to reduce engine pumping losses
                would add costs with diminishing efficiency returns.'' Similarly,
                BorgWarner commented that they ``do not expect that variable
                compression ratio (VCR) or homogeneous charge compression ignition
                (HCCI) will see broad application in the short term, if ever. While
                each of these technologies can offer marginal efficiency gains at some
                engine speed-load conditions, the use of down-sized boosted engines
                with 8-10 speed transmissions makes it possible to run engines at near
                optimum conditions and effectively minimizes gains from VCR or HCCI.
                VCR mechanisms result in additional mass, cost and complexity, and true
                HCCI has yet to be demonstrated in a production vehicle. The agencies
                do not believe that OEMs will judge these technologies to be cost
                effective.'' \703\
                ---------------------------------------------------------------------------
                 \701\ Kargul, J., Stuhldreher, M., Barba, D., Schenk, C. et al.,
                ``Benchmarking a 2018 Toyota Camry 2.5-Liter Atkinson Cycle Engine
                with Cooled-EGR,'' SAE Technical Paper 2019-01-0249, 2019,
                doi:10.4271/2019-01-0249.
                 \702\ NHTSA-2018-0067-12431, at 8.
                 \703\ NHTSA-2018-0067-11895.
                ---------------------------------------------------------------------------
                 So, while previous analyses may have shown pathways to compliance
                with increasingly complex advanced gasoline engines, the NPRM and final
                rule analyses more appropriately reflect that the most complex gasoline
                engine technologies will account for a smaller share of manufacturers'
                products during the rulemaking timeframe. However, despite this fact,
                the NPRM and final rule analysis include more advanced powertrain
                technologies than previous analyses, in part to account for important
                considerations like intellectual property and the fact that some
                manufacturers have already started down the path of incorporating a
                certain advanced engine technology in their product portfolio, and that
                abrupt switching to another advanced engine technology would result in
                unrealistic stranding of capital costs. In addition, greater precision
                in how cumulative technologies applied to engines, as estimated through
                the Autonomie effectiveness modeling, appropriately reflects the
                diminishing returns to efficiency benefits that those advanced engines
                can provide. Moreover, as identified by a wide range of commenters,
                battery costs are projected to fall in the rulemaking timeframe to a
                point where, in the compliance modeling, it becomes more cost effective
                to add electrification technologies to vehicles than to apply other
                advanced gasoline engine technologies.
                 Finally, the agencies declined to incorporate some information and
                data for the NPRM or final rule central analysis for reasons discussed
                in the following sections. In general, the data produced by agencies or
                submitted by commenters failed to isolate effectiveness impacts of
                individual technologies (or in some cases a combination of two or
                several technologies). The data included effects from additional
                unaccounted and undocumented technologies. Because the effectiveness
                improvement measured or claimed resulted from more than just the
                reported sources, the actual effectiveness of the technology or
                technologies is obfuscated and easily under or over predicted. Using
                effectiveness values generated in this manner carries a high risk of
                double counting effectiveness and undercounting costs.
                 In many cases, this problem exists where data or information is
                based on laboratory testing or on-road testing of production vehicles
                or components including engines and transmissions. Production vehicles
                and components usually include multiple technology improvements from
                one redesign to the next, and rarely incorporate just a single
                technology change. Furthermore, technology improvements on production
                vehicles in some cases cannot be readily observed, such as the level of
                mechanical friction in an engine, and isolation and identification of
                the improvement attributable to each technology would be impractical
                given the costs and time required to do so. That said, in some cases,
                where possible to do so, the agencies used the data or information from
                production vehicles to corroborate information from the Autonomie
                simulations. However, the agencies declined to apply that data or
                information directly in the analysis if the effectiveness improvement
                attributable to a particular technology could not be isolated.
                 The agencies made these updates from prior analyses not, as some
                commenters have suggested, to ``artificially overstate technology
                costs,'' \704\ or to ``ignore the knowledge and expertise of the EPA
                engineering
                [[Page 24385]]
                and compliance staff,'' \705\ ``so that the model in many instances
                selects more expensive, less fuel efficient technology while excluding
                less expensive and more efficient alternatives,'' \706\ but because the
                updates reflected the agencies' reasonable assessment of the current
                state of vehicle technologies and their costs, and the state of future
                vehicle technologies and costs in the rulemaking timeframe.
                ---------------------------------------------------------------------------
                 \704\ NHTSA-2018-0067-11741 at 7.
                 \705\ NHTSA-2018-0067-11741 at I-23.
                 \706\ NHTSA-2018-0067-12123.
                ---------------------------------------------------------------------------
                 Separate from the decision to update assumptions used for the NPRM
                analysis from prior analyses, the agencies did refine some technology
                effectiveness and cost assumptions from the NPRM to this final rule
                analysis. In addition to being appropriate for technical reasons, this
                should address some commenters' overarching concerns about understated
                technology effectiveness and overstated technology costs. For example,
                several commenters noted that the costs of BISG/CISG systems were
                higher for small Cars/SUVs and medium cars than for medium SUVs and
                pickup trucks, which the Alliance and FCA described as ``implausible''
                and ``misaligned with industry understanding,'' and which ICCT
                described as ``contrary to basic engineering logic, which holds that a
                system which would be smaller and have lower energy and power
                requirements would be less expensive, not more.'' \707\ The agencies
                agree, and have made changes to address this issue, as described in
                Section VI.C.3.a) Electrification.
                ---------------------------------------------------------------------------
                 \707\ NHTSA-2018-0067-11741.
                ---------------------------------------------------------------------------
                 After considering comments, the agencies also added several engine
                technologies and technology combinations for the final rule analysis.
                These included a basic high compression ratio Atkinson cycle engine, a
                variable compression ratio engine, a variable turbo geometry engine,
                and a variable turbo geometry with electric assist engine (VTGe). The
                NPRM discussed and provided engine maps for each of these technologies.
                The agencies also added new technology combinations including diesel
                engines with cylinder deactivation, turbocharged engines with advanced
                cylinder deactivation, diesel engines paired with manual transmissions,
                and diesel engines paired with 12-volt start-stop technology.
                Transmission revisions included updating the effectiveness of 6-speed
                automatic transmissions, applying updated shift logic for 10-speed
                automatic transmissions, and increasing the gear span for efficient 10-
                speed automatic transmissions. Mass reduction technology was expanded
                to include up to 20 percent curb weight reduction, compared with up to
                10 percent for the NPRM. These changes, and the comments upon which
                they were based, are described in further detail in the following
                sections.
                1. Engine Paths
                 The internal combustion (IC) engine is a heat engine that converts
                chemical energy in a fuel into mechanical energy. Chemical energy of
                the fuel is first converted to thermal energy by means of combustion or
                oxidation with air inside the engine. This thermal energy raises the
                temperature and pressure of the gases within the engine, and the high-
                pressure gas then expands against the internal mechanisms of the
                engine. This expansion is converted by the mechanical linkages of the
                engine to a rotating crankshaft, which is the output of the engine. The
                crankshaft, in turn, is connected to a transmission to transmit the
                rotating mechanical energy to the desired final use, particularly the
                propulsion of vehicles.
                 IC engines can be categorized in a number of different ways
                depending upon which technologies are designed into the engine: By type
                of ignition (e.g., spark ignition or compression ignition), by engine
                cycle (e.g., Otto cycle or Atkinson cycle), by valve actuation (e.g.,
                overhead valve (OHV), single overhead camshaft (SOHC), or dual overhead
                camshaft (DOHC)), by basic design (e.g., reciprocating or rotary), by
                configuration and number of cylinders (e.g., inline four-cylinder (I4)
                or V-shaped six-cylinder (V6)), by air intake (e.g., forced induction
                (turbo or super charging) or naturally aspirated), by method of fuel
                delivery (e.g., port injection or direction injection), by fuel type
                (e.g., gasoline or diesel), by application (e.g., passenger car or
                light truck),or by type of cooling (e.g., air-cooled or water-cooled).
                For each combination of technologies among the various categories,
                there is a theoretical maximum efficiency for all engines within that
                set. There are various metrics that can be used to compare engine
                efficiency, and the four metrics the agencies use or discuss in this
                preamble are:
                 Brake specific fuel consumption (BSFC), which is the mass
                of fuel consumed per unit of work output (amount of fuel used to
                produce power);
                 Brake thermal efficiency (BTE), which is the total fuel
                energy released per unit of work output (percentage of fuel used to
                produce power);
                 Fuel consumption (gallons per mile), which looks at the
                gallons of fuel consumed per unit of work output (mile travelled); and
                 Fuel economy (in MPG), which is the amount of work output
                (miles travelled) per unit (gallon) of fuel consumed.
                 When comparing the efficiency of IC engines, it is important to
                identify the metric(s) used and the test cycle for the measurement
                because results vary widely when engines operate over different test
                cycles. Two-cycle fuel economy tests used to certify vehicles'
                compliance with the CAFE standards tend to overestimate the average
                fuel economy motorists will typically achieve during on-road
                operation.\708\ In the NPRM and for this final rule analysis, the
                agencies considered technology effectiveness for the 2-cycle test
                procedures and AC and off-cycle test procedures to evaluate how
                technologies could be applied for manufacturers to comply with
                standards. The agencies also considered real world operation beyond
                these test procedures when considering IC engine technologies in order
                to assure the technologies were configured and specified in a manner
                that could be used in real world vehicle applications.
                ---------------------------------------------------------------------------
                 \708\ 77 FR 62988.
                ---------------------------------------------------------------------------
                a) Fuel Octane
                 As mentioned in other sections of the Preamble, the agencies go to
                great lengths to ensure engine technologies considered for potential
                compliance pathways are feasible for real-world implementation and
                effectiveness. An important facet of this evaluation are both the fuels
                that are used for efficiency testing and also the fuels that consumers
                may purchase in the marketplace.
                 In the NPRM, the agencies included a general overview of fuel
                octane (stability) level, including levels currently available, and the
                potential impact of fuel octane on engines developed for the U.S.
                market.\709\ The agencies described that a typical, overarching goal of
                optimal spark-ignited engine design and operation is to maximize the
                greatest amount of energy from the fuel available, without manifesting
                detrimental impacts to the engine over expected operating conditions.
                Design factors, such as compression ratio, intake and exhaust value
                control specifications, and combustion chamber and piston
                characteristics, among others, are all impacted by the octane of the
                fuel consumers are anticipated to use.\710\
                ---------------------------------------------------------------------------
                 \709\ PRIA at 253.
                 \710\ In addition, PRIA Chapter 6 contains a brief discussion of
                fuel properties, octane levels used for engine simulation and in
                real-world testing, and how octane levels can impact performance
                under these test conditions.
                ---------------------------------------------------------------------------
                [[Page 24386]]
                 The agencies also discussed potential challenges associated with
                octane levels available currently, and how those octane levels may play
                a role in potential vehicle fuel efficiency improvements. Vehicle
                manufacturers typically develop their engines and engine control system
                calibrations based on the fuel available to consumers. In many cases,
                manufacturers may recommend a fuel grade for best performance and to
                prevent potential damage. In some cases, manufacturers may require a
                specific fuel grade for both best performance, to achieve advertised
                power ratings, and/or to prevent potential engine damage.
                 Consumers, though, may or may not choose to follow the
                manufacturer's recommendation or requirement for a specific fuel grade
                for their vehicle. As such, vehicle manufacturers often choose to
                employ engine control strategies for scenarios where the consumer uses
                a lower than recommended, or required, fuel octane level, as a way to
                mitigate potential engine damage over the life of a vehicle. These
                strategies limit the extent to which some efficiency improving engine
                technologies can be implemented, such as increased compression ratio
                and intake system and combustion chamber designs that increase burn
                rates and rate of in-cylinder pressure rise. If the minimum octane
                level available in the market were higher (especially the current sub-
                octane regular grade in the mountain states), vehicle manufacturers
                might not feel compelled to design vehicles sub-optimally to
                accommodate such blends.
                 When knock (also referred to as detonation) is encountered during
                engine operation, at the most basic level, non-turbocharged engines can
                adjust the timing of the spark that ignites the fuel, as well as the
                amounts of fuel injected at each intake stroke (``fueling''). In
                turbocharged applications, knocking is typically controlled by
                adjusting boost levels along with spark timing and/or the amount of
                fuel injected. Past rulemakings discussed other techniques that may be
                employed to allow higher compression ratios, including optimizing spark
                timing, and adding of cooled exhaust gas recirculation (EGR).
                Regardless of the type of spark-ignition engine or technology employed,
                efforts to reduce or prevent knock with the lower-octane fuels that are
                available in the market result in the loss of potential power output,
                creating a ``knock-limited'' constraint on performance and efficiency.
                 The agencies noted that despite limits imposed by available fuel
                grades, manufacturers continue to make progress in extracting more
                power and efficiency from spark-ignited engines. Production engines are
                safely operating with regular 87 AKI fuel with compression ratios and
                boost levels once viewed as only possible with premium fuel. According
                to the Department of Energy, the average gasoline octane level has
                remained fundamentally flat starting in the early 1980's and decreased
                slightly starting in the early 2000s. During this time, however, the
                average compression ratio for the U.S. fleet has increased from 8.4 to
                10.52, a more than 20 percent increase. As explained by the Department
                of Energy, ``[t]here is some concern that in the future, auto
                manufacturers will reach the limit of technological increases in
                compression ratios without further increases in the octane of the
                fuel.'' \711\ As such, manufacturers are still limited by the fuel
                grades available to consumers and the need to safeguard the durability
                of their products for all of the available fuels; thus, the potential
                improvement in the design of spark-ignition engines continues to be
                overshadowed by the fuel grades available to consumers.
                ---------------------------------------------------------------------------
                 \711\ Fact of the Week, Fact #940: August 29, 2016 Diverging
                Trends of Engine Compression Ratio and Gasoline Octane Rating, U.S.
                Department of Energy, https://www.energy.gov/eere/vehicles/fact-940-august-29-2016-diverging-trends-engine-compression-ratio-and-gasoline-octane (last visited Mar. 21, 2018).
                ---------------------------------------------------------------------------
                 EPA and NHTSA also described ongoing research and positions from
                automakers and advocacy groups on fuel octane levels, including
                comments received during past agency rulemakings and on the 2016 Draft
                TAR regarding the potential for increasing octane levels in the U.S.
                market. The agencies described arguments for adjusting to octane
                levels, including making today's premium grade the base grade of fuel
                available, which could enable low cost design changes to improve fuel
                economy and reduce tailpipe CO2 emissions. Challenges
                associated with this approach include the increased cost to consumers
                who drive vehicles designed for current regular octane grade fuel, who
                would not benefit from the use of the higher cost higher-octane fuel.
                The costs of such a transition to higher-octane fuel would be high and
                persist well into the future, since unless current regular octane fuel
                were unavailable in the North American market, manufacturers would be
                effectively unable to redesign their engines to operate on higher-
                octane fuel. In addition, the full benefits of such a transition would
                not be realized until vehicles with such redesigned engines were
                produced for a sufficient number of model years largely to replace the
                current on-road vehicle fleet. The transition to net positive benefits
                would take many years.
                 The agencies also described input received from renewable fuel
                industry stakeholders and from the automotive industry supporting high-
                octane gasoline fuel blends to enable fuel economy and CO2
                improving technologies such as higher compression ratio engines.
                Stakeholders suggested that mid-level (e.g., E30) high-octane ethanol
                blends should be considered and that EPA should consider requiring that
                mid-level blends be made available at service stations. Stakeholders
                supporting higher-octane blends suggested that higher-octane gasoline
                could provide auto manufacturers with more flexibility to meet more
                stringent standards by enabling opportunities for use of lower tailpipe
                CO2 emitting technologies (e.g., higher compression ratio
                engines, improved turbocharging, optimized engine combustion).
                 The agencies sought additional comment in the NPRM on various
                aspects of current fuel octane levels and how fuel octane could play a
                role in the future. More specifically, the agencies sought comment on
                how increasing fuel octane levels could have an impact on product
                offerings and engine technologies, as well as what improvements to fuel
                economy and tailpipe CO2 emissions could result from higher-
                octane fuels. The agencies sought comment on an ideal octane level for
                mass-market consumption, and whether there were downsides with
                increasing the available octane levels and, potentially, eliminating
                lower-octane fuel blends. EPA also requested comment on whether and how
                EPA could require the production and use of higher-octane gasoline
                consistent with Title II of the Clean Air Act.
                 The agencies received numerous, wide-ranging comments in response
                to the NPRM discussion, and some direct responses to the agencies'
                requests for comments. The commenters included fuel producers,
                individual vehicle manufactures, environmental groups, vehicle
                suppliers, fuel advocacy groups, and agricultural organizations, among
                others. Commenters provided a broad range of comments ranging from
                explication of the many challenges to increasing available octane
                levels, to claims of the substantial efficiency
                [[Page 24387]]
                increases that could be easily obtained by requiring higher-octane
                levels.
                 Several ethanol industry stakeholders commented in support of
                requiring higher-octane fuels using mid-level ethanol blends. The High-
                Octane, Low Carbon (HOLC) Alliance commented that it believes ``NHTSA
                and EPA have a critical opportunity to cost-effectively ensure progress
                in fuel efficiency and CO2 emissions standards. Scientific
                experts agree that high-octane, low-carbon fuel can yield greater fuel
                economy and emissions benefits when paired with internal combustion
                engines (ICEs). But, to realize such benefits, automobile manufacturers
                require approval sooner rather than later to such fuels. Alternatively,
                automobile manufacturers will be limited in their ability to maximize
                the environmental performance of their vehicles until non-liquid fuel
                engines become more readily available. In finalizing the Proposed Rule,
                the HOLC Alliance strongly urges EPA and NHTSA to establish a pathway
                forward toward incentivizing the production and adoption of higher-
                octane, lower carbon fuels. By doing so, EPA and NHTSA can continue to
                incrementally increase CO2 and fuel economy standards,
                respectively.'' \712\
                ---------------------------------------------------------------------------
                 \712\ HOLC Alliance, Detailed Comments, EPA-HQ-OAR-2018-0283-
                4196.
                ---------------------------------------------------------------------------
                 Renewable Fuels Associations (RFA) commented that ``it strongly
                believes vehicles and fuels must be considered together as integrated
                systems. As EPA has recognized in the past, a `systems approach enables
                emission reductions that are both technologically feasible and cost
                effective beyond what would be possible looking at vehicle and fuel
                standards in isolation.' Because ethanol-based high-octane low-carbon
                fuel blends would enable cost-effective gains in fuel economy and
                carbon dioxide reductions, the agencies should take steps to support
                [high-octane low-carbon] fuels in the final SAFE rule.'' \713\
                ---------------------------------------------------------------------------
                 \713\ RFA, Detailed Comments, EPA-HQ-OAR-2018-0283-4409.
                ---------------------------------------------------------------------------
                 RFA cited several studies indicating benefits are available from
                raising the floor of fuel octane levels currently available, and,
                particularly, ``[t]he results from the studies reviewed generally
                support a main conclusion that splash blending ethanol is a highly
                effective means of raising the octane rating of gasoline and enabling
                low-cost efficiencies and reduced emissions in modern spark-ignition
                engines.'' \714\ In addition, National Corn Growers Association stated
                that, ``[w]ithout a change in fuel, automakers are reaching the limits
                on the efficiency gains that can be achieved with technology changes.''
                \715\
                ---------------------------------------------------------------------------
                 \714\ RFA, Detailed Comments, EPA-HQ-OAR-2018-0283-4409.
                 \715\ National Corn Growers Association, https://www.ncga.com/file/1621/NCGA%20Comments20Docket%20No.%20EPA-HQ-OAR-2018-0283%20and%20NHTSA-2018-0067.pdf.
                ---------------------------------------------------------------------------
                 The National Corn Growers Association, in conjunction with
                associated corn growing and agricultural groups, pointedly stated the
                EPA should, ``[s]et a minimum fuel octane level of 98 RON and phase out
                low octane fuels as new optimized vehicles enter the market in MY
                2023,'' and concluded that approving a ``midlevel ethanol blend vehicle
                certification fuel would enable automakers to expedite design and
                testing of optimized vehicles for use with this new fuel.'' \716\
                ---------------------------------------------------------------------------
                 \716\ National Corn Growers Association, https://www.ncga.com/file/1621/NCGA%20Comments%20Docket%20No.%20EPA-HQ-OAR-2018-0283%20and%20NHTSA-2018-0067.pdf.
                ---------------------------------------------------------------------------
                 The 25x25 Alliance commented that ``to meet the dual goals of
                greater fuel efficiency and reduced GHG emissions, the utilization of
                higher compression spark ignition internal combustion engines will be
                essential. Increasing engine compression improves thermal efficiency.
                However, as compression increases, higher-octane fuels will be needed
                to prevent engine knock. Automakers and advocacy groups have expressed
                support for increases to fuel octane levels for the US market. Ethanol
                with its octane rating of 113 offers engine knock resistance at a lower
                cost than any other octane booster in gasoline. In addition, ethanol's
                lower direct and life-cycle GHG emissions as compared to gasoline are
                well documented. For this reason, a fuel produced from a mixture of
                ethanol and gasoline and used in conjunction with advanced high
                compression engines presents itself as a technology pathway capable of
                complying with new CAFE/GHG standards.'' They continue, ``HOLC
                supporters recognize numerous barriers and other associated regulatory
                hurdles must be resolved before HOLC ethanol fuels are adopted at large
                scale. . . 25x25 believes it is imperative that the vehicle and fuel be
                treated as a comprehensive system. To date CAFE/GHG standards have
                largely focused on vehicle engine technology. Advanced engine vehicles
                perform best in concert with fuels of suitable properties and
                composition to optimally enable and power them.'' \717\
                ---------------------------------------------------------------------------
                 \717\ 25x25 Alliance, Detailed Comments, EPA-HQ-OAR-2018-0283-
                4210.
                ---------------------------------------------------------------------------
                 The American Coalition for Ethanol (ACE) commented that ``high-
                octane blends comprised of 25 to 30 percent ethanol would help bring
                down the cost for consumers compared to the premium-priced octane level
                advocated by oil refiners. Ethanol has a blending octane rating of
                nearly 113 and trades at a steep discount to gasoline. In many
                wholesale markets today, ethanol costs at least 60 cents per gallon
                less than gasoline. Ethanol delivers the highest octane at the lowest
                cost, allowing automakers to benefit by continuing to develop high-
                compression engine technologies and other product offerings to achieve
                efficiency improvements and reduced emissions. The ideal way to
                transition from today's legacy fleet to new vehicles with advanced
                engine technologies designed to run optimally on a high-octane fuel is
                to utilize FFVs as bridge vehicles that can provide immediate demand
                for mid-level ethanol blends.'' \718\
                ---------------------------------------------------------------------------
                 \718\ ACE, Detailed Comments, EPA-HQ-OAR-2018-0283-4033.
                ---------------------------------------------------------------------------
                 Growth Energy commented that with a mid-level ethanol blend,
                automakers not only get higher-octane that they can use to optimize
                engines and gain further fuel efficiency, they will also see a fuel
                that has demonstrably lower carbon dioxide emissions.\719\ The Illinois
                Corn Growers' Association et al., commented that ``NHTSA and EPA must
                adapt the existing regulatory structure to reflect the specific
                characteristics of mid-level blend fuels. Working together, the ethanol
                industry, automakers, EPA and NHTSA can bring about, during the period
                covered by the SAFE program, a new generation of high efficiency
                internal combustion engines optimized to take advantage of this new
                fuel's unique properties.'' \720\
                ---------------------------------------------------------------------------
                 \719\ Growth Energy, Detailed Comments, EPA-HQ-OAR-2010-0799-
                9540-A2.
                 \720\ Comment removed because it contains copyrighted data,
                Illinois Corn Growers Association, et al., https://www.regulations.gov/document?D=EPA-HQ-OAR-2018-0283-4198.
                ---------------------------------------------------------------------------
                 Ethanol industry commenters provided comment on several EPA actions
                they believe would be necessary to support higher-octane mid-level fuel
                blends:
                 Set a minimum fuel octane level and phase out low-octane
                fuels as new optimized vehicles enter the market;
                 Approve a high-octane, mid-level ethanol blend vehicle
                certification fuel;
                 Correct the fuel economy formula by updating the R-Factor
                to be at or nearly ``1'' to reflect documented operation of modern
                engine technology;
                 Extend a RVP waiver of 1 psi to all gasoline containing at
                least 10 percent ethanol;
                 Adopt the Argonne National Laboratory GREET model to
                determine updated lifecycle carbon emissions for ethanol;
                [[Page 24388]]
                 Establish meaningful credits to automakers to incentivize
                transition to higher-octane fuel vehicles and continue to support flex-
                fuel vehicles; and
                 Provide equal treatment to vehicle technologies that
                reduce carbon emissions.
                 The Clean Fuels Development Coalition, et al. suggested that, ``the
                `ideal octane level' to optimize LDV performance, fuel efficiency, and
                reduce harmful emissions and consumer costs is 98-100 RON produced with
                E30+ `clean octane.' '' \721\ Concurrently, the HOLC Alliance and ACE,
                among others, also supported that 98 to 100 RON would be ideal octane
                levels for the nation.\722\
                ---------------------------------------------------------------------------
                 \721\ Clean Fuels Development Coalition, et al., Detailed
                Comments, NHTSA-2018-0067-11988.
                 \722\ HOLC Alliance, Detailed Comments, EPA-HQ-OAR-2018-0283-
                4196; ACE, Detailed Comments, EPA-HQ-OAR-2018-0283-4033.
                ---------------------------------------------------------------------------
                 BorgWarner, a supplier to major automobile manufacturers, commented
                that ``[f]uel octane is a limiting factor in the selection of
                compression ratio for all spark-ignition engines and the amount of
                boost for turbocharged engines. Higher-octane is particularly effective
                for using higher compression ratios with boosted engines,'' and stated
                that ``[t]here is substantial merit to raising the minimum octane
                required because current fuel pricing penalizes consumers for using
                higher-octane fuel. A base octane of 95 RON would be consistent with
                Europe. This would allow consistent development of engines for the
                broader US-EU market. Prior to the introduction of ethanol into
                gasoline, the base blend for regular fuel was typically 92 RON.
                Addition of 10% ethanol to this base blend gave 95 RON regular, so the
                base blend would be reformulated to retain the 92 RON at a lower cost.
                Returning to the previous base blend would be cost effective to the
                consumer.'' \723\
                ---------------------------------------------------------------------------
                 \723\ BorgWarner, Detailed Comments, EPA-HQ-OAR-2018-0283-4174.
                ---------------------------------------------------------------------------
                 Auto manufacturers also provided comment on the topic of higher-
                octane fuels. The Alliance of Automobile Manufacturers (the Auto
                Alliance) commented that it ``has long advocated for the availability
                of cost-effective, higher-octane fuel. The Alliance also believes the
                Agencies should require a transition to a higher minimum-octane
                gasoline (minimum 95-98 RON). There are several ways to produce higher-
                octane grade gasoline, such as expanding the ethanol availability, but
                the Alliance does not promote any sole or particular pathway.'' \724\
                The Alliance reiterated its position regarding fuel octane levels
                where, ``[t]he Alliance has long supported two goals regarding the
                octane (anti-knock) properties of gasoline: (1) The availability of
                cost effective higher-octane fuels, greater than 95 Research Octane
                Number (RON) and (2) the immediate elimination of subgrade fuel less
                than 87 anti-knock index (AKI).'' The Alliance also noted that ``[t]he
                higher-octane fuel that is available today is sold as a premium grade.
                To support future engine technologies, the approach taken with today's
                premium fuel option would not be expected to provide an attractive
                value proposition to the customer; therefore, a new higher minimum-
                octane gasoline, 95-98 RON, is needed to achieve anticipated
                performance.''
                ---------------------------------------------------------------------------
                 \724\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 Ford Motor Company agreed with the Auto Alliance's collective
                comments on fuel octane level and added specific support to raising
                minimum octane levels, stating that ``Ford concurs with those comments
                and supports increasing the marketplace octane rating in the U.S. to a
                minimum of 95 Research Octane Number (RON).'' Ford also generally
                supported the agencies' fuel octane discussion in terms of impacts to
                vehicle performance, where ``[h]igher octane gasoline enables
                opportunities for the use of key energy-efficient technologies,
                including: Higher compression ratio engines, lighter and smaller
                engines, improved turbocharging, optimized engine combustion phasing/
                timing, and low temperature combustion strategies. All of these
                technologies paired with higher-octane gasoline permit smaller engines
                to meet the demands of the consumer while at the same time providing
                higher overall efficiencies.'' \725\
                ---------------------------------------------------------------------------
                 \725\ Ford, Detailed Comments, EPA-HQ-OAR-2018-0283-5691.
                ---------------------------------------------------------------------------
                 Volkswagen commented ``[t]here may be several potential ways to
                achieve a high-octane fuel that may be more costly to the vehicle than
                others. Achieving an E10 high-octane fuel may mean a different hardware
                set than on E20 or E30 high-octane fuel. Elimination of sub-grades of
                market fuel (less than 87AKI) quickly is very important. If current 87
                AKI and 85 AKI fuels remain in the market for backward compatibility
                (such as if an E30 were chosen as the high-octane fuel of the future),
                a robust method at the fuel dispensing station and incorporated into
                the fueling station equipment to prevent mis-fueling is necessary.
                However, an E10 high-octane pathway might have far fewer compatibility
                problems and might bring extra fuel economy to the drivers of those
                current vehicles.'' \726\
                ---------------------------------------------------------------------------
                 \726\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
                ---------------------------------------------------------------------------
                 The agencies also received comments from the petroleum industry
                regarding higher-octane fuels. API commented that ``[g]iven the
                multiple engine technology pathways available to the automakers for
                achieving future fuel economy and CO2 emissions targets, the
                challenge of determining future market fuel gasoline octane number
                needs is complex and not yet settled. API believes that the octane
                number issue should be part of a comprehensive transport policy that
                addresses both vehicles and fuels as a system. API and its members are
                engaged in collaborations with the automakers and other stakeholders to
                better understand future fuel requirements for emerging powertrain
                technologies.'' API also commented ``the future for gasoline octane
                number will be driven by the stringency of regulations that set future
                fuel economy and CO2 requirements, the collective responses
                of the automakers to those regulations, consumer preferences regarding
                vehicles and fuels, and fuel supply economics. EPA's authority to
                regulate gasoline octane number is doubtful. Therefore, EPA should not
                attempt to regulate gasoline octane number at this time.'' \727\
                ---------------------------------------------------------------------------
                 \727\ API, Detailed Comments, EPA-HQ-OAR-2018-0283-5458.
                ---------------------------------------------------------------------------
                 In terms of challenges associated with potential high-octane fuel
                deployment, the American Fuel & Petrochemical Manufacturers (AFPM)
                commented that, ``[a]side from a lack of legal authority, EPA faces
                numerous technical, logistical, and legal challenges and uncertainties
                in requiring the use of higher-octane fuels. Any such requirement would
                need a separate rulemaking dedicated to such a purpose with an
                extensive technical record in support, including test data on vehicles
                designed for the higher-octane fuel and on the existing fleet with and
                without higher-octane.'' \728\
                ---------------------------------------------------------------------------
                 \728\ AFPM, Detailed Comments, EPA-HQ-OAR-2018-0283-5698.
                ---------------------------------------------------------------------------
                 AFPM also commented that it does not support the potential
                regulatory requirement for the production or use of higher octane
                gasoline as a compliance option. AFPM commented that EPA lacks the
                authority to require the use of higher octane fuels under CAA Sec.
                211(c)(1)(A). AFPM further commented ``[t]he only vehicles legally
                permitted to use more than 15 percent ethanol blends are flex-fuel
                vehicles, which are currently certified to utilize both E10 and E85.
                Without an alternative certification for an auto
                [[Page 24389]]
                manufacturer to build an E30 certified vehicle, which would require
                extensive testing and certification procedures as well as sufficient
                market availability of the certification fuel, it would be
                inappropriate for the Administration to consider such vehicles as a
                viable option in the 2022-2026 compliance period.''
                 Gasoline retailers also commented regarding higher-octane fuels.
                NACS and SIGMA commented that they support examining the use of such
                fuels as a potential path towards future emissions reductions and that
                it will be important that the agencies appropriately consider and
                address a variety of related issues, including:
                 1. How to allow and handle the expanded sales of higher-octane
                fuels, which may include fuels that currently face barriers to sale,
                such as E15;
                 2. Streamlining the registration and regulation of higher-level
                blends of ethanol;
                 3. Addressing misfueling liability concerns of retailers;
                 4. Streamlining federal labeling requirements and ensuring federal
                preemption of state requirements; and
                 5. Addressing any other regulatory and legislative challenges
                associated with the use of higher-octane fuels.\729\
                ---------------------------------------------------------------------------
                 \729\ Joint submission on behalf of NACS and SIGMA, Detailed
                Comments, EPA-HQ-OAR-2018-0283-5824.
                ---------------------------------------------------------------------------
                 NATSO commented that ``the Agencies should under no circumstances
                consider `requiring that mid-level [ethanol] blends be made available
                at service stations' '' and went on to say that ``retailers would need
                to be assured that they will not be held responsible for customers that
                misfuel . . . Federal dispenser labeling requirements would have to be
                streamlined and state requirements would have to be preempted. . . Auto
                manufacturers would have to warrant all new higher-octane vehicles up
                to at least E15 depending upon vehicles' capabilities, and would have
                to affirmatively state which cars in the existing fleet can run on E15
                and ensure that the cars are warrantied or retroactively warrantied as
                such.'' \730\
                ---------------------------------------------------------------------------
                 \730\ NATSO, Detailed Comment, EPA-HQ-OAR-2018-0283-5484.
                ---------------------------------------------------------------------------
                 UCS commented that ``[a]n orderly transition to high-octane fuel
                would take several years to complete. It will take time for the
                necessary regulations to be finalized, for vehicles optimized for high-
                octane gasoline to come to market and to build out the fuel
                distribution infrastructure to make this fuel broadly available. And
                even once high-octane gasoline is in use, it will take more time for
                automakers to phase-in new models optimized for high-octane fuel and to
                fully replace the legacy E10 fleet. Another factor to consider is that
                the rising share of high-octane gasoline will be buffered by falling
                sales of gasoline, given increasing fuel efficiency, such that the
                overall demand for ethanol will change more slowly. The agencies'
                expectation is that high-octane gasoline will not significantly enter
                commerce before 2026, and subsequently will only gradually gain market
                share through 2040. There is no realistic prospect of completing this
                process before 2025 or 2026, the timeframe of this rulemaking. The
                appropriate context for this discussion within vehicle rules is the
                next round of fuel economy and emission standards. Even then, an
                expeditious rulemaking process will be required to achieve adequate
                regulatory clarity to facilitate rapid adoption post-2026.'' UCS also
                commented ``[we] strongly oppose granting fuel economy credits based on
                the technical potential of vehicles to operate on high-octane fuel
                before there is clear evidence that high-octane fuel is in use and the
                potential fuel economy benefits are being realized on the road.'' \731\
                ---------------------------------------------------------------------------
                 \731\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
                ---------------------------------------------------------------------------
                 The agencies have reviewed the submissions received in response to
                their solicitation of comments concerning fuel octane levels and
                recognize the potential that higher-octane fuels, coupled with advanced
                engine technologies, can provide for improvements to fuel economy and
                tailpipe CO2 emissions. The agencies agree with commenters
                that establishing a higher minimum octane for gasoline is a complex
                undertaking that would require consideration of a wide array of
                difficult issues. In light of the complexity of the constellation of
                issues, the fact that EPA did not propose new octane requirements, and
                that EPA's authority to set fuel requirements resides in CAA section
                211(c)(1), the agencies recognize that the present rulemaking is not
                the appropriate vehicle to set octane levels. If EPA pursues future
                rulemaking action on this topic, it would consider these comments in
                that context and in consideration of the appropriate statutory
                provisions. The agencies note that the current vehicle certification
                process provides a path to certify a vehicle requiring the use of high-
                octane fuel, which allows the impact of such fuels to be captured over
                the required certification test cycles for CO2 emissions and
                fuel economy.
                 EPA also is declining to adopt new incentives for flex-fueled
                vehicles (FFVs) (vehicles designed to operate on gasoline or E85 or a
                mixture), as some commenters suggested. FFV incentives were not
                identified by EPA in its request for comments in the proposed rule and
                are outside the scope of this rulemaking.
                 The analyses conducted for this rulemaking assumed the use of Tier
                3 fuels, where applicable, which are considered directly
                representative, or a reasonable proxy for, fuels available for
                consumers to purchase. As explained in the previous paragraph, agency
                actions related to test fuels, consumer available fuels, or flexible-
                fuel incentives are out of scope of this rulemaking. However, to the
                extent that the agencies consider any additional rulemaking actions
                related to fuel octane requirements and/or availability, the agencies
                note that further analysis to set CAFE and CO2 standards
                would also reflect any potential, related impacts of those potential
                changes.
                b) Engine Maps
                 Engine paths include numerous engine technologies that
                manufacturers can use to improve fuel economy and reduce CO2
                emissions. Some engine technologies can be incorporated into existing
                engine design architectures with minor or moderate changes to the
                engine, but many engine technologies require an entirely new engine
                architecture or a major refresh. For this final rule analysis, twenty-
                three unique engine technologies are available for adoption, and are
                evaluated uniquely across the ten separate vehicle types (technology
                classes).
                 For the NPRM and final rule analysis, the impact of engine
                technologies on fuel consumption, torque, and other metrics was
                characterized using GT-POWER(copyright) modeling conducted by IAV
                Automotive Engineering, Inc. (IAV). IAV is one of the world's leading
                automotive industry engineering service partners and has extensive
                experience in testing and modeling engines and combustion. GT-POWER is
                a commercially available engine modeling tool with detailed cylinder
                and combustion modeling capabilities.\732\ GT-POWER is used to simulate
                engine behavior and provides data on engine metrics, including power,
                torque, airflow, volumetric efficiency, fuel consumption, turbocharger
                performance, and other parameters. The primary outputs of IAV's use of
                GT-POWER for this
                [[Page 24390]]
                analysis are the development of engine maps that provide operating
                characteristics of engines equipped with specific technologies.
                ---------------------------------------------------------------------------
                 \732\ More information regarding GT Power Modeling is available
                at https://www.gtisoft.com/gt-suite-applications/propulsion-systems/gt-power-engine-simulation-software.
                ---------------------------------------------------------------------------
                 When an engine is running, at any given point in time, the
                operation can be characterized by the engine's crankshaft rotational
                speed (typically in revolutions per minute, or RPM) and engine output
                (torque) level. Engines can operate at a range of engine speed and
                torque levels. Engine maps provide a visual representation of various
                engine performance characteristics at each engine speed and torque
                combination across the operating range of the engine. A common example
                of a performance characteristic is BSFC.\733\ Other characteristics
                include engine emissions, engine efficiency, and engine power.
                ---------------------------------------------------------------------------
                 \733\ The amount of fuel needed to achieve a specific power, or
                how efficiently an engine uses fuel to produce work.
                ---------------------------------------------------------------------------
                 Engine maps have the appearance of topographical maps, typically
                with engine speed on the horizontal axis and engine torque on the
                vertical axis. A third engine characteristic, BSFC, is displayed as
                contours, defining the operating regions for that BSFC with each
                contour showing all operating points at a specified BSFC value. Once
                created, the data they contain is referenced for engine fuel
                consumption at a given engine speed and torque operating point.
                 For the NPRM and final rule analysis, the agencies relied on IAV to
                develop engine maps representing each of the engine technologies. IAV
                used benchmark production engine test data, component test data, and
                manufacturers and suppliers' technical publications to develop a one-
                dimensional GT-POWER engine model for the baseline engine technology
                configuration. Technologies were incrementally added to the baseline
                model to assess their impact on fuel consumption. The following is a
                representative example of how IAV created the engine maps used in this
                analysis.
                 First, IAV defined the characteristics of Eng01 (a base VVT engine)
                and optimized it for all the combustion parameters while minimizing
                fuel consumption and maintaining performance. The result of this was a
                fuel map as a function of BMEP and engine RPM. IAV then took the same
                Eng01 and adopted characteristics of SGDI technology to the base
                engine. The new engine (Eng18, VVT and SGDI) was then optimized for all
                combustion parameters while minimizing fuel consumption and maintaining
                performance. The result was an engine fuel map for Eng18, as a function
                of BMEP and engine speed. The engine map is directly comparable to the
                engine map for Eng01 and the difference in those engine maps
                specifically identifies the effectiveness impact of VVT and SGDI
                technologies. This process was repeated for all of the IAV engine maps
                that used Eng01 (VVT) as the baseline engine. This methodology ensured
                the engine maps represent the maximum improvement in BSFC for each
                engine configuration change, while considering real world design
                constraints.
                 IAV used its global engine database that includes benchmarking
                data, engine test data, single cylinder test data, prior modeling
                studies, and technical publications and information presented at
                conferences to populate the assumptions and inputs used for engine map
                modeling, and to validate the ultimate results.\734\ Argonne used the
                engine maps resulting from this analysis as inputs for the Autonomie
                full vehicle modeling and simulation.
                ---------------------------------------------------------------------------
                 \734\ Friedrich, I., Pucher, H., and Offer, T., ``Automatic
                Model Calibration for Engine-Process Simulation with Heat-Release
                Prediction,'' SAE Technical Paper 2006-01-0655, 2006, https://doi.org/10.4271/2006-01-0655. Rezaei, R., Eckert, P., Seebode, J.,
                and Behnk, K., ``Zero-Dimensional Modeling of Combustion and Heat
                Release Rate in DI Diesel Engines,'' SAE Int. J. Engines 5(3):874-
                885, 2012, https://doi.org/10.4271/2012-01-1065. Multistage
                Supercharging for Downsizing with Reduced Compression Ratio (2015).
                MTZ Rene Berndt, Rene Pohlke, Christopher Severin and Matthias
                Diezemann IAV GmbH. Symbiosis of Energy Recovery and Downsizing
                (2014). September 2014 MTZ Publication Heiko Neukirchner, Torsten
                Semper, Daniel Luederitz and Oliver Dingel IAV GmbH.
                ---------------------------------------------------------------------------
                 As described in the NPRM and PRIA, the agencies developed engine
                maps for technologies that are in production today or that are expected
                to be available in the rulemaking timeframe. The agencies recognize
                that engines with the same combination of technologies produced by
                different manufacturers will have differences in BSFC and other
                performance measures, due to differences in the design of engine
                hardware (e.g., intake runners and head ports, valves, combustion
                chambers, piston profile, compression ratios, exhaust runners and
                ports, turbochargers, etc.), control software, and emission
                calibration. Therefore, the engine maps are intended to represent the
                levels of performance that can be achieved on average across the
                industry in the rulemaking timeframe.
                 Accordingly, the agencies noted that it was expected that the
                engine maps developed for this analysis will differ from engine maps
                for manufacturers' specific engines. For a given engine configuration,
                some production engines may be less efficient and some may be more
                efficient than the engine maps presented in the analysis. However, the
                agencies intended and expected that the incremental changes in
                performance modeled for this analysis, due to changes in technologies
                or technology combinations, will be similar to the incremental changes
                in performance observed in manufacturers' engines for the same changes
                in technologies or technology combinations. Most importantly, using a
                single engine model as a reference provides a common base for
                comparison of all incremental changes resulting from technology
                changes, and anchors incremental technology effectiveness values to a
                common reference. The effectiveness values from the internal simulation
                results were validated against detailed engine maps produced from
                engine benchmarking programs, as well as published information from
                industry and academia, ensuring reasonable representation of simulated
                engine technologies.\735\
                ---------------------------------------------------------------------------
                 \735\ Bottcher, L., Grigoriadis, P. ``ANL--BSFC map prediction
                Engines 22-26.'' IAV (April 30, 2019). 20190430_ANL_Eng 22-26
                Updated_Docket.pdf.
                ---------------------------------------------------------------------------
                 As discussed in the NPRM, the agencies updated the list of engine
                technologies, before and after the Draft TAR, based on stakeholder
                comments and consultations with CARB, Argonne, and IAV. The technology
                list was built on the technologies that were considered in the 2012
                final rule, and included technologies that are being implemented or
                that are under development and feasible for production in the
                rulemaking timeframe. The agencies noted that some advanced engines
                were included in the simulation that were, and often still are, not yet
                in production, and the engine maps for those engines were either based
                on CBI or theoretical data. The agencies also stated in the NPRM that
                the final rule analysis may include updated engine maps for existing
                modeled engines, or entirely new maps added to the analysis if either
                action could improve the quality of the fleet-wide analysis.
                 While there are a large number of possible combinations of engine
                technologies, the agencies categorized the IAV engine maps used in the
                NPRM full vehicle simulations into six categories. The categories were
                based on engine architecture and include: Dual overhead camshaft (DOHC)
                engines, single overhead camshaft (SOHC) engines, turbocharged engines,
                hybrid Atkinson cycle engines,\736\ non-hybrid
                [[Page 24391]]
                Atkinson mode engines, and diesel engines. Another unique technology
                that was available for adoption for the NPRM analysis was the advanced
                cylinder deactivation (ADEAC) for the SOHC and DOHC engines, however
                this technology was modeled using a fixed effectiveness value rather
                than an engine map, because the agencies did not have sufficient data
                to be used as input to the engine map or full vehicle simulation
                modeling. In addition, the agencies provided potential engine maps and
                additional specifications for several other technologies that could be
                considered for the final rule analysis. These included a basic high
                compression ratio Atkinson mode engine, a Miller cycle engine, and an
                engine with an electric assist.
                ---------------------------------------------------------------------------
                 \736\ These types of Atkinson cycle engines are mainly for
                hybrid applications like Toyota Prius or Ford C-Max.
                ---------------------------------------------------------------------------
                 The full list of engine maps used in the NPRM is presented in Table
                VI-39 below.
                BILLING CODE 4910-59-P
                [[Page 24392]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.143
                BILLING CODE 4910-59-C
                 The full list of engine maps used in this final rule analysis is
                presented in Table VI-40.
                [[Page 24393]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.144
                [[Page 24394]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.145
                [[Page 24395]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.146
                BILLING CODE 4910-59-C
                 Comments on engine maps varied, with industry commenters generally
                supporting the maps used in the NPRM analysis and CARB and
                environmental advocate commenters generally objecting to the maps. The
                Alliance argued that previously-modeled fuel efficiency improvements
                for downsized, turbocharged engine technologies were ``highly
                optimistic,'' and stated that the updated engine maps used for the NPRM
                analysis were an improvement.
                 ICCT argued that the IAV engine maps used for the NPRM analysis
                were out of date, and better engine maps benchmarked by EPA staff were
                available and should have been used instead.\737\ UCS similarly stated
                that Argonne work used for previous CAFE technical documents had relied
                on outdated engine maps, and that the new IAV engine maps used in this
                rulemaking were developed for a different purpose and had not been
                benchmarked against the latest engines either on the road or in
                development.\738\ ICCT questioned whether the agencies had validated
                engines 13 and 14 with physical testing and/or simulation modeling to
                the level of quality of EPA's simulation modeling.\739\ ICCT further
                asserted that EPA's benchmarked engine maps had been ``knowingly
                disregarded'' for the NPRM analysis, and stated that the NPRM analysis
                was therefore arbitrary.\740\ ICCT commented that the agencies must
                conduct and disclose a systematic investigation and comparison of
                engine benchmarking, engine modeling, and transmission modeling
                completed by EPA, Ricardo, and Argonne for model year 2014-2018
                vehicles. ICCT recommended that the agencies rely on engine maps used
                for past EPA ALPHA modeling while the agencies conduct such an
                investigation.
                ---------------------------------------------------------------------------
                 \737\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-49.
                 \738\ Union of Concerned Scientists, Technical Appendix, Docket
                No. NHTSA-2018-0067-12039, at 4.
                 \739\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-46.
                 \740\ ICCT, Docket No. NHTSA-2018-0067-11741, at I-49.
                ---------------------------------------------------------------------------
                 The agencies believe it is most important for engine map data to
                provide accurate BSFC information for known technologies and technology
                levels. The agencies disagree with statements that IAV engine maps are
                outdated. The majority of the engine maps were developed specifically
                to support the midterm review and encompass engine technologies that
                are present in the analysis fleet and technologies that could be
                applied in the rulemaking timeframe. In many cases those engine
                technologies are mainstream today and will continue to be during the
                rulemaking timeframe. For example, the engines on some MY 2017 vehicles
                in the analysis fleet have technologies that were initially introduced
                ten, or more, years ago. Having engine maps representative of those
                technologies is important for the analysis. The most basic engine
                technology levels also provide a useful baseline for the incremental
                improvements for other engine technologies. The timeframe for the
                testing or modeling is unimportant, because time by itself doesn't
                impact engine map data. A given engine or model will produce the same
                BSFC map regardless of when testing or modeling is conducted.
                Simplistic discounting of engine maps based on temporal considerations
                alone could result in discarding useful technical information. Also,
                narrow use of temporal considerations would also result in the
                discarding of several engine maps from Ricardo that were used for the
                EPA Draft TAR and Proposed Determination analyses.\741\ Therefore, with
                the engine maps used representing current technologies regardless of
                development date, the agencies do not agree with commenter assertions.
                ---------------------------------------------------------------------------
                 \741\ Ricardo, Inc. ``Computer Simulation of Light-Duty Vehicle
                Technologies for Greenhouse Gas Emission Reduction in the 2020-2025
                Timeframe.'' Ricardo (December 2011). https://nepis.epa.gov/Exe/ZyPDF.cgi/P100D57R.PDF?Dockey=P100D57R.PDF. Last accessed Jan 14,
                2020.
                ---------------------------------------------------------------------------
                 The same commenters also appear to misunderstand how the agencies'
                effectiveness data, including engine maps, were used in the NPRM
                analysis (and in past rulemakings). The analysis never applies absolute
                BSFC levels from the engine maps to any vehicle model or configuration
                for the rulemaking analysis. The absolute fuel economy values from the
                full vehicle Autonomie simulations are used only to determine
                incremental effectiveness for switching from one technology to another
                technology. The incremental effectiveness is applied to the absolute
                fuel economy of vehicles in the analysis fleet, which are based on CAFE
                compliance data. For subsequent technology changes, incremental
                effectiveness is applied to the absolute fuel economy level of the
                previous technology configuration. Therefore, for a technically sound
                analysis, it is most important that the differences in BSFC among the
                engine maps be accurate, and not the absolute values of the individual
                engine maps. However, achieving this can be challenging.
                 A technically sound approach is to use a single or very small
                number of baseline engine configurations with well-defined BSFC maps,
                and then, in a very systematic and controlled process, add specific
                well-defined technologies and create a BSFC map for each unique
                [[Page 24396]]
                technology combination. This could theoretically be done through engine
                or vehicle testing, but testing would need to be conducted on a single
                engine, and each configuration would require physical parts and
                associated engine calibrations to assess the impact of each technology
                configuration, which is impractical for the rulemaking analysis because
                of the extensive design, prototype part fabrication, development, and
                laboratory resources that are required to evaluate each unique
                configuration. Modeling is an approach used by industry to assess an
                array of technologies with more limited testing. Modeling offers the
                opportunity to isolate the effects of individual technologies by using
                a single or small number of baseline engine configurations and
                incrementally adding technologies to those baseline configurations.
                This provides a consistent reference point for the BSFC maps for each
                technology and for combinations of technologies which enables the
                differences in effectiveness among technologies to be carefully
                identified and quantified. The agencies selected this approach for the
                NPRM and final rule. Engine maps were created by IAV using this
                technically sound and rigorous methodology. Both absolute engine maps
                and the incremental differences in engine maps were presented in the
                PRIA.
                 Using a mix of engine maps from engine modeling and from
                benchmarking data provides no common reference for measuring impacts of
                adding specific technological improvements. In addition, as discussed
                in further detail in Section VI.C.1.e), manufacturers often implement
                multiple fuel-saving technologies simultaneously when redesigning a
                vehicle and it is not possible to isolate the effect of individual
                technologies by using laboratory measurements of a single production
                engine or vehicle with a combination of technologies. Because so many
                vehicle and engine changes are involved, it is not possible to
                attribute effectiveness improvements accurately for benchmarked engines
                to specific technology changes. This leads to overcounting or
                undercounting technology effectiveness.
                 Further, while two or more different manufacturers may produce
                engines with the same high level technologies (such as a DOHC engine
                with VVT and SGDI), each manufacturer's engine will have unique
                component designs that cause its version of the engine to have a unique
                engine map. For example, engines with the same high level technologies
                have unique intake manifold and exhaust manifold runners, cylinder head
                ports and combustion chamber geometry that impact charge motion,
                combustion and efficiency, as well as unique valve control, compression
                ratios, engine friction, cooling systems, and fuel injector spray
                characteristics, among other factors. The agencies developed and used a
                single engine map to represent each technology and each combination of
                engine technologies.
                 Therefore, it should not be expected that any of the agencies'
                engine maps would necessarily align with a specific manufacturer's
                engine, unless of course the engine map was developed from that
                specific engine. The agencies do not agree that comparing an engine map
                used for the rulemaking analysis to a single specific benchmarked
                engine has technical relevance, beyond serving as a general
                corroboration for the engine map. When a vehicle is benchmarked, the
                resulting data is dictated by the unique combination of technologies
                and design constraints for the whole vehicle system. For these reasons,
                the agencies do not agree with ICCT that Eng13 and Eng14 should be
                validated by conducting full vehicle modeling and comparing the results
                with a single benchmarked vehicle. The engine maps used in this
                analysis are precisely controlled for specific incremental technology
                adoption and not for comparisons of absolute performance of a specific
                vehicle's engine.
                 Differences are also explained by the NPRM and final rule analyses
                using large-scale full vehicle Autonomie simulations to estimate
                effectiveness instead of rough LPM approximations based on limited
                ALPHA simulation work.\742\ These issues are discussed in more detail
                in Section VI.B.3.
                ---------------------------------------------------------------------------
                 \742\ 2016 EPA Proposed Determination TSD at p.2-276 to 2-279
                ---------------------------------------------------------------------------
                 Accordingly, the agencies declined directly to use the Ricardo and
                other EPA engine maps created from engine benchmarking as inputs for
                this rulemaking because, among other reasons discussed below, they did
                not afford the opportunity to evaluate the effectiveness improvements
                for specific, individual technologies. For example, the 2018 Toyota
                Camry 2.5L engine that EPA benchmarked had a broad array of observable
                technologies, and several more that were not observable.\743\ However,
                there was no baseline from which to isolate or compare any of the
                individual technology improvements. For example, Toyota commented on
                this benchmarking, stating:
                ---------------------------------------------------------------------------
                 \743\ EPA Test Data. 2018 Toyota Camry 2.5L A25A-FKS Engine Tier
                3 Fuel. Available at https://www.epa.gov/sites/production/files/2019-04/2018-toyota-2.5l-a25a-fks-engine-tier3-fuel-test-data-package-dated-04-08-19.zip. Last accessed Nov. 20, 2019.
                 Past Toyota comments on Atkinson-cycle benefits have addressed
                only those derived from variable valve timing (VVT) with late intake
                valve closing (LIVC) that enables a 13:1 compression ratio. The
                total 18.6 percent improvement of the 2018 Camry 2.5L over the
                previous generation also includes benefits from cEGR and internal
                engine design changes such as to the block, cylinder head, pistons,
                valvetrain, as well as drivetrain and body/chassis
                enhancements.\744\
                ---------------------------------------------------------------------------
                 \744\ NHTSA-2018-0067-12431. Supplemental Comments--Toyota Motor
                North America, at p. 1-2.
                 Toyota's comments emphasize that the efficiency improvements in
                this engine were driven by several additional technological
                improvements, and not merely the cEGR, Atkinson cycle engine and higher
                compression ratio design that was assumed for the EPA Draft TAR and
                Proposed Determination analyses.\745\
                ---------------------------------------------------------------------------
                 \745\ EPA PD TSD at 2-229.
                ---------------------------------------------------------------------------
                 The agencies do agree component, engine, and vehicle test data are
                very important for validating systems models, such as Autonomie, and
                for validating model inputs, such as engine maps. Accordingly, the
                agencies did fully consider engine maps used in prior rulemakings,
                along with a broad array of other data as part of the process for
                evaluating the IAV engine maps used for the NPRM and the final rule
                analysis simulation work. Engine maps from Ricardo, EPA benchmarking,
                NHTSA-sponsored benchmarking,\746\ information from technical papers
                and conferences,\747\ extensive data and
                [[Page 24397]]
                expertise from the Argonne AMTL vehicle testing group and Energy
                modeling group, \748\ and the 2015 NAS report,\749\ were all sources
                used to confirm that incremental technology effectiveness estimates
                were appropriate. The engine maps developed by IAV provided reliable
                and reasonable estimates for the incremental impacts of engine
                technologies. The use of this approach explains some of the
                effectiveness differences between the NPRM and final rule analyses, and
                the EPA Draft TAR and Proposed Determination analyses.
                ---------------------------------------------------------------------------
                 \746\ NHTSA Benchmarking, ``Laboratory Testing of a 2017 Ford F-
                150 3.5 V6 EcoBoost with a 10 speed transmission.'' DOT HS 812 520.
                 \747\ Maruyama, F., Kojima, M., and Kanda, T., ``Development of
                New CVT for Compact Car,'' SAE Technical Paper 2015-01-1091, 2015,
                doi:10.4271/2015-01-1091. Shelby, M., Leone, T., Byrd, K., and Wong,
                F., ``Fuel Economy Potential of Variable Compression Ratio for Light
                Duty Vehicles,'' SAE Int. J. Engines 10(3):2017, doi:10.4271/2017-
                01-0639. Eisazadeh-Far, K. and Younkins, M., ``Fuel Economy Gains
                through Dynamic-Skip-Fire in Spark Ignition Engines,'' SAE Technical
                Paper 2016-01-0672, 2016, doi:10.4271/2016-01-0672. Wade, R.,
                Murphy, S., Cross, P., and Hansen, C., ``A Variable Displacement
                Supercharger Performance Evaluation,'' SAE Technical Paper 2017-01-
                0640, 2017, doi:10.4271/2017-01-0640. Hakariya, M., Toda, T., and
                Sakai, M., ``The New Toyota Inline 4-Cylinder 2.5L Gasoline
                Engine,'' SAE Technical Paper 2017-01-1021, 2017, doi:10.4271/2017-
                01-1021. Ogino, K., Yakabe, Y., and Chujo, K., ``Development of the
                New V6 3.5L Gasoline Direct Injection Engine,'' SAE Technical Paper
                2017-01-1022, 2017, doi:10.4271/2017-01-1022. Shibata, M., Kawamata,
                M., Komatsu, H., Maeyama, K. et al., ``New 1.0L I3 Turbocharged
                Gasoline Direct Injection Engine,'' SAE Technical Paper 2017-01-
                1029, 2017, doi:10.4271/2017-01-1029. Conway, G., Robertson, D.,
                Chadwell, C., McDonald, J. et al., ``Evaluation of Emerging
                Technologies on a 1.6 L Turbocharged GDI Engine,'' SAE Technical
                Paper 2018-01-1423, 2018, doi:10.4271/2018-01-1423.
                 \748\ ANL Energy Group. https://www.anl.gov/es; ANL AMTL group.
                https://www.anl.gov/es/advanced-mobility-technology-laboratory.
                 \749\ National Research Council. 2015. Cost, Effectiveness, and
                Deployment of Fuel Economy Technologies for Light-Duty Vehicles.
                Washington, DC--The National Academies Press, at pp. 294-305.
                https://doi.org/10.17226/21744.
                ---------------------------------------------------------------------------
                 In considering ICCT's comment about using IAV engine maps or EPA's
                engine maps, as an exercise, the agencies compared two IAV engine maps
                to the EPA's benchmarked Toyota 2.5L naturally aspirated engine and
                Honda's 1.5L turbocharged downsized engine.750 751 The IAV
                engines were modeled and simulated in a midsize non-performance vehicle
                with an automatic transmission and the same road load technologies,
                MR0, ROLL0 and AERO0, to isolate for the benefits associated with the
                specific engine maps.\752\ Eng 12, a 1.6L, 4 cylinder, turbocharged,
                SGDI, DOHC, dual cam VVT, VVL engine was selected as the closest engine
                configuration to the Honda 1.5L. Eng 22b, a 2.5L, 4 cylinder, VVT
                Atkinson cycle engine, was selected as the closest engine configuration
                to the Toyota 2.5L. As discussed before, both the Toyota 2.5L naturally
                aspirated engine and Honda's 1.5L engine have incorporated a number of
                fuel saving technologies including improved accessories and engine
                friction reduction. In order to assure an ``apples-to-apples''
                comparison, both IACC and EFR technologies were applied to the IAV
                engine maps. IACC technology provides an additional 3.6% incremental
                improvement and EFR provides an additional 1.4% incremental improvement
                beyond the IAV engine maps for midsize non-performance vehicles.\753\
                ---------------------------------------------------------------------------
                 \750\ Toyota 2.5L TNGA Prototype Engine From 2016 SAE Paper--
                ALPHA Map Package. Version 2017-12. Ann Arbor, MI: US EPA National
                Vehicle and Fuel Emissions Laboratory, National Center for Advanced
                Technology, 2017.
                 \751\ Honda 1.5L Turbo Prototype Engine From 2016 SAE Paper--
                ALPHA Map Package. Version 2017-12. Ann Arbor, MI: US EPA National
                Vehicle and Fuel Emissions Laboratory, National Center for Advanced
                Technology, 2017.
                 \752\ See ANL--All Assumptions_Summary_FRM_06172019_FINAL and
                ANL--Summary of Main Component Performance
                Assumptions_FRM_06172019_FINAL for midsize class characteristics.
                 \753\ The NPRM and this final rule analysis allowed the adoption
                of IACC technologies in the CAFE model that provided an additional
                3.6% incremental improvement for the midsize car vehicle class. As
                discussed in [Section VI.C Other Technologies], these benefits are
                not shown in the IAV engine simulated results, so they were added
                manually for this comparison.
                ---------------------------------------------------------------------------
                 The comparison shows effectiveness of the IAV engine maps and
                effectiveness values for the final rule analysis are in line with the
                Honda 1.5L and the Toyota 2.5L benchmarked engines. Figure VI-15 below
                shows the effectiveness improvements for the EPA benchmarked engines
                and the corresponding IAV engine maps incremental to a baseline
                vehicle. Accordingly, the agencies believe that the methodology used in
                this analysis, and the engine maps and incremental effectiveness values
                used, are in line with benchmarking data and are reasonable for the
                rulemaking analysis. The agencies believe the approach used in this
                rulemaking analysis appropriately allows the agencies to account for a
                wide array of engine technologies that could be adopted during the
                rulemaking timeframe. Declining to use manufacturer-specific engines
                allows the agencies to ensure that all effectiveness and cost
                improvements due to the incremental addition of fuel economy improving
                technologies are appropriately accounted for.
                [[Page 24398]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.147
                 Next, Roush Industries (``Roush''), writing on behalf of the
                California Air Resources Board, commented that the NPRM-modeled engines
                vary in cylinder size, which would significantly alter combustion, heat
                transfer, knock tolerance, and other important operating
                parameters.\754\ Roush stated that a more accurate simulation, which
                would improve incremental fuel economy improvement, should maintain a
                consistent cylinder displacement (500cc) and vary the number of
                cylinders or expected fuel consumption maps.\755\
                ---------------------------------------------------------------------------
                 \754\ Roush Industries on behalf of California Air Resources
                Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
                0067-11984, at 12.
                 \755\ Roush Industries on behalf of California Air Resources
                Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
                0067-11984, at 12.
                ---------------------------------------------------------------------------
                 The agencies believe that holding cylinder volume constant is the
                appropriate approach to research seeking to identify the impacts of
                technological changes on BSFC, torque, power, and other
                characteristics, when holding cylinder volume constant. However, as
                explained in Section VI.B.3.a)(2) Maintaining Vehicle Attributes and
                Section VI.B.3.a)(6) Performance Neutrality, CAFE and CO2
                rulemaking analyses attempt to maintain vehicle attributes, including
                performance, and hold all of the attributes constant when showing
                pathways that improve fuel economy. Therefore, the agencies' analyses
                require engine maps that attempt to hold performance constant--not
                necessarily cylinder size. Since certain fuel economy improving
                technologies would increase performance if cylinder size is held
                constant, such as when adding turbocharging technology, the agencies
                appropriately include changes in displacement and cylinder volume for
                technologies that have a significant impact on engine torque and power,
                such as turbocharging. For a number of fuel economy improving
                technologies that had smaller impacts on engine torque and power, the
                engine maps were created with cylinder volume held constant. Table VI-
                39 identifies the engine displacement information for each of the
                engine maps. For example, the same engine displacement (2.0 L) and
                cylinder displacement (500 cc) was used for creating engine maps for
                naturally aspirated engines Eng01, Eng02, Eng03, Eng04, Eng05a, Eng5b,
                Eng06a, Eng07a, and Eng08a, whereas engine displacement (1.6 L) and
                cylinder displacement (400 cc) is used for creating the engine map for
                turbocharged engine Eng12 in order to maintain performance. The
                agencies have concluded that the approach used for the NPRM and the
                final rule analysis is the most technically sound approach given the
                data needs and assessments required for CAFE and CO2
                rulemaking.
                 Roush also commented as follows:
                [S]everal of the base engine maps used in the 2018 PRIA analysis
                exhibit maximum thermal efficiency (lowest fuel consumption) at
                2000-3000 rpm and at maximum load, which is unrealistic for normal
                passenger vehicle engines. Such maps will over predict fuel economy
                for extremely down-sized applications (very small engine in a heavy
                vehicle). This is because there is no fuel economy penalty for
                running the engine at a high loads point where, in reality, BSFC is
                high due to retarding spark timing to prevent knocking and fuel
                enrichment to reduce exhaust temperatures to protect exhaust valves
                and turbocharger components.\756\
                ---------------------------------------------------------------------------
                 \756\ Roush Industries on behalf of California Air Resources
                Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
                0067-11984, at 11.
                 For example, Roush stated that Eng12 is predicted to have its
                highest efficiency at very high load and high engine speeds with no
                degradation in brake specific fuel consumption (BSFC) at engine speeds
                between 2,000 rpm and 4,500 rpm all the way up to peak load, which is
                unrealistic because
                [[Page 24399]]
                turbocharged engines at high loads require retarded spark timing to
                prevent knock and fuel enrichment to prevent overheating of the
                turbocharger and related components.\757\ Roush stated that these
                factors would increase fuel consumption and reduce efficiency under
                real-world conditions.\758\ Roush also stated that another effect of
                the Eng12 fuel consumption curve would be to predict unreasonably good
                fuel consumption at very high power levels for downsized turbocharged
                engines. Roush stated this could bias technology pathways in over-
                predicting fuel economy benefits for small engines installed in heavier
                vehicles, causing an overly optimistic predicted performance of the
                vehicle with regard to drivability, acceleration, and fuel consumption,
                which would create unrealistic real-world pathways to compliance.\759\
                ---------------------------------------------------------------------------
                 \757\ Roush Industries on behalf of California Air Resources
                Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
                0067-11984, at 18.
                 \758\ Roush Industries on behalf of California Air Resources
                Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
                0067-11984, at 19.
                 \759\ Roush Industries on behalf of California Air Resources
                Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
                0067-11984, at 23.
                ---------------------------------------------------------------------------
                 As discussed in the Argonne model documentation for the final rule
                analysis, the simulations used to determine incremental effectiveness
                for the NPRM and final rule analyses were conducted using 2-cycle test
                procedures, because they are the test procedures used for CAFE and
                CO2 compliance.\760\ Therefore, the engines maps are
                intended to represent BSFC accurately under those test conditions and
                do not need to capture BSFC under every operating condition. During 2-
                cycle test conditions, engines do not operate for extended periods at
                the speed and high load conditions noted by Roush. A few vehicle and
                engine combinations may operate at those speed and load points only
                briefly during the 2-cycle CAFE and CO2 tests. Engines are
                capable of operating for short periods of time under higher exhaust
                temperature conditions and manufacturers commonly delay fuel enrichment
                until it is needed to protect engine components (in particular exhaust
                valves and exhaust manifolds) from excessive temperatures that can
                impact engine durability. Fuel enrichment can be delayed because it
                takes a period of time at higher temperature for components to heat up
                and reach a temperature that would impact durability. Because these
                high speed and load conditions occur for a relatively short time during
                the CAFE and CO2 test cycles, and then return to lower speed
                and/or load conditions with lower exhaust temperature, engines operate
                for the entire CAFE and CO2 test cycles without triggering
                fuel enrichment. The fuel enrichment delay also enables vehicles to
                comply with criteria emission regulations and improves real world fuel
                economy. Therefore, the engine maps used for the NPRM and final rule
                analysis fully represent how engines operate during CAFE and
                CO2 test cycles, and properly do not include fuel enrichment
                at all 2-cycle operating conditions. Also, a trained knock model was
                used to develop the engine maps, and the spark timing reflects
                appropriate levels for engine operation during the delay in fuel
                enrichment.
                ---------------------------------------------------------------------------
                 \760\ A Detailed Vehicle Simulation Process To Support CAFE and
                CO2 Standards for the MY 2021-2026 Final Rule Analysis.
                ---------------------------------------------------------------------------
                 Next, regarding developing the NPRM engine maps to account for Tier
                3 test fuel, the Alliance and Ford stated that the engine maps using
                Tier 3 test fuel represented an improvement over prior analyses. The
                Alliance stated that previous EPA modeling had incorrectly used Tier 2
                premium octane fuel to predict the benefits of engine technologies,
                which overstated fuel economy gains that would be achievable when using
                regular-grade octane Tier 3 fuel. Ford provided similar comments, and
                also noted that regular grade octane fuel will be required for
                compliance after the 2020 model year.\761\
                ---------------------------------------------------------------------------
                 \761\ Ford Motors, Attachment, Docket No. EPA-HQ-OAR-2018-0283-
                5691, at 7.
                ---------------------------------------------------------------------------
                 In contrast, ICCT and UCS both commented that the agencies had
                incorrectly updated the IAV engine maps developed with Tier 2 test fuel
                to account for Tier 3 fuel.\762\ ICCT stated that the update reduced
                the effectiveness of the turbo technologies and suggested that the fuel
                update adjustment should not have been done at all, stating
                manufacturers that label vehicles as ``premium fuel recommended'' are
                required to show no emissions changes over all test cycles when using
                premium octane fuel and therefore reducing effectiveness for fuel
                differences, as the agencies did with the IAV engine maps, is
                unrealistic and inappropriate.
                ---------------------------------------------------------------------------
                 \762\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-82; Union of Concerned
                Scientists, Technical Appendix, Docket No. NHTSA-2018-0067-12039, at
                p. 15.
                ---------------------------------------------------------------------------
                 UCS also commented more specifically on the impact of the
                adjustment from Tier 2 to Tier 3 fuel related to the knock threshold
                for advanced engines, noting that manufacturers consider different
                approaches to different fuels, and not all of those approaches
                necessitate reductions in efficiency, as the agencies' assumption
                suggests. UCS stated that charge cooling can reduce knock in direct
                injection engines, resulting in an ``effective octane'' difference of a
                six point increase for E10, thus potentially compensating for the
                difference in octane between Tier 2 (E0 93 AKI) and Tier 3 (E10 87 AKI)
                fuels. UCS argued that excluding this consideration led the agencies to
                restrict advanced engines like HCR2 and reduce the effectiveness of
                turbocharged engines with CEGR. UCS suggested that there would be a
                reduction in the costs between the baseline and proposed standards if
                the analysis allowed the application of HCR2 engines and corrected the
                effectiveness of turbocharged CEGR engines.
                 Both ICCT and UCS also stated that the adjustment ignored a 2018
                EPA study showing that, while fuel consumption increases with the
                switch from Tier 2 to Tier 3 test fuel, emissions are reduced, meaning
                that the agencies' adjustment is wrong ``for some technologies because
                [CO2]-per-mile emissions can be lower with the switch to
                higher octane ethanol blends.'' UCS also stated that the adjustment
                factor applied is wrong for two reasons, first because converting
                solely with energy density would assume a 3.7 percent increase in fuel
                consumption compared to the observed 2.7 percent increase, and second
                because the adjustment goes in the wrong direction when applied to
                CO2 emissions, which show a reduction of 1.4 percent on the
                test cycle. UCS stated that the Autonomie model accordingly overstates
                CO2 emissions on Tier 3 fuel by 4.2 percent. UCS argued that
                the adjustment to account for Tier 3 test fuel therefore double counts
                any penalty in fuel economy and ignores CO2 tailpipe
                reductions, which would result in an improvement on the test cycle.
                Because the CAFE test procedure already has an adjustment in place to
                correct for fuel properties relative to 1975 test fuel, but carbon-
                related exhaust emissions do not, UCS stated that the fuel adjustment
                could lead to drastically conservative fuel economy and CO2
                curves.
                 ICCT stated that the agencies could fix this issue by relying on
                EPA's engine maps, where EPA had accounted for cost and effectiveness
                of technology used to protect operation on regular octane fuel by
                increasing costs and reducing effectiveness.
                 Some of these comments can be addressed with a simple
                clarification: The NPRM contained text that was
                [[Page 24400]]
                inconsistent regarding how the analysis accounted for the engine maps
                (which were based on Tier 3 fuel). The separate model documentation
                correctly described that, for the NPRM analysis, the agencies developed
                fuel maps for Tier 3 fuel and did not adjust the final Autonomie
                outputs.\763\ The NPRM text, however, incorrectly stated that ``(a)n
                adjustment factor was applied to the Autonomie simulation results to
                adjust them to reflect Tier 2 certification fuel. Argonne adjusted the
                vehicle fuel economy results to present certification fuel by using the
                ratio of the lower heating values to the rest and certification
                fuels.'' In fact, no adjustments were made to the NPRM Autonomie
                simulation outputs, as the modeled engine maps were appropriately
                modeled using Tier 3 fuel.
                ---------------------------------------------------------------------------
                 \763\ NHTSA-2018-0067-0007 at 177-178 and 191.
                ---------------------------------------------------------------------------
                 As discussed in detail in VI.C.1.a) Fuel Octane, engine
                specifications used to create the engine maps for the NPRM and the
                final rule were developed using Tier 3 fuel. Tier 3 fuel was used to
                ensure the engines were capable of operating on real world regular
                octane (87 pump octane = (R+M/2)). This capability is in line with what
                manufacturers must do to ensure engines have acceptable noise,
                vibration, harshness, drivability and performance levels, and will not
                fail prematurely when operated on regular octane fuel. If the agencies
                developed engine maps based on Tier 2 fuel alone, the engine maps would
                reflect the engines' ability to have higher compression ratios and to
                operate with greater levels of spark advance than could be implemented
                by manufacturers, who must take into account operation on regular
                octane fuels used by a majority of U.S. consumers.\764\ Not considering
                regular octane fuel operation by manufacturers would lead to engine
                durability, and engine noise, vibration, harshness, and drivability
                issues. Manufacturers have told the agencies that even for vehicles
                designed to operate on high octane fuel, the engines and controls must
                be designed to operate on every fuel octane level available in the U.S.
                to avoid these issues.\765\ Thus, developing engine maps based on Tier
                2 fuel alone would incorrectly overstate the BSFC improvements
                achievable in the real world.
                ---------------------------------------------------------------------------
                 \764\ Tamm, D.C., Devenish, G.N. Finelt, D.N. Kalt, L.K.
                ``Analysis of Gasoline Octane Costs'' Baiker and O'brien, Inc.
                Prepared for EIA. October 18, 2018. https://www.eia.gov/analysis/octanestudy/pdf/phase1.pdf at 11-13.
                 \765\ Ford Motor Company. NHTSA-2016-0068-0048 at 3. Auto
                Alliance comments for 2016 draft TAR. Attachment 7 Limitations of
                Ricardo Fuel Economy Analysis of Downsizing. NHTSA-2016-0068-0070.
                ---------------------------------------------------------------------------
                 Based on these comments and considerations, the agencies determined
                the engine maps developed for the NPRM appropriately account for fuel
                octane, and better approximate BSFC achieved by the majority of engines
                used in the U.S. vehicle fleet. The agencies believe ICCT's and other
                commenters' assertions that the engine maps should reflect Tier 2 fuel
                and not be updated for Tier 3 fuel would ignore these important
                considerations, and would provide engine maps that could not be
                achieved by engines in the real world. The agencies determined that
                engine maps developed for the Draft TAR and EPA Proposed Determination
                that were based on Tier 2 fuel should not be used for the NPRM and
                final rule analyses for these reasons.
                 EPA is addressing the impact of Tier 3 fuel on fuel economy and
                CO2 emissions compliance test results as part of a separate
                rulemaking. The separate rulemaking may establish an adjustment to
                account for the impacts of the change in test fuel. Those impacts are
                beyond the scope of this rulemaking. The analysis for this rule uses
                fuel economy and CO2 emissions of the vehicles in the MY
                2017 analysis fleet as the reference for absolute fuel economy and
                CO2 emissions. The analysis starts with absolute compliance
                data from MY 2017 and adopts technologies incrementally to determine
                future compliance. Because MY 2017 absolute compliance values are based
                on Tier 2 fuel, and standards are based on the use of Tier 2 fuel,
                there is no need to make any adjustments for the differences in energy
                content and carbon content of Tier 2 and Tier 3 fuel.\766\
                ---------------------------------------------------------------------------
                 \766\ During the 1980s, the U.S. Environmental Protection Agency
                (EPA) incorporated the R factor into fuel economy calculations in
                order to address concerns about the impacts of test fuel property
                variations on corporate average fuel economy (CAFE) compliance,
                which is determined using the Federal Test Procedure (FTP) and
                Highway Fuel Economy Test (HFET) cycles. The R factor is defined as
                the ratio of the percent change in fuel economy to the percent
                change in volumetric heating value for tests conducted using two
                differing fuels.
                ---------------------------------------------------------------------------
                 The agencies considered ICCT's statement that manufacturers that
                label vehicles as ``premium fuel recommended'' are required to show no
                emissions changes over all test cycles when using regular octane fuel,
                and therefore reducing effectiveness for fuel differences as the
                agencies did with the IAV engine maps is unrealistic and inappropriate.
                The agencies believe these conclusions are technically incorrect. The
                existence of an EPA compliance regulation does not impact the laws of
                nature, which govern issues associated with the impact of fuel octane
                on the ability to improve engine BSFC and on engine durability, noise,
                vibration, harshness, and drivability. It is widely recognized and
                accepted that higher octane fuels allow engines to be designed with
                higher compression ratios, faster combustion rates, and more optimal
                spark advance, which improve BSFC. Section VI.C.1.a) discusses comments
                advocating for increasing the minimum fuel octane specification to
                enable these improvements. The engine maps developed by IAV and used
                for the Draft TAR and NPRM were consistent with these trends and showed
                that BSFC is better with Tier 2 (higher octane) fuel than Tier 3 (lower
                octane) fuel.\767\ ICCT did not provide any data supporting the concept
                that there is no shift in BSFC, fuel economy, or CO2
                emissions when engines are optimized with different octane fuels, or
                between Tier 2 and Tier 3 fuel. It is appropriate to note that the EPA
                regulation does provide a tolerance which in practice allows a small
                level of shift in emissions.\768\
                ---------------------------------------------------------------------------
                 \767\ See BSFC difference between engines modeled with Tier 3
                fuel versus high octane fuel by IAV in PRIA 6.3.2.2.20.9 at 288 to
                PRIA 6.3.2.20.11 at 292.
                 \768\ 40 CFR 1066.210 (b) Accuracy and Precision.
                ---------------------------------------------------------------------------
                 Regarding comments that certain combinations of technologies can
                enable BSFC improvements while controlling spark knock, the agencies in
                fact considered a very broad array of engine technology combinations
                for the analysis, including several added technologies as discussed
                further below. The agencies believe the rigorous methodology used to
                develop the engine maps resulted in engine maps representing the
                maximum improvement in BSFC for each engine configuration, while also
                addressing real world constraints. Engine maps for the new technologies
                were presented in PRIA Chapter 6.3.2.2.16.4. The PRIA also discussed
                that IAV maps were developed considering a very comprehensive list of
                combustion operating parameters as part of the IAV GT-Power engine
                modeling. IAV's GT-Power engine modeling included sub-models to account
                for heat release through a predictive combustion model, knock
                characteristic through a kinetic fit knock model, physics-based heat
                flow model physics based friction model, and IAV's proprietary
                Optimization Tool Box.\769\ These independent models were
                [[Page 24401]]
                run concurrently to make sure engine design requirements were met for
                each engine configuration that was modeled.
                ---------------------------------------------------------------------------
                 \769\ IAV's Optimization Tool Box is a module of IAV Engine. IAV
                Engine, as the basic platform for designing engine mechanics,
                provides a large number of tools that have proven their worth across
                the globe in several decades of automotive development work at IAV.
                The modules help designers, computation engineers and simulation
                specialists in designing mechanical engine components--for example,
                in laying out valvetrains and timing gears as well as crankshafts.
                ---------------------------------------------------------------------------
                 Finally, in response to the agencies' request for comment on
                including the additional engine maps presented in the NPRM as potential
                technological pathways, several commenters stated that the agencies
                should include those technologies, in addition to other emerging engine
                technologies.\770\ After considering these comments, the agencies added
                several engine technologies and technology combinations to the final
                rule analysis. The additions included a basic high compression ratio
                Atkinson mode engine (HCR0), a variable compression ratio engine (VCR),
                a variable turbo geometry engine (VTG), and a variable turbo geometry
                with electric assist engine (VTGe). The agencies also added advanced
                cylinder deactivation technology (TURBOAD) to Eng12 (TURBOD) in the
                Autonomie modeling for the final rule analysis. Like with ADEAC, the
                agencies did not have IAV engine maps for TURBOAD, so the agencies took
                the effectiveness values as predicted by full vehicle simulations of a
                TURBOD and added 1.5 percent or 3 percent respectively for I-4 engines
                and V-6 or V-8 engines, as explained in more detail further below. The
                agencies also included more iterations of existing technologies, like
                diesel engines with cylinder deactivation, diesel engines paired with
                manual transmissions, and diesel engines paired with 12-volt start stop
                technology, in addition to more combinations of hybrid technologies
                that are discussed further in Section VI.C.3, below.
                ---------------------------------------------------------------------------
                 \770\ ICCT Docket # NHTSA-2018-0067-11741 at I-19--I-22; CARB
                Docket # NHTSA-2018-0067-11873 at 107-108.
                ---------------------------------------------------------------------------
                 The following sections list and describe the comprehensive set of
                engine technologies and combinations of engine technologies that have
                been included in the analysis. The agencies also discuss the additional
                engine technologies added for the final rule, and reasons for excluding
                a small number of technologies proffered by commenters. The agencies
                believe the wide array of engine technologies included in the final
                rule analysis and the methodology used to develop the engine maps to
                measure the effectiveness of those technologies reasonably represents
                the scope of technologies that should be considered during the
                rulemaking timeframe.
                c) Engine Modeling in the CAFE Model
                (1) Basic Engines
                 The NPRM described that there are a number of engine technologies
                that manufacturers can use to improve fuel economy and CO2
                emissions. Some engine technologies can be incorporated into existing
                engines with minor or moderate changes to the engines, but many engine
                technologies require an entirely new engine architecture. The terms
                ``basic engine technologies'' and ``advanced engine technologies'' are
                used only to define how the CAFE model applies a specific engine
                technology and handles incremental costs and effectiveness
                improvements. ``Basic engine technologies'' refer to technologies that,
                in many cases, can be adapted to an existing engine with minor or
                moderate changes to the engine, compared to ``advanced engine
                technologies'' that generally require significant changes or an
                entirely new engine architecture.
                 In the CAFE model, basic engine technologies may be applied in
                combination with other basic engine technologies; advanced engine
                technologies (defined by an engine map) stand alone as an exclusive
                engine technology. The words ``basic'' and ``advanced'' are not meant
                to confer any information about the level of sophistication of the
                technology. Also, many advanced engine technology definitions include
                some basic engine technologies, but these basic technologies are
                already accounted for in the costs and effectiveness values of the
                advance engine. The ``basic engine technologies'' need not be (and are
                not) applied in addition to the ``advanced engine technologies'' in the
                CAFE model.
                (a) DOHC
                 In the NPRM analysis, the agencies characterized dual overhead cam
                (DOHC) engine technology as ``basic.'' DOHC engine configurations have
                two camshafts per cylinder head, one operating the intake valves and
                one operating the exhaust valves. Four basic engine technologies--
                variable valve timing (VVT), variable valve lift (VVL), stoichiometric
                gasoline direction injection (SGDI), and basic cylinder deactivation
                (DEAC)--were considered for DOHC engines. Implementing these
                technologies involves changes to the cylinder head of the engine, but
                the engine block, crankshaft, pistons, and connecting rods require few,
                if any, changes.
                 Variable valve timing (VVT) is a family of valve-train designs that
                dynamically adjusts the timing of the intake valves, exhaust valves, or
                both, in relation to piston position. VVT can reduce pumping losses,
                provide increased engine torque and horsepower over a broad engine
                operating range, and allow unique operating modes, such as Atkinson
                cycle operation, to further enhance efficiency. VVT is nearly
                universally used in the MY 2017 fleet.\771\ In the NPRM analysis, the
                VVT technology modeled by IAV was based on dual (independent) cam
                phasing. This was a more advanced VVT technology that allowed
                controlling of valve overlap, which can be used to control internal EGR
                to minimize fuel consumption at low engine loads.\772\ VVT enables
                control of many aspects of air flow, exhaust scavenging, and combustion
                relative to fixed valve timing engines. Engine parameters such as
                volumetric efficiency, effective compression ratio, and internal
                exhaust gas recirculation (iEGR) can all be enabled and accurately
                controlled by a VVT system.
                ---------------------------------------------------------------------------
                 \771\ 98.1 percent of MY2017 vehicles are equipped with VVT. EPA
                Report. The 2018 EPA Automotive Trends Report. https://nepis.epa.gov/Exe/ZyPDF.cgi/P100W5C2.PDF?Dockey=P100W5C2.PDF at
                Table 4.1 Production Share by Engine technology.
                 \772\ 2015 NAS at p. 32.
                ---------------------------------------------------------------------------
                 Variable valve lift (VVL) dynamically adjusts the distance a valve
                travels from the valve seat optimizing airflow over a broad range of
                engine operating conditions. The technology can increase effectiveness
                by reducing pumping losses and may improve efficiency by affecting in-
                cylinder charge (fuel and air mixture), motion, and combustion. VVL is
                less common in the 2017 fleet than VVT. Some manufacturers have
                implemented a limited, discrete approach to VVL where just two valve
                lift profiles are available versus a full-range, continuously variable
                implementation.
                 Stoichiometric gasoline direct injection (SGDI) sprays fuel at high
                pressure directly into the combustion chamber, which provides cooling
                of the in-cylinder charge via in-cylinder fuel vaporization to improve
                spark knock tolerance and enable an increase in compression ratio and/
                or more optimal spark timing for improved efficiency. SGDI appears in
                about half of basic engines produced in MY 2017, and the technology is
                used in many advanced engines as well.\773\
                ---------------------------------------------------------------------------
                 \773\ 49.7 percent of MY2017 vehicles are equipped with SGDI.
                EPA Report. The 2018 EPA Automotive Trends Report. https://nepis.epa.gov/Exe/ZyPDF.cgi/P100W5C2.PDF?Dockey=P100W5C2.PDF at
                Table 4.1 Production Share by Engine technology.
                ---------------------------------------------------------------------------
                 Basic cylinder deactivation (DEAC) disables intake and exhaust
                valves and
                [[Page 24402]]
                turns off fuel injection for the deactivated cylinders during light-
                load operation. The engine runs temporarily as though it were a smaller
                engine, which reduces pumping losses and improves efficiency. In the MY
                2017 fleet, manufacturers used DEAC on V6, V8, V10, and V12 engines in
                OHV, SOHC, and DOHC engine configurations. With some engine
                configurations in some operating conditions, DEAC creates noise-
                vibration-and-harshness (NVH) challenges. NVH challenges are
                significant for V6 and I4 DEAC configurations, and limit the operating
                range where DEAC can operate. For I4 engine configurations with smaller
                displacements, there are fewer operating conditions where engine load
                is low enough to use DEAC, which limits effectiveness. No manufacturers
                produced I4 DEAC engines in MY 2017. Typically, the smaller the engine
                displacement, the less opportunity DEAC provides to improve fuel
                consumption.
                 The agencies provided engine fuel maps for each of the eight DOHC
                engines (Eng01, Eng02, Eng03, Eng04, Eng18, Eng19, Eng20, and Eng21)
                used for the NPRM analysis. Each of these engines incrementally added
                technology to Eng01, a basic VVT engine, while holding all other
                factors constant like ambient temperature, ambient pressure, and fuel
                type.
                 For the NPRM analysis, the agencies estimated the effectiveness of
                DEAC using full vehicle modeling and simulation. In the NPRM PRIA
                6.2.1.2, the agencies discussed how Autonomie uses a specific control
                logic for cylinder deactivation for naturally aspirated engines that
                takes into consideration for noise, vibration, and harshness.\774\ For
                the final rule analysis, the agencies took steps to use full vehicle
                modeling and simulation to apply DEAC to both naturally aspirated and
                turbocharged engines. The same control logic was applied to the
                turbocharged engine cylinder deactivation (TURBOD) for the final rule
                analysis.
                ---------------------------------------------------------------------------
                 \774\ NHTSA-2018-0067-1972. ``Preliminary Regulatory Impact
                Analysis (PRIA) The Safer Affordable Fuel-Efficient (SAFE) Vehicles
                Rule for Model Year 2021-2026 Passenger Cars and Light Trucks,'' at
                191.
                ---------------------------------------------------------------------------
                 The agencies used the same assumptions for advanced cylinder
                deactivation (ADEAC) in the final rule analysis. In the NPRM the
                agencies stated engine maps were not available at the time of the
                analysis, and said that ADEAC was estimated to improve a basic engine
                with VVL, VVT, SGDO, and DEAC by three percent (for 4 cylinder engines)
                and six percent (for engines with more than 4 cylinders).\775\ The new
                technology combination for turbocharged advanced cylinder deactivation
                (TURBOAD) uses a similar approach for determining effectiveness. The
                agencies have applied a one-and-a-half percent effectiveness
                improvement estimate for 4-cylinder or smaller engines and a three
                percent effectiveness estimate for 6-cylinder or larger engines
                relative to TURBOD.
                ---------------------------------------------------------------------------
                 \775\ 83 FR 430039 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 For the final rule analysis the basic engine path for DOHCs are
                shown in Figure VI-16 and the high-level engine specifications are
                shown in Table VI-41. The baseline basic DOHC engine, Eng01, was the
                starting point and other engine technologies were incrementally adopted
                to determine effectiveness. Adoption of DEAC technology for
                turbocharged engines is discussed in Section VI.C.1.e)(2). Similarly,
                ADEAC technology is discussed in Section VI.C.1.e)(4).
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.148
                [[Page 24403]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.149
                BILLING CODE 4910-59-C
                (b) SOHC
                 Similar to DOHC engines, SOHC engines were characterized as
                ``Basic'' engine technologies in the NPRM analysis. They are
                characterized by having a single camshaft in the cylinder head
                operating both the intake and exhaust valves. Four basic engine
                technologies, VVT, VVL, SGDI, and DEAC were considered for SOHC
                engines. Implementing these technologies involves changes to the
                cylinder head of the engine, but the engine block, crankshaft, pistons,
                and connecting rods require few, if any, changes.
                 The agencies provided engine fuel maps for each of these types of
                SOHC engines and requested comments. Engine maps 5b, 6a, 7a, and 8a
                were modeled SOHC engines. The SOHC engine models used engine 5a, which
                was based on Eng01 as a reference, by removing one camshaft. Eng5a was
                included for the Draft TAR, but not included for the NPRM analysis due
                to high BSFC from higher friction that was inherited from the DOHC
                engine design. A level 0.1 bar of friction reduction over the entire
                operating range for engine maps 5b, 6a, 7a, and 8a was applied to
                represent improvements over existing engine designs. The addition of
                friction reduction to these engines was a result of consideration of
                deliberative interagency comments received during the Draft TAR review
                process noting higher fuel consumption on the baseline SOHC engine 5a
                relative to other modern SOHC engines.
                 Meszler on behalf of NRDC commented that ``[a]lthough variable
                valve timing (VVT) technology is identified as an available refresh
                technology, the NPRM CAFE model (unlike the version used for the 2016
                TAR analysis) actually assumes that all baseline vehicles include VVT
                technology. As a result, the approximately 9 percent of model year 2016
                sales that do not actually include VVT are not credited with any
                efficiency benefit for adoption of the technology . . . . '' \776\
                ---------------------------------------------------------------------------
                 \776\ Meszler, at 32.
                ---------------------------------------------------------------------------
                 We agree with this comment, and for the final rule analysis updated
                the CAFE model to add a non-VVT level engine in the 2017 analysis fleet
                and to allow those vehicles to adopt VVT technologies at a refresh or
                redesign. However, the agencies did not have engine maps for the non-
                VVT engines, so the agencies applied a fixed-value effectiveness
                estimate from similar VVT engine maps to represent the effectiveness
                for non-VVT engines. The agencies used the effectiveness of a similar
                configuration technology package of another engine to represent non-VVT
                engines. Non-VVT SOHC engines may add any combination of VVL with SGDI
                and DEAC. The agencies believe that the estimated effectiveness used
                for VVT engines was appropriate because the effectiveness offset is in
                line with 2015 NAS estimates for VVT engines with respect to VVL
                engines.777 778
                ---------------------------------------------------------------------------
                 \777\ Baseline effectiveness references for SOHC;VVT; SGDI;
                AT5;CONV;ROLL0;MR0;AERO0, SOHC;VVT; DEAC; AT5;CONV;ROLL0;MR0;AERO0,
                SOHC;VVT;VVL; DEAC; AT5;CONV;ROLL0;MR0;AERO0, and SOHC;VVT;
                SGDI;DEAC; AT5;CONV;ROLL0;MR0;AERO0 were used to represent SOHC;VVL;
                SGDI; AT5;CONV;ROLL0;MR0;AERO0, SOHC;VVL;DEAC;
                AT5;CONV;ROLL0;MR0;AERO0, and SOHC;VVL; SGDI;DEAC;
                AT5;CONV;ROLL0;MR0;AERO0 baseline combinations. These combinations
                represented only 2% of the models and 3.1% sales by volume in the MY
                2017 baseline fleet.
                 \778\ 2015 NAS Table 2.7 and Table 2.8 at 32-33.
                ---------------------------------------------------------------------------
                 The basic engine path for SOHC engines used in this final rule is
                shown in Figure VI-17 and the specifications are shown in Table VI-42.
                Note, that Eng5a is only a reference used to build the rest of the SOHC
                engines.
                BILLING CODE 4910-59-P
                [[Page 24404]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.150
                [GRAPHIC] [TIFF OMITTED] TR30AP20.151
                BILLING CODE 4910-59-C
                (2) Turbocharged Downsized Engines
                 Engine maps 12, 13, and 14 modeled turbocharged downsized engines.
                Turbocharged downsized engines are characterized by technology that can
                create greater-than-atmospheric pressure in the engine intake manifold
                when higher output is needed. The raised pressure results in an
                increased volume of airflow into the cylinder supporting combustion,
                increasing the specific power of the engine. An increased specific
                power means the engine can generate more power per unit of volume,
                which allows engine volume to be reduced while maintaining performance,
                thereby increasing fuel efficiency. IAV Eng12 was the base engine for
                all simulated turbocharged engines and was validated using engine
                dynamometer test data.\779\
                ---------------------------------------------------------------------------
                 \779\ Bottcher, L. Grigoriads, P. ``ANL--BSFC map prediction
                Engines 22-26'' April, 30, 2019. IAV_20190430_Eng 22-26
                Updated_Docket.pdf.
                ---------------------------------------------------------------------------
                 One notable change that the agencies made for the NPRM analysis
                based on stakeholder comments to the Draft TAR was to update the turbo
                family engine maps to assume operation on regular octane fuel (Tier 3,
                or 87 AKI), instead of premium fuel (Tier 2, or 93 AKI), to assure the
                maps accounted for real world constraints that impact durability and
                drivability, and noise, vibration, and harshness. Using regular octane
                fuel is consistent with the fuel octane that manufacturers specify be
                used in the majority of vehicles (manufacturers generally only specify
                premium fuel is required for higher performance models, although that
                is not always the case), and enables the modeling to account for
                important design and calibration issues associated with regular octane
                fuel. The agencies noted in the NPRM that using the updated engine maps
                addressed over-estimation of potential fuel economy improvements and
                ensured that the analysis reflected real-world constraints faced by
                manufacturers to assure engine durability and acceptable drivability.
                Importantly, assuming no change in fuel octane required to
                [[Page 24405]]
                operate a vehicle ensures that the agencies are modeling technology
                pathways that can improve fuel economy while maintaining vehicle
                performance, capability, and other attributes.
                 Compared with the NHTSA analysis in the Draft TAR, the turbocharged
                and downsized engine maps adjusted at high torque and low speed
                operation, and at high speed operation to account for knock limitations
                when using regular octane fuel. The knock model used to develop the
                turbocharged engines was trained on production and development engines
                tested at IAV to quantify the effects of different octane fuels.\780\
                Below the knock threshold, there is no change to the fuel consumption
                maps. The agencies noted that with the fuel octane change there are
                generally two major effects in the regions where the engine is knock-
                limited: First, spark timing is retarded causing a reduction in
                combustion efficiency and hence an increase in BSFC, and second, an
                increase in combustion and exhaust temperatures requiring fuel
                enrichment to cool those temperatures for engine component protection
                and resulting in increased BSFC.781 782
                ---------------------------------------------------------------------------
                 \780\ Knock models are based on Gamma Technology's kinetic fit
                model per the technical paper titled, ``A combustion model for IC
                engine combustion simulations with multi-component fuels,'' by
                YoungChul Ra, Rolf D. Reitz--Engine Research Center, University of
                Wisconsin-Madison.
                 \781\ Fuel enrichment is extra fuel is injected at the intake
                manifold port or directly into the cylinder. Fuel vaporization and
                the fuel's thermal mass reduces combustion and exhaust temperatures.
                Changes to the air/fuel ratio also impact combustion speed which
                impacts the knock limit.
                 \782\ Singh, E. and Dibble, R., ``Effectiveness of Fuel
                Enrichment on Knock Suppression in a Gasoline Spark-Ignited
                Engine,'' SAE Technical Paper 2018-01-1665, 2018, https://doi.org/10.4271/2018-01-1665.
                ---------------------------------------------------------------------------
                 The agencies also noted that for Eng14, the turbocharged downsized
                engine with cooled exhaust gas recirculation (cEGR), cEGR was added at
                the higher speeds where further reduction in combustion temperature was
                required. The higher specific heat capacity of cEGR reduced the need
                for fuel enrichment by lowering combustion temperatures and limiting
                the amount of spark retardation necessary to manage spark knock. With
                increasing load, cEGR is also used to lower combustion temperatures to
                reduce NOx emissions. The agencies explained that because IAV's models
                are not trained for emissions, cEGR was only considered for areas that
                are knock-limited and/or to reduce combustion temperatures. Because
                cEGR has the impact of slowing down burn rates, the amount of cEGR that
                could be utilized was balanced to maintain efficient combustion.
                Combustion stability was also evaluated to assure cEGR rates did not
                cause excessive cycle-to-cycle combustion variations, which adversely
                impact drivability.\783\
                ---------------------------------------------------------------------------
                 \783\ Heywood. B. J, Internal Combustion Engine Fundamentals, at
                413-37, McGraw-Hill (1988).
                ---------------------------------------------------------------------------
                 Some commenters criticized these downsized turbocharged IAV maps,
                referencing deliberative EPA comments docketed pursuant to the Clean
                Air Act procedural requirements at 42 U.S.C. 7607, which stated that
                the assumptions for Eng12's fuel octane, heating value, and carbon
                content were not representative of certification fuel and did not
                appear to be consistently used for the various engine maps, concluding
                that the resultant engine maps were not representative of
                CO2 performance of turbocharged engines over the
                certification cycle. ICCT stated it appeared these concerns had not
                been addressed for the NPRM, and that ``this problem essentially
                affect[ed] all engines on the turbocharged engine pathway.'' \784\
                ---------------------------------------------------------------------------
                 \784\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-46.
                ---------------------------------------------------------------------------
                 The agencies disagree with ICCT's comments relating both to whether
                fuel specifications were used consistently and whether the fuel
                specifications for fuel octane, heating value and carbon content were
                representative of the same fuel. First, the EPA deliberative comments
                were resolved in the deliberative process through the clarification
                that a single fuel specification was used to develop all of the engines
                and engine maps. Therefore, the engine maps are internally consistent.
                The fuel specification was presented in the NPRM section PRIA Chapter
                6.3.2.2.17. Second, the agencies considered future fuel and emissions
                standards by using regular octane fuel for this analysis. The
                assumptions for the fuel used in this analysis align with the EPA's
                Tier 3 standards that went into effect January 1, 2017.\785\ For the
                reasons discussed further above, the agencies believe it is important
                to use Tier 3 fuel for engine maps used for rulemaking analysis.
                ---------------------------------------------------------------------------
                 \785\ Final Rule for Control of Air Pollution from Motor
                Vehicles: Tier 3 Motor Vehicle Emission and Fuel Standards. https://www.epa.gov/regulations-emissions-vehicles-and-engines/final-rule-control-air-pollution-motor-vehicles-tier-3. Last accessed September
                26, 2019. Docket EPA-HQ-OAR-2011-0135.
                ---------------------------------------------------------------------------
                 Roush claimed that the turbocharged engine maps used in the
                analysis were responsible for an overly-conservative estimate of
                underlying combustion engine efficiencies, arguing that many production
                engines available today use the same technology packages identified in
                the PRIA but with significantly higher efficiencies.\786\ Roush noted
                that the base turbocharged engine map used in the PRIA, Eng12, is
                assumed to have variable valve lift (VVL), but with a turbocharged
                engine the benefit of VVL over dual variable valve timing (VVT) is
                limited.\787\ Roush argued that almost all vehicle manufacturers use
                lower-cost dual VVT systems in their turbocharged engines, and that the
                agencies' base turbocharged engine assumption is unrealistic with a
                correspondingly high cost.\788\
                ---------------------------------------------------------------------------
                 \786\ Roush Industries on behalf of California Air Resources
                Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
                0067-11984, at 16.
                 \787\ Roush Industries on behalf of California Air Resources
                Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
                0067-11984, at 17.
                 \788\ Roush Industries on behalf of California Air Resources
                Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
                0067-11984, at 17.
                ---------------------------------------------------------------------------
                 Roush contrasted its critique of Eng12 with an EPA ALPHA run of a
                2016 Honda Civic 1.5L turbocharged engine (L15B7) with continuously
                variable intake and exhaust camshaft phasing (CVVT), which is less
                expensive than the CVVL, arguing that it showed greater efficiency over
                more of the engine map at a lower cost than Eng12. Roush further argued
                that since the L15B7 engine is the first generation of the new Honda
                turbocharged engine, ``even further fuel consumption improvement is
                highly likely in the period through MY2025.'' \789\
                ---------------------------------------------------------------------------
                 \789\ Roush Industries on behalf of California Air Resources
                Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
                0067-11984, at 18.
                ---------------------------------------------------------------------------
                 As the agencies explained further above, from a technical
                perspective there is no reason why the 2016 Honda Civic 1.5 L Turbo
                should have an engine map that is the same as Eng12, Eng13, or Eng14.
                The turbocharged engine technologies represented by Eng12, Eng13 and
                Eng14 are not representative of any specific engine from any one
                manufacturer. Honda's 1.5L turbocharged engine incorporates a unique
                combination of technologies including electric wastegate, sodium-filled
                exhaust valves, light weight internal components, friction reduction
                technologies, 2-stage oil pump, low viscosity oil (0W-20), and a unique
                exhaust system.\790\
                ---------------------------------------------------------------------------
                 \790\ Honda Press Release. ``2016 Honda Civic Sedan Press Kit--
                Powertrain'' October 18, 2015. https://hondanews.com/en-US/releases/2016-honda-civic-sedan-press-kit-overview?page=178. Last accessed
                Feb. 12, 2020.
                ---------------------------------------------------------------------------
                 While there are an enormous number of different technology
                combinations that manufacturers could apply on their
                [[Page 24406]]
                engines, the agencies' analysis must select a reasonable number of
                configurations--in fact, the agencies analyze thousands of unique make/
                model/powertrain combinations and apply them to over one hundred
                thousand unique technology combinations for each of ten classes for
                this rulemaking. See Section VI.B.3.a)(6) and Section VI.B.3 for more
                details. For turbocharged engines, the agencies selected eight
                combinations which represent a wide range of technologies, combinations
                of technologies, and effectiveness improvements for the rulemaking
                analysis, as listed in Table VI-40. Three of the combinations were
                added based on commenter's recommendations. While it is possible to
                identify other combinations, such as the unique technologies Honda
                chose for its 1.5L Turbo engine, agencies do not believe it would be
                appropriate to select all of the technologies on one specific
                manufacturer's engine for the rulemaking analysis. Doing so would,
                appropriately, raise questions about the availability of proprietary
                designs and controls to other manufacturers, among other
                considerations.
                 The agencies also believe that the engine maps for Eng12, Eng13 and
                Eng14 show reasonable differences in BSFC maps that characterize the
                impact of each of these technology combinations, and differences
                relative to naturally aspirated engines. As discussed further above,
                incremental differences in BSFC are used for the rulemaking analysis.
                Roush's comments center on the comparison of absolute effectiveness
                values for a specific production vehicle, and do not address
                incremental effectiveness among a range of technologies, nor the
                appropriate baseline reference for the Honda 1.5L Turbo for technology
                content and for effectiveness. The ALPHA simulation for the 2016 Honda
                Civic 1.5L turbocharged engine provides absolute test data and has no
                baseline for assessing incremental effectiveness. Because there is no
                baseline, there is no basis for identifying which specific technologies
                have changed, nor any basis for determining the incremental
                effectiveness of each individual technology.
                 Regarding Roush's comment that that further fuel consumption
                improvement for the Honda L15B7 is highly likely in the period through
                MY 2025, Roush provided no information or data on what specific
                technologies would further improve the fuel consumption of that engine.
                With no defined new technology to consider, there is no basis for
                estimating the costs, nor for estimating the effectiveness of Roush's
                assertion. Without further information, the agencies can only point to
                the additional engine technologies considered for this final rule,
                discussed further below.
                 ICCT also stated that IAV's handling of cooled EGR (cEGR) in the
                engine maps was inappropriate, as IAV analyzed cEGR as a knock-
                abatement technology instead of a fuel efficiency technology. ICCT
                stated that this is reason that the NPRM analysis showed no benefit to
                cEGR, and if the agencies had used EPA's properly modeled cEGR
                effectiveness based on validated data, the effectiveness of cEGR would
                have been more realistic.
                 Similarly, Roush commented that cEGR application in the modeled
                turbocharged engines is excluded in engine operating modes that highly
                influence vehicle fuel economy. Roush contrasted Eng13, a turbocharged
                engine with VVT, direct injection, and cEGR, with the Mazda 2.5L
                SkyActiv Turbo engine available in the 2016 Mazda CX-9, which also
                employs cEGR.
                 The agencies believe Eng14 was created and modeled using a sound
                technical methodology, using constraints that the industry uses to
                ensure the engines would meet durability and customer acceptability
                criteria. IAV turbocharged engines adopted VVT and VVL to maximize
                volumetric efficiency and improve the combustion process. Engines with
                VVT control intake and exhaust valve timing to recycle burned exhaust
                gas into the combustion chamber. The recycling of exhaust gases using
                VVT is commonly called internal EGR. Cooled EGR (cEGR) is a second
                method for diluting the incoming air that takes exhaust gases, passes
                them through a cooler to reduce their temperature, and then mixes them
                with incoming air in the intake manifold. Diluting the incoming air
                with inert exhaust gas reduces pumping losses, thereby improving BSFC.
                The dilution also reduces combustion rates, temperatures, and
                pressures, which mitigates spark knock and reduces the need for fuel
                enrichment at higher loads to control exhaust temperature for component
                durability (typically, exhaust valves and exhaust manifold). Not only
                does this exhaust gas displace some incoming air, but it also heats the
                incoming air and lowers its density. Both interactions lower the
                volumetric efficiency of the engine.\791\ Cooled EGR is a more
                effective way of reducing combustion temperature in higher load and
                higher speed engines like turbocharged engines.
                ---------------------------------------------------------------------------
                 \791\ Volumetric efficiency (VE) in internal combustion engine
                engineering is defined as the ratio of the mass density of the air-
                fuel mixture drawn into the cylinder at atmospheric pressure (during
                the intake stroke) to the mass density of the same volume of air in
                the intake manifold. Ideally, you want this to be high as possible
                to maximize thermal efficiency during the power stroke (combustion
                phase).
                ---------------------------------------------------------------------------
                 As mentioned above, IAV developed engine specifications, including
                the rate of internal EGR and cEGR, using variation in combustion
                criteria used by industry to ensure the engines would meet durability
                and customer acceptability criteria. In addition to reducing pumping
                losses, EGR slows the combustion rate and causes combustion to be less
                consistent cycle-to-cycle as the concentration increases. Industry and
                researchers use a measurement known as coefficient of variation of
                indicated mean effective pressure (COV of IMEP) to evaluate combustion
                stability. Industry commonly recognizes values greater than 3.0 percent
                as unacceptable because above those levels, the combustion instability
                creates a noticeable and objectionable drivability problem for vehicle
                occupants, referred to as ``surge.'' Surge is perceived as the vehicle
                accelerating and decelerating erratically, instead of running smoothly.
                IAV set EGR rates at each of the engine operating conditions at the
                highest level that did not exceed 3.0 percent COV of IMEP. Therefore,
                the IAV engine maps did maximize efficiency within real-world
                constraints, similar to how manufacturers develop their engines. At the
                lower speed and load conditions of the 2-cycle tests, the COV of IMEP
                threshold was reached using internal EGR alone, so additional cEGR was
                not applied. At higher load conditions, such as the US06 cycle, cEGR
                was applied.
                 ICCT's statement that the engine maps were only developed
                considering knock-abatement is inaccurate. In the PRIA Chapter
                6.3.2.2.11, the agencies discussed the application of internal EGR in
                combination with cEGR for Eng14. VVT technology, with which Eng14 is
                equipped, maximizes EGR usage first in areas where the engine primarily
                operates, such as low load and low speed area like city cycle and
                highway cycle tests used in CAFE compliance testing. Cooled EGR is
                applied at higher speed and higher load conditions, such as the US06
                test cycle.
                 Using EPA's modeled cEGR would have resulted in infeasible engine
                maps because they were developed assuming the exclusive use of high
                octane Tier 2 fuel, and using a COV of IMEP threshold of 5 percent,
                which is beyond the level that is deemed acceptable to consumers in the
                real world.\792\ The use of these
                [[Page 24407]]
                criteria results in engine maps with BSFC levels that cannot be
                achieved by manufacturers that must ensure their engines are durable
                and are acceptable to customers with fuels that are used and available.
                The reference engine for EPA's cEGR concept was a 2010 Ricardo
                prototype V6 engine that used 98 RON fuel (93AKI or premium fuel) to
                determine effectiveness.\793\ The problems associated with using high
                octane Tier 2 to develop engine maps are discussed in detail in Section
                VI.C.1.a). The issues associated with excessive cEGR rates and COV of
                IMEP, are discussed immediately above. In addition, the cEGR engine
                maps that EPA used were never evaluated with regular octane Tier 3 fuel
                to assess the further degradation in BSFC and COV of IMEP that would
                occur where spark advance would need to be decreased to address spark
                knock, as decreasing spark advance directionally makes both BSFC and
                COV of IMEP worse.\794\ Also, because some models are still under
                development, ALPHA effectiveness estimates in the Draft TAR and derived
                for the Proposed Determination do not provide the best available basis
                for assessing effectiveness impacts.\795\ Therefore, the assumptions
                used for the EPA Draft TAR and Proposed Determination engine maps
                overstate feasible improvements and therefore do not provide meaningful
                comparisons to the engine maps used for the NPRM and final rule
                analyses.
                ---------------------------------------------------------------------------
                 \792\ EPA Proposed Determination TSD at 2-295.
                 \793\ 2016 EPA Technical Support Document at p. 2-312 in section
                2.3.4.1.9 Table 2.69. EPA-420-R-16-021, November 2016. Available at
                https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100Q3L4.pdf.
                 \794\ 2016 EPA Technical Support Document at p. 2-312 in section
                2.3.4.1.9. EPA-420-R-16-021, November 2016. Available at https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100Q3L4.pdf.
                 \795\ Advanced Light-Duty Powertrain and Hybrid Analysis (ALPHA)
                Tool. Available at https://www.epa.gov/regulations-emissions-vehicles-and-engines/advanced-light-duty-powertrain-and-hybrid-analysis-alpha#v1.0. Version 2.2. Incomplete Models in
                ALPHA2.2_TechWalkExamples\Ford Tech Walk\publish_Escape_AWD_matrix.
                ---------------------------------------------------------------------------
                 Finally, with regards to Roush's comparison of Eng13 to the 2016
                Mazda SkyActiv-G 2.5L Turbo, the agencies believe these engines use
                technologies that are sufficiently different so as to render a
                comparison not useful, even for a very rough validation of Eng13. Most
                fundamentally, as discussed in PRIA Chapter 6.3.2.2.11 and 6.3.2.2.13,
                the Mazda 2.5L Turbo is a Miller cycle engine, whereas Eng13 is an Otto
                cycle engine. Also, the Mazda 2.5L Turbo has cEGR, whereas Eng13 does
                not.\796\ On a more detailed level, as described in PRIA Chapter
                6.3.2.2.20.10, Eng13 has a BSFC of 238 g/kwh, whereas Roush refers to
                an engine having a BSFC of 250 g/kwh.\797\ The agencies therefore
                believe comparing the 2016 Mazda SkyActiv-G 2.5L Turbo to Eng13 is not
                a useful or relevant comparison. In the PRIA, the agencies included an
                engine map for a Miller cycle engine and requested comments on whether
                it should be included in the final rule analysis. Based on the
                comments, as discussed further below, the agencies added a Miller cycle
                engine to the final rule analysis.
                ---------------------------------------------------------------------------
                 \796\ NHTSA Benchmarking, ``Laboratory Testing of a 2016 Mazda
                CX9 2.5 I4 with a 6 Speed Transmission.'' DOT HS 812 519.
                 \797\ NHTSA-2018-0067-11984 at p. 20 of 37 Figure 8.
                ---------------------------------------------------------------------------
                (3) Non-HEV Atkinson Mode Engines
                 Manufacturers use a variety of designs and technologies to obtain
                an engine's highest thermal efficiency while maintaining drivability
                and performance. While the Otto cycle has historically been used by the
                vast majority of gasoline based engines, one way to improve thermal
                efficiency is by using alternative combustion cycles. One such
                alternative combustion cycle that can be used in place of the Otto
                cycle to achieve a higher maximum thermal efficiency is the Atkinson
                cycle. Atkinson cycle operation is achieved by modifying the Otto cycle
                engines' crank and valvetrain mechanics to maintain compression ratio
                while increasing expansion ratio.798 799 800 Specifically,
                in Otto cycle operation, the exhaust valve is opened near the end of
                the power stroke, allowing exhaust gases out of the cylinder. The
                pressure in the cylinder is still about three to five atmospheres.\801\
                Currently, there are two common approaches to achieving Atkinson Cycle
                operation: Either the exhaust valve timing or the intake valve timing
                are modified. In the first instance, the exhaust valve is not opened
                until enough expansion has occurred for the cylinder pressure to be
                equivalent to atmospheric pressure. The energy that typically is lost
                when the exhaust valve opens in Otto cycle is captured in the Atkinson
                cycle, leading to higher thermal efficiency. Modifying the intake valve
                timing, the most common way to achieve Atkinson cycle operation,
                involves allowing the intake valve to stay open during some portion of
                compression stroke. As a result, some of the fresh charge is driven
                back into the intake manifold by the raising piston so the cylinder is
                never completely filled with air, allowing optimized capture of
                combustion-created pressure.
                ---------------------------------------------------------------------------
                 \798\ Otto cycle is a four-stroke cycle that has four piston
                movements over two engine revolutions for each cycle. First stroke:
                Intake or induction; seconds stroke: Compression; third stroke:
                Expansion or power stroke; and finally, fourth stroke: Exhaust.
                 \799\ Compression ratio is the ratio of the maximum to minimum
                volume in the cylinder of an internal combustion engine.
                 \800\ Expansion ratio is the ratio of maximum to minimum volume
                in the cylinder of an IC engine when the valves are closed (i.e.,
                the piston is traveling from top to bottom to produce work).
                 \801\ Pulkrabek. W.W. ``Engineering Fundamentals of the Internal
                Combustion Engine.'' 2nd edition. Pearson Prentice Hall, at p. 118.
                ---------------------------------------------------------------------------
                 While Atkinson cycle engines have higher theoretical thermal
                efficiency compared to Otto cycle engines, the Atkinson cycle engine
                delivers that higher efficiency at the cost of power density.\802\ The
                reduced power density is because of lower operation pressures in the
                cylinder than in a typical Otto cycle engine. Accordingly, Atkinson
                cycle engines have been ideal for hybrid vehicles because their
                electric motor can make up for lost power density.
                ---------------------------------------------------------------------------
                 \802\ Power density is the engine power per unit of displacement
                (= [Engine Power]/[Engine Displacement]).
                ---------------------------------------------------------------------------
                 As vehicle technologies have become more sophisticated,
                descriptions of Atkinson cycle engines and Atkinson mode engine
                technologies have been used interchangeably, and often incorrectly, in
                association with high compression ratio (HCR) engines by the agencies
                and stakeholders. Although they both achieve an overall higher thermal
                efficiency than Otto cycle-only engines, they differ in execution
                depending on engine load. For the following discussion, Atkinson
                technologies considered in the analysis can be categorized into three
                groups: (1) Atkinson engines, (2) Atkinson-mode engines, and (3)
                Atkinson-enabled engines, which are variable valve timing engines with
                late intake closing that enables the Atkinson cycle mode. As discussed
                earlier, because power density is traded for efficiency, there is a
                limit to where Atkinson technology can be applied. While any vehicle
                could, theoretically, adopt an Atkinson-mode engine or an engine that
                enables operating in Atkinson cycle mode, the difference in vehicle
                application (high-performance versus standard-performance vehicles,
                towing requirements, trucks) leads to different effectiveness levels.
                The range of effectiveness appeared to create confusion among
                stakeholders regarding how the technology is applied to vehicles for
                compliance modeling and simulation.
                 Atkinson engines are engines that operate full-time in the Atkinson
                cycle. As mentioned above, the most common method of operation used by
                Atkinson engines currently is late intake closing.
                [[Page 24408]]
                This approach allows backflow from the combustion chamber into the
                intake manifold, reducing the dynamic compression ratio, but providing
                a higher expansion ratio. This improves thermal efficiency but reduces
                power density. As a result of limited engine operation, these engines
                tend to have lower specific power.\803\ The lower specific power tends
                to relegate these engines to hybrid vehicles applications, as coupling
                the engines to electric motors can compensate for the lower specific
                power. The Toyota Prius is an example of a vehicle that uses an
                Atkinson engine. Typically, vehicles that use an Atkinson cycle engine
                incorporate various fuel-efficient technologies like aerodynamic
                improvements, advanced continuously variable transmissions, mass
                reduction, and many other technologies to minimize engine load and
                attain high thermal efficiency.\804\ The 2017 Toyota Prius achieved a
                peak thermal efficiency of 40 percent.\805\
                ---------------------------------------------------------------------------
                 \803\ Specific power is the maximum power produced per
                displacement typically in units of hp/L or kw/l.
                 \804\ Toyota. ``Under the Hood of the All-new Toyota Prius.''
                Oct. 13, 2015. Available at https://global.toyota/en/detail/9827044.
                Last accessed Nov. 22, 2019.
                 \805\ Matsuo, S., Ikeda, E., Ito, Y., and Nishiura, H., ``The
                New Toyota Inline 4 Cylinder 1.8L ESTEC 2ZR-FXE Gasoline Engine for
                Hybrid Car,'' SAE Technical Paper 2016-01-0684, 2016, https://doi.org/10.4271/2016-01-0684.
                ---------------------------------------------------------------------------
                 Atkinson-mode engines are engines that use both the Otto cycle and
                Atkinson cycle during operation, switching between the modes of
                operation based on engine loads. During high loads the engine will
                operate in the power-dense Otto cycle mode, while at low loads the
                engine will operate in the higher-efficiency Atkinson cycle mode. The
                magnitude of efficiency improvement experienced by a vehicle using this
                technology is directly related to how much of the vehicle's operation
                time is spent in Atkinson mode. This means vehicles that typically
                operate at a high load, like a truck towing a trailer, will spend more
                time in the Otto mode and less time in the Atkinson cycle mode, and
                will achieve a lower overall efficiency improvement over a traditional
                Atkinson engine that operates full-time in the Atkinson cycle. As a
                result, manufacturers will try to use this type of engine in
                conjunction with other technologies that reduce engine load, which
                allows the engine to operate more frequently in Atkinson cycle mode.
                For example, manufacturers could reduce parasitic losses by
                incorporating more efficient accessory technologies, or reducing
                overall vehicle mass and aerodynamic drag. These technologies are
                enablers for Atkinson-mode engines. When these types of technologies
                are adopted, it reduces the parasitic losses and, in turn, reduces the
                time the engine is in high load region. An example of an Atkinson-mode
                engine is the MY 2017 Mazda 3.
                 The last type of Atkinson-type engine, the Atkinson-enabled engine,
                can be characterized by primarily running the Otto cycle, but can
                achieve Atkinson-mode using variable valve timing (VVT) technology.
                Some engines use changes in VVT on the intake side to enable Atkinson
                cycle operation in low load, low speed operation, like city driving.
                These types of engines are typically used in applications that
                generally require higher specific power such that it would be
                infeasible to use Atkinson-mode engines or Atkinson engines. These
                vehicles tend to have higher load demands due to towing requirements,
                payload requirements, greater aerodynamic drag from larger frontal
                areas, greater tire rolling resistance from larger tires and higher
                driveline losses from four-wheel drive or all-wheel drive (e.g., SUVs
                and pickup trucks). These higher load demands tend to push these
                engines more frequently to the less efficient region of the engine map
                and limit the amount of Atkinson operation. An example of the Atkinson-
                enabled engine is the Toyota MY 2017 Tacoma 3.5L 6-cylinder engine.
                 EPA developed two engine maps representing non-hybrid Atkinson
                engines to support the 2016 Draft TAR, Proposed Determination, and
                first Final Determination.\806\ Referred to as ATK and ATK2, the
                engines represented a current non-hybrid Atkinson cycle engine based on
                the 2.0L 2014 Mazda SkyActiv-G (ATK) engine, and a future Atkinson
                engine concept based on the Mazda engines, but adding cooled EGR,
                cylinder deactivation, and an increased compression ratio (14:1)
                developed for full vehicle modeling and simulation (ATK2). For the 2016
                Draft TAR, the agencies adopted EPA's high compression ratio (HCR)
                engine maps as Eng24 and Eng25, which corresponded to HCR1 and HCR2 in
                the CAFE modeling.
                ---------------------------------------------------------------------------
                 \806\ 2016 LD Draft Technical Assessment Report (TAR), Vehicle
                Greenhouse Gas Emission Standards and Corporate Average Fuel Economy
                Standards for Model Years 2022-2025; at p. 5-282. Proposed
                Determination on the Appropriateness of the Model Year 2022-2025
                Light-Duty Vehicle Greenhouse Gas Emissions Standards under the
                Midterm Evaluation; pp. 22 & A-7. Final Determination on the
                Appropriateness of the Model Year 2022-2025 Light-Duty Vehicle
                Greenhouse Gas Emissions Standards under the Midterm Evaluation,
                Response to Comments; pp. 29 & 52.
                ---------------------------------------------------------------------------
                 The Alliance had provided significant comments on the 2016 Draft
                TAR regarding the engine maps for HCR engines.\807\ The Alliance
                detailed concerns regarding the feasibility and effectiveness of Eng24
                (HCR1) and Eng25 (HCR2). Many of the comments on the 2016 Draft TAR
                noted that the modeling projected an implausible rapid fleet
                penetration for these technologies, and overestimated effectiveness.
                Commenters stated the overestimation was due largely to modeling with
                use of high-octane fuel and the addition of other technologies like
                cEGR and cylinder deactivation (DEAC) using theoretical assumptions
                that exceed the bounds of operation of components. In contrast, other
                commenters had stated that EPA's work on the future Atkinson concept
                ``has shown this pathway to be a promising alternative way to match the
                levels of improvement from a 27-bar BMEP turbocharged engine,'' and
                that ``it is prudent to assume that the robust body of evidence EPA is
                putting together based on benchmarking and modeling data is a
                reasonable assessment of the technology's potential.'' \808\
                ---------------------------------------------------------------------------
                 \807\ Alliance of Automobile Manufacturers, Alliance of
                Automobile Manufacturers Comments on Draft Technical Assessment
                Report: Midterm Evaluation of Light-Duty Greenhouse Gas Emission
                Standards and Corporate Average Fuel Economy Standards for Model
                Years 2022-2025 (EPA-420-D-16-900, July 2016), at 45 (Sept. 26,
                2016), Docket ID EPA-HQ-OAR-2015-0827-4089 and NHTSA-2016-0068-0072.
                 \808\ Union of Concerned Scientists Comments Concerning the
                Draft Technical Assessment Report for the Mid-term Evaluation of
                Model Year 2022-2025 Light-duty Vehicle Greenhouse Gas Emissions and
                Fuel Economy Standards, at 10-11.
                ---------------------------------------------------------------------------
                 For the NPRM analysis, the agencies included EPA's engine maps. The
                agencies allowed HCR1 to be applied only for a few manufacturers that
                indicated they would pursue this technology pathway versus alternative
                pathways, such as downsized turbocharged engines. The agencies were
                also careful to maintain vehicle performance and utility attributes
                when considering the application of Atkinson-type technologies. Current
                Atkinson capable engines have incorporated other technologies to reduce
                load in order to maximize time in Atkinson operation and to offset the
                power loss partially. This includes improved accessories, addition of
                friction reduction technologies, and other technologies that reduce
                engine load. Although modern improvements to engines have allowed
                Atkinson operation to occur more often (because of lower engine loads)
                for passenger cars, larger vehicles capable of carrying more cargo and
                occupants, and towing larger and heavier trailers, have more limited
                potential Atkinson operation. Those
                [[Page 24409]]
                adoption features are discussed further in Section VI.C.1.e) Adoption
                Features, below.
                 As stated in the NPRM, the agencies excluded the HCR2 concept
                engine from the central analysis for several reasons. First, the
                concept was not subjected to validation to assess its technical
                feasibility. The concept was only modeled with high octane Tier 2 fuel.
                The HCR2's capability to operate on regular octane Tier 3 fuel was
                assessed using non-cycle specific operation, necessitating adjustments
                to the final results to account for Tier 3 fuel properties from Tier 2
                operation, instead of simply operating the engine on Tier 3 to generate
                effectiveness estimates.\809\ As discussed further above and in Section
                VI.C.1.a), fuel octane affects engine durability, performance,
                drivability, and noise, vibration and harshness. Assumptions about
                compression ratio, EGR rates, and use of cylinder deactivation were not
                adequately validated. PRIA Chapter 6.3.2.2.20.18 discussed many
                questions about HCR2 technology's practicability as specified,
                especially in high load, low engine speed operating conditions. There
                also has been no observable physical demonstration of the technology
                assumptions. Many manufacturer engine experts questioned its technical
                feasibility and commercial practicability during the model years
                covered by the rulemaking. Stakeholders like the Alliance had
                previously asked for the engine to be removed from the rulemaking
                analyses until the performance could be validated with engine
                hardware.\810\ For these reasons, the agencies considered the HCR2
                engine too speculative to include in the NPRM central analysis.
                However, the agencies did provide a sensitivity analysis that included
                the HCR2 engine.
                ---------------------------------------------------------------------------
                 \809\ EPA PD TSD at 2-210.
                 \810\ NHTSA-2016-0068-0070 at 45.
                ---------------------------------------------------------------------------
                 Comments on HCR1 and HCR2 varied, with commenters split on issues
                like whether HCR2 was speculative or real, whether there was technology
                in the fleet that could adequately be represented by HCR2, and the
                effectiveness of HCR2 in the analysis.
                 The Alliance commented in support of the decision to exclude HCR2
                from the analysis, citing previous comments to the Draft TAR and
                proposed determination ``detailing concerns of feasibility and
                effectiveness of the non-hybrid Atkinson engine technology packages,
                including cooled exhaust gas recirculation (``CEGR'') and cylinder
                deactivation.'' \811\ Specifically, the Alliance's comments ``noted
                that the modeling projected an implausibly rapid fleet penetration of
                this complex engine technology and overestimated its effectiveness, due
                largely to modeling with high-octane fuel and the theoretical addition
                of CEGR plus cylinder deactivation.'' The Alliance concluded that ``the
                inexplicably high benefits ascribed to this theoretical combination of
                technologies has not been validated by physical testing.'' Ford
                commented that previous assessments had ``over-estimated both the
                effectiveness and near-term penetration of advanced Atkinson technology
                powertrains,'' stating that ``[t]he effectiveness of the `futured'
                Atkinson package (HCR2) that includes cooled exhaust gas recirculation
                (CEGR) and cylinder deactivation (DEAC) is excessively high, primarily
                due to overly-optimistic efficiencies in the base engine map,
                insufficient accounting of CEGR and DEAC integration losses, and no
                accounting of the impact of 91RON Tier 3 test fuel. Given the
                speculative and optimistic modeling of this technology combination,
                Ford supports limiting the use of HCR2 technology to reference only, as
                described in the Proposed Rule.'' \812\ Separately, in support of its
                overarching comments that the NPRM modeling better reflected reality
                over prior regulatory assessments, Toyota commented that the
                effectiveness estimates for Atkinson cycle engine technology in the
                NPRM may still have been overstated.\813\
                ---------------------------------------------------------------------------
                 \811\ NHTSA-2018-0067-12073.
                 \812\ NHTSA-2018-0067-11928.
                 \813\ NHTSA-2018-0067-12150.
                ---------------------------------------------------------------------------
                 In contrast, CARB, ICCT, Meszler Engineering Services, UCS, and
                other stakeholders commented in different respects, with the broad
                themes being: (1) That the change in approach towards HCR engines from
                the Draft TAR and Proposed Determination to the NPRM was not justified,
                was inadequately justified, or was based on justification from the
                industry and not the agencies' own independent judgment; (2) that HCR2
                as defined by EPA does exist and therefore should be used in the
                analysis; and (3) that even if HCR2 technology does not exist exactly
                as EPA defined it, other technologies in the fleet provide the same
                level of efficiency improvement as HCR2 and therefore it should be used
                in the analysis. Many of these commenters stated that if HCR2 had been
                allowed in the compliance analysis, as shown in the NPRM sensitivity
                analysis allowing HCR2 to be applied, compliance costs would have been
                reduced dramatically, ``on par with NHTSA and EPA estimates in the
                TAR.'' 814 815
                ---------------------------------------------------------------------------
                 \814\ NHTSA-2018-0067-11741.
                 \815\ NRDC, Attachement2_CAFE Model Tech Issues.pdf. Docket No.
                NHTSA-2018-0067-11723, at 7-13. ICCT, Full Comments Summary. Docket
                No. NHTSA-2018-0067-117411, at I-2.
                ---------------------------------------------------------------------------
                 Specifically, ICCT, CARB, and UCS took issue with the agencies'
                description of HCR2 technology as speculative, stating that description
                contrasted with how EPA described the technology in prior documents.
                ICCT commented that ``in the Draft TAR and Final Determination, EPA
                observed the real-world advances toward production vehicles using HCR2
                technology, and determined that that technology could be adopted by
                automakers during the compliance period.'' \816\ ICCT stated that in
                the NPRM, ``without rational explanation, the agencies now describe
                this technology as `speculative' and have omitted the technology from
                their primary compliance scenarios altogether.'' CARB similarly
                commented that ``[t]he fact that the Agencies, especially EPA, make [a
                statement that HCR2 is entirely speculative] is genuinely impossible to
                credit.'' \817\ In support, all three commenters referenced EPA's
                hardware testing of a European Mazda engine,\818\ with ICCT stating
                that HCR2 was dismissed as entirely speculative ``despite the careful
                benchmarking of improved HCR engines by EPA,'' while CARB and UCS
                similarly cited this hardware testing to rebut the Alliance's assertion
                that the effectiveness values for HCR2 was ``seriously overestimated.''
                ---------------------------------------------------------------------------
                 \816\ NHTSA-2018-0067-11741.
                 \817\ NHTSA-2018-0067-11873.
                 \818\ Schenk, C. and Dekraker, P., ``Potential Fuel Economy
                Improvements from the Implementation of cEGR and CDA on an Atkinson
                Cycle Engine,'' SAE Technical Paper 2017-01-1016, 2017, doi:10.4271/
                2017-01-1016.
                ---------------------------------------------------------------------------
                 ICCT also took issue with the NPRM statements that ``many engine
                experts questioned [HCR2's] technical feasibility and near-term
                commercial practicability,'' \819\ and that ``[s]takeholders asked for
                the engine to be removed from compliance simulations until the
                performance could be validated with engine hardware,'' with references
                to comments from Fiat-Chrysler (stating ``Remove ATK2 from OMEGA model
                until the performance is validated'' and ``ATK2--High Compression
                engines coupled with Cylinder Deactivation and Cooled EGR are unlikely
                to deliver modeled results, meet customer needs, or be ready for
                commercial application.''),\820\ and comments from the Alliance of
                Automobile Manufacturers, stating that
                [[Page 24410]]
                ``[There] is no current example of combined Atkinson, plus cooled EGR,
                plus cylinder deactivation technology in the present fleet to verify
                EPA's modeled benefits and . . . EPA could not provide physical test
                results replicating its modeled benefits of these combined
                technologies.'' \821\ ICCT stated that the agencies did not identify
                any such comments or evidence from engine experts, or agency analysis
                of them. ICCT stated that ``it is clear that NHTSA is deferring to
                stakeholders, and that EPA has been forced to defer to NHTSA.''
                ---------------------------------------------------------------------------
                 \819\ 83 FR 43038.
                 \820\ Id. (citing NHTSA-2016-0068-0082).
                 \821\ Id. (citing EPA-HQ-OAR-2015-0827-6156).
                ---------------------------------------------------------------------------
                 ICCT also cited interagency review documents where EPA stated
                ``[t]here are Atkinson engine vehicles on the road today (2018 [Toyota]
                Camry and Corolla with cooled EGR and the 2019 Mazda CX5 and Mazda6
                with cylinder deac) that use high geometric compression ratio Atkinson
                cycle technology that is improved from the first generation, MY2012
                vintage ``HCR1'' technology. While it is true that no production
                vehicle has both cooled EGR and cylinder deac, as the EPA ``HCR2''
                engine did, nonetheless, these existing engines demonstrate better
                efficiency than estimated by EPA. Therefore, it would be appropriate to
                continue to use EPA's cooled EGR + deac engine map to represent
                ``HCR2'' engines.'' \822\
                ---------------------------------------------------------------------------
                 \822\ NHTSA-2018-0067-11741, Attachment3_ICCT 15page summary and
                full comments appendix, at I-10 (citing Docket Entry: E.O. 12866
                Review Materials for The Safer Affordable Fuel-Efficient (SAFE)
                Vehicles Rule for Model Years 2021-2026 Passenger Cars and Light
                Trucks NPRM, Docket ID EPA-HQ-OAR-2018-0283-0453 (hereinafter
                ``EO12866 Review Materials''), File:
                ``EO_12866_Review_EPA_comments_on_the_NPRM_sent_to_OMB,_June_29,_2018
                '' at 82, https://www.regulations.gov/document?D=EPA-HQ-OAR-2018-0283-0453).
                ---------------------------------------------------------------------------
                 More specifically regarding the technical specifications of the
                HCR2 engine, ICCT and others stated that EPA had already addressed
                concerns brought by the Alliance \823\ on (1) the base engine fuel
                consumption maps used as the foundation of the HCR2 engine map; \824\
                (2) practical limitations for cEGR to limit engine knock; \825\ (3) the
                reliance on the availability of cylinder deactivation at unrealistic
                speed and load operating points; (4) the impact of 91 RON market and
                certification test fuels; and (5) the ability to implement HCR2
                technology in existing vehicle architectures.\826\
                ---------------------------------------------------------------------------
                 \823\ EPA-HQ-OAR-2015-0827-4089; EPA-HQ-OAR-2015-0827-6156.
                 \824\ NHTSA-2018-0067-11741 (``EPA showed how its ``difference''
                engine maps validly represented performance of the ATK2 [HCR2]
                packages including on different fuels (pp. 301-02); and that the
                difference maps submitted in the industry comment ``provided no
                information to compare vintage or application of the actual engine
                or engines tested, and did not state whether or not testing was
                conducted,'' lacking any information on ``test and/or analytical
                methods, assumptions, fuel properties, environment test conditions,
                how the engine was controlled or how control was modeled, the number
                of data points gathered to generate the AAM `difference map' to
                assure that identical testing and a sufficient fit of data was
                performed'' (p. 301). In addition, EPA showed that concerns about
                knock due to use of cooled exhaust gas recirculation had been
                considered and resolved by ignition improvements (p. 302).'').
                 \825\ NHTSA-2018-0067-12039 (``The agencies appear to have
                relied upon the differences between anti-knock properties of Tier 2
                and Tier 3 fuels, mistakenly focusing solely on octane while
                ignoring ethanol content. . . . this fails to acknowledge the anti-
                knock benefit of charge cooling related to ethanol, which more than
                compensates for the change in octane. HCR2 therefore should not be
                omitted out of concerns around knock.'').
                 \826\ NHTSA-2018-0067-11741. ICCT stated that EPA had previously
                concluded that existing engine architectures were ``well adapted for
                [HCR] technology, and well adapted for the emerging next level HCR2
                package of technologies, since the foundational technologies of
                gasoline direct injection, increased valve phasing authority, higher
                compression ratios, and cooled exhaust gas recirculation are already
                in widespread use.'' ICCT also commented that ``EPA correctly
                observed that there was sufficient lead time to adopt the HCR2
                technology before MY2022 and that it could be incorporated without
                requiring major vehicle redesigns.''
                ---------------------------------------------------------------------------
                 CARB, UCS, and ICCT all stated, in different terms, that even if
                HCR2 technology does not exist exactly as EPA defined it, other
                technologies that exist in the fleet provide the same level of
                efficiency improvement as HCR2, specifically referencing the MY 2018
                Toyota Camry engine and various Mazda engines, and claiming that HCR2
                should therefore be used in the analysis. Specifically, CARB stated
                that these engines ``are already achieving similar efficiency as the
                modeled HCR2 package even though they don't have the full complement of
                technologies (i.e., CEGR and DEAC) used in the HCR2 package.'' \827\
                CARB stated that these engines' ``existence as production engines today
                certainly speaks to the feasibility of this technology for modeling
                that goes out to 2030MY.'' \828\ Similarly, UCS stated that while the
                2018 Toyota Camry engine ``does not have all of the features of the
                HCR2 package constructed by EPA, it achieves similar levels of
                performance, thus rendering the agencies' rationale for excluding HCR2
                moot--this is a production vehicle using Tier 3 fuel which achieves
                performance equivalent to HCR2.'' \829\ Similarly, ICCT cited their own
                analysis of the 2018 Toyota Camry for the propositions that the package
                of technologies on the Camry exceeds the efficiency gains projected by
                EPA's OMEGA model, meaning that EPA's projections for the HCR2 engine
                might understate its effectiveness, and the early problems with low-end
                torque losses associated with Atkinson cycle engines have been
                completely solved.\830\ ICCT stated that ``[t]his evaluation of a real
                world vehicle that comes close to meeting all of the elements of an
                HCR2 engine makes it clear that HCR2 engines are far from a speculative
                technology.''
                ---------------------------------------------------------------------------
                 \827\ NHTSA-2018-0067-11873.
                 \828\ NHTSA-2018-0067-11873.
                 \829\ NHTSA-2018-0067-12039.
                 \830\ NHTSA-2018-0067-11741.
                ---------------------------------------------------------------------------
                 ICCT and CARB also took issue with the agencies' justification for
                not using the HCR2 engine map as a simulation proxy for other new
                engine technology, specifically the statement that:
                 It is important to conduct a thorough evaluation of the actual
                new production engines to measure the brake specific fuel
                consumption and to characterize the improvements attributable to
                friction and thermal efficiency before drawing conclusions. Using
                vehicle level data may misrepresent or conflate complex interactions
                between a high thermal efficiency engine, engine friction reduction,
                accessory load improvements, transmission technologies, mass
                reduction, aerodynamics, rolling resistance, and other vehicle
                technologies.\831\
                ---------------------------------------------------------------------------
                 \831\ 83 FR 43038.
                 Both commenters also took issue with the agencies' statement that
                existing technologies in the NPRM version of the CAFE model could work
                together appropriately to represent an HCR1 engine with additional
                efficiency improvements.\832\
                ---------------------------------------------------------------------------
                 \832\ 83 FR 43038.
                ---------------------------------------------------------------------------
                 ICCT stated that the complexity associated with the package of
                improvements in the Camry engine was common to all of the technology
                packages included in either OMEGA or CAFE modeling, and was neither a
                new issue nor an issue that precludes making reasonable engineering
                judgments. ICCT stated that the agencies projected efficiency estimates
                for other technology packages without engine maps from a production
                engine, citing the agencies' approach to modeling ADEAC technology, and
                concluded that the purpose of full vehicle simulation modeling is to
                project the efficiency impact when several different parts of the
                vehicle are simultaneously upgraded. ICCT stated that ``[i]f reasonable
                estimates could be made for ADEAC without fully validated engine maps,
                there is no reason to exclude other technologies on these grounds,
                especially considering the deep expertise by the agencies and their
                state-of-the-art technology simulation capabilities with the ALPHA
                modeling.'' Similarly, HDS noted that in contrast to the agencies'
                exclusion of HCR2 due to
                [[Page 24411]]
                unresolved issues associated with knock mitigation and cylinder
                deactivation, ``the 2018 analysis included Advanced Cylinder De-
                activation (ADEAC) which has recently come to market readiness.'' \833\
                ---------------------------------------------------------------------------
                 \833\ NHTSA-2018-0067-11985.
                ---------------------------------------------------------------------------
                 Merriam-Webster's dictionary defines speculative as ``involving,
                based on, or constituting intellectual speculation,'' and also,
                ``theoretical rather than demonstrable.'' \834\ To be clear, most
                engines maps used in this analysis--IAV engine maps included--are
                theoretical, although they are built based on benchmarked engine data,
                and additional fuel-economy-improving technologies are added through
                modeling and simulation. But that does not mean that these engines are
                speculative. Although the IAV engine maps are not meant to model any
                manufacturer's particular engine, many, if not all, technology
                combinations have been implemented in real-world engines.
                ---------------------------------------------------------------------------
                 \834\ Definition of ``speculative,'' https://www.merriam-webster.com/dictionary/speculative.
                ---------------------------------------------------------------------------
                 The agencies qualified the HCR2 engine as speculative because ``no
                production engine as outlined in the EPA SAE paper has ever been
                commercially produced or even produced as a prototype in a lab setting.
                Furthermore, the engine map has not been validated with hardware and
                bench data, even on a prototype level (as no such engine exists to test
                to validate the engine map).'' \835\ It is important to distinguish
                theoretical engines maps with technology combinations that have been
                proven through real-world testing and operation, from the HCR2 engine
                map, that was created using a combination of validated individual
                component models, but the resulting engine system model and generated
                engine map were not fully validated against actual hardware.
                ---------------------------------------------------------------------------
                 \835\ 83 FR 43038.
                ---------------------------------------------------------------------------
                 The Alliance and individual automakers have repeatedly provided
                comments on agency actions with their assessment of the feasibility of
                the HCR2 engine, including comments ICCT referenced, stating the EPA
                had addressed concerns brought by the Alliance in the Proposed
                Determination Technical Support Document.\836\ The agencies agree with
                ICCT that EPA provided responses to comments about HCR2 assumptions and
                engine maps in the Technical Support Document, the Proposed
                Determination, and the 2017 Final Determination. However, the agencies
                considered the matter further after receiving extensive comments on
                HCR2 for the NPRM. The agencies have concluded responses did not
                directly and fully address the technical concerns raised by the
                Alliance. Further, new data and information has become available since
                the Proposed and Final Determination that is directly relevant to the
                use of EPA's engine maps in this analysis.
                ---------------------------------------------------------------------------
                 \836\ Also important to note regarding ICCT's comment, the
                Alliance comment cited in the NPRM came from a section of the
                Alliance's comments titled, ``EPA's Response to Alliance Comments
                Regarding Atkinson Cycle Engine Technology Benefits is Inadequate,''
                which seems to suggest that EPA did not address concerns brought by
                the Alliance in the Proposed Determination Technical Support
                Document.
                ---------------------------------------------------------------------------
                 First, it is important to provide background information about
                ICCT's comments referencing previous discussions from the TAR, Proposed
                Determination and Final Determination. For the 2016 Draft TAR, EPA
                initially created the ATK1 and ATK2 engine maps based on the MY 2014
                Mazda 2.0L SKYACTIV-G engine. The EPA benchmarked the Mazda engine,
                then modeled increasing the efficiency of the Mazda engine map by
                simulating the application of additional technologies using GT-Power
                models. The Alliance and FCA commented on the 2016 Draft TAR suggesting
                the EPA's development of the ATK1 and ATK2 engine maps were flawed
                because the maps were developed based on optimistic baseline engine
                characterization of the Mazda engine. The Alliance provided evidence of
                the flaws in EPA's characterization by comparing EPA's published base
                engine data, developed using Tier 2 certification gasoline, to engine
                data benchmarked by USCAR. USCAR benchmarked their own Mazda Skyactiv
                engine map using a 91 RON fuel. The comparison resulted in the creation
                of a ``difference map'' that showed where the two data sets diverged.
                The ``difference map'' implied there were areas of significant
                divergence, calling into question the data upon which the ATK1 and ATK2
                models are based. The EPA responded stating ``[the Alliance] did not
                provide data or other information to substantiate its claim that EPA's
                engine dynamometer fuel consumption measurements using a MY2014 Mazda
                OEM production 2.0L SKYACTIV-G, upon which the ATK2 packages from the
                TAR analysis are based, were in any way unrepresentative of this
                engine's actual performance.'' \837\ ICCT cited in their NPRM comments
                that the EPA's discussion of these ``difference maps'' supported their
                statement that ``[i]n fact, in the Technical Support Document for EPA's
                Proposed and 2017 Final Determination, EPA addressed all these concerns
                brought forth by the Alliance [regarding HCR2] (including the costs and
                effectiveness impacts of using regular octane fuel instead of premium
                fuel).''
                ---------------------------------------------------------------------------
                 \837\ EPA PD TSD at 2-299.
                ---------------------------------------------------------------------------
                 It is understandable why ICCT may have thought this discussion
                addressed concerns raised about the HCR2 map; however, review of the
                Alliance's original Draft TAR comments makes it clear the Alliance's
                initial comments addressed the benchmarking of the MY 2014 Mazda 13:1
                SKYACTIV-G engine itself. The Alliance's original comments, expressed
                concern over the modeled effectiveness of the advanced Atkinson
                technology packages because of the baseline engine data used. The
                Alliance suggested the effectiveness is likely overestimated due to
                multiple flaws in the benchmarking and modeling approaches taken by
                EPA. Only the benchmarking is addressed by EPA's response to the
                ``difference maps,'' not the concerns about modeling approach.
                 The Alliance's concerns about modeling included the accuracy of the
                base engine fuel consumption maps (to the extent the baseline engine
                maps were overly optimistic, the modeled ATK maps were optimistic),
                limitations for cEGR to mitigate engine knock, limitations of cylinder
                deactivation, and the impact of fuels.\838\ After further review, the
                agencies determined the Alliance's concerns were not fully addressed,
                resulting in a closer review of the ATK model development process.
                ---------------------------------------------------------------------------
                 \838\ EPA-HQ-OAR-2015-0827-4089.
                ---------------------------------------------------------------------------
                 Review of the engine model development showed the engine map was
                generated assuming the use of high octane fuel, and the follow-up
                engine dynamometer validation testing also used high octane fuel.\839\
                The characterization of the baseline Mazda Skyactiv engine showed 1-3
                percent increase in thermal efficiency across a large portion of the
                engine map when operated on Tier 2 fuel versus lower octane
                fuel.840 841 The increase in engine
                [[Page 24412]]
                thermal efficiency, caused by the higher octane fuel, is anticipated to
                be amplified when applying ATK technologies. ATK technologies increase
                efficiency by increasing the pressure in cylinder during combustion;
                however, at the same time the increased pressure increases risk of
                knock. For more discussion on engine knock, see Section VI.C.1.a).
                Ultimately, it is expected that the ATK1 and ATK2 engines would show a
                larger improvement in thermal efficiency as a result of being developed
                assuming a high-octane fuel versus the 1-3 percent improvement observed
                on the baseline Mazda Skyactiv engine.
                ---------------------------------------------------------------------------
                 \839\ Ellies, B., Schenk, C., and Dekraker, P., ``Benchmarking
                and Hardware-in-the-Loop Operation of a 2014 MAZDA SkyActiv 2.0L
                13:1 Compression Ratio Engine,'' SAE Technical Paper 2016-01-1007,
                2016, doi:10.4271/2016-01-1007.
                 \840\ The engine was first run on LEVIII-compliant certification
                fuel which has a 7 psi vapor pressure and 88aki. This fuel is
                similar to Tier 3 fuel with exception of the vapor pressure which is
                required to be 9 psi to meet Tier 3 certification. It was then
                tested on Tier 2 certification fuel (93aki) to assess effects of
                higher octane fuel on engine operation and efficiency.
                 \841\ Ellies, B., Schenk, C., and Dekraker, P., ``Benchmarking
                and Hardware-in-the-Loop Operation of a 2014 MAZDA SkyActiv 2.0L
                13:1 Compression Ratio Engine,'' SAE Technical Paper 2016-01-1007,
                2016, doi:10.4271/2016-01-1007.
                 Schenk, C. and Dekraker, P., ``Potential Fuel Economy
                Improvements from the Implementation of cEGR and CDA on an Atkinson
                Cycle Engine,'' SAE Technical Paper 2017-01-1016, 2017, doi:10.4271/
                2017-01-1016.
                ---------------------------------------------------------------------------
                 A further limitation was revealed during the agencies review of the
                ATK model development. The limitation was in how COV of IMEP, an
                important indicator of combustion stability, was not accounted for
                directly in the model. The 0-D/1-D models used for investigating cEGR
                effectiveness could not adequately simulate changes to COV of IMEP. To
                compensate for the lack of an appropriate model, limits on cEGR were
                based on literature values for unrelated engine technologies.\842\ As a
                result, there was no direct evaluation of combustion stability while
                evaluating the feasibility of the engine concept.
                ---------------------------------------------------------------------------
                 \842\ Schenk, C. and Dekraker, P., ``Potential Fuel Economy
                Improvements from the Implementation of cEGR and CDA on an Atkinson
                Cycle Engine,'' SAE Technical Paper 2017-01-1016, 2017, doi:10.4271/
                2017-01-1016.
                ---------------------------------------------------------------------------
                 In contrast, for the NPRM and final rule analysis, IAV engines were
                optimized using Tier 3 fuel, to balance performance and fuel
                consumption. The majority of baseline vehicles are specified to operate
                on 87 AKI fuel, therefore lower octane fuel was used to maintain
                baseline functionality. The IAV engine maps were all derived from a
                consistent baseline engine and were also optimized using a validated
                kinetic knock model, and using a COV of IMEP threshold of 3 percent.
                 These differences in model construction caused an inconsistency
                that resulted in unrealistic improvements in fuel economy and
                CO2 emissions for the HCR engine technologies, whereas the
                IAV engine maps reflect more realistic accounting for the improvements.
                The use of high octane fuel and lack of combustion stability modeling
                are complimentary issues that have compounded effects when combined.
                For example, the use of high octane fuel allows more advanced spark
                timing which both increases efficiency and improves combustion
                stability, allowing higher cEGR rates before reaching acceptable limits
                for drivability. The compound effect is greater than the simply adding
                together individual effects, causing a potentially further unrealistic
                increase in effectiveness. At a minimum, it is uncertain how using Tier
                3 fuel in the HCR2 engine would impact the BSFC of the engine, as there
                was no direct evaluation of the feasibility of the engine concept's
                ability to operate on regular octane fuel. The cost for the
                effectiveness of the HCR2 technology also is inconsistent with the cost
                of the effectiveness improvement values for the technologies in the
                2015 NAS report.\843\ In considering all of this information, the
                agencies, believe the HCR2 engine map overstates the capabilities of
                the technology and decided not to use that engine map for the final
                rule analysis.
                ---------------------------------------------------------------------------
                 \843\ 2015 NAS at p. 90 and 91.
                ---------------------------------------------------------------------------
                 However, the agencies believe the HCR1 engine map does reflect
                improvements that are representative of the technology in the
                rulemaking timeframe. For the final rule, to reflect better the
                incremental effectiveness for a low-cost version of HCR technology, the
                agencies added the HCR0 engine for the analysis. The specification of
                this engine was provided in the NPRM PRIA as Eng22b. Using this engine
                improves the estimated incremental effectiveness because the
                incremental engine changes from were directly specified for the
                modeling. HCR0 is the first engine in the HCR path that a manufacturer
                could adopt. Accordingly, the non-HEV Atkinson engine maps used for the
                NPRM and final rule central analysis fit into the three defined
                categories as follows: (1) Eng26 is an HEV Atkinson Cycle engine; (2)
                in the NPRM analysis, Atkinson-mode engines were characterized by Eng24
                (HCR1), and for the final rule analysis, Atkinson-mode engines are
                characterized by Eng22b (HCR0) and Eng24 (HCR1); and (3) Atkinson-
                enabled engines are characterized by the different VVT engine
                technologies identified earlier in basic engine discussions and shown
                on Table VI-41 and Table VI-42.
                 Regarding the ability of manufacturers to adapt the engine
                architecture to practical use, the agencies see merit in observations
                from both manufacturers and other groups. ICCT is correct in their
                observation that some production engines have integrated combinations
                of the technologies, including SGDI, VVT and cEGR. Furthermore, the
                agencies agree with ICCT that an engine could be built integrating all
                the technologies represented in the HCR2 engine model. However, the
                agencies also agree with the Alliance's comments to the 2016 Draft TAR
                that applying all the technologies to an engine that only has some of
                the technologies would require a significant redesign of the powertrain
                package. The redesign would need to accommodate the new hardware
                integration, controls and emissions calibration, OBD development and
                other major efforts. As discussed further in Section VI.C.1.e), the
                agencies believe these considerations impact how quickly and widely the
                technology could be implemented in the rulemaking timeframe.
                 The agencies also disagree with commenters that the HCR2 engine map
                should be used as a proxy for other vehicles in the fleet that achieve
                high thermal efficiency. None of the existing vehicles that commenters
                cited, like the 2019 Toyota Camry and Corolla with cEGR or the 2019
                Mazda CX5 and Mazda 6 with cylinder deactivation, include the same
                combination of technologies as the HCR2 engine. Unlike other engine
                technologies in the NPRM and the final rule analysis, no engines in the
                market or in prototype stages exist that have the combined technology
                specifications of the HCR2. Accordingly, there is no production vehicle
                that demonstrates the combination of technologies as applied in the
                HCR2 engine that (1) is feasible, and (2) can achieve the same
                effectiveness as the modeled HCR2 engine. The NPRM highlighted concerns
                about using the HCR2 engine map as a proxy for new engine technologies
                that achieve high thermal efficiency, specifically that:
                 It is important to conduct a thorough evaluation of the actual
                new production engines to measure the brake specific fuel
                consumption and to characterize the improvements attributable to
                friction and thermal efficiency before drawing conclusions. Using
                vehicle level data may misrepresent or conflate complex interactions
                between a high thermal efficiency engine, engine friction reduction,
                accessory load improvements, transmission technologies, mass
                reduction, aerodynamics, rolling resistance, and other vehicle
                technologies.\844\
                ---------------------------------------------------------------------------
                 \844\ 83 FR 43038.
                 The agencies continue to believe this is true, and Toyota's
                comments that the Camry improvements were due to more than just the
                engine improvements, as discussed further below, provide further
                support to this conclusion.
                 Several commenters cited EPA's SAE paper discussing the use of the
                HCR2 engine model and comparing it to the benchmarking of a 2018 Toyota
                Camry
                [[Page 24413]]
                2.5L engine.845 846 The commenters cited the HCR2 engine's
                similarities to the Toyota Camry engine as a reason to employ the
                technology model broadly across the entire vehicle fleet, including
                applying it to pickup trucks such as the Toyota Tacoma. In the paper,
                EPA benchmarked a 2018 Toyota Camry 2.5L Atkinson cycle engine equipped
                with cEGR. EPA created a full vehicle model (the exemplar vehicle)
                based on the benchmarked data for use in the ALPHA modeling tool. The
                full vehicle simulation was used to compare the HCR2 engine to the
                Camry's 2.5L engine, and showed some similarities. The paper implied
                that it is possible to adopt more technologies to the MY 2018 Camry,
                like cylinder deactivation, to meet future standards.
                ---------------------------------------------------------------------------
                 \845\ Kargul, J., Stuhldreher, M., Barba, D., Schenk, C. et al.,
                ``Benchmarking a 2018 Toyota Camry 2.5-Liter Atkinson Cycle Engine
                with Cooled-EGR,'' SAE Int. J. Adv. & Curr. Prac. in Mobility
                1(2):601-638, 2019, https://doi.org/10.4271/2019-01-0249.
                 \846\ Duleep, K.G., ``Review of the Technology Costs and
                Effectiveness Utilizing in the Proposed SAFE Rule,'' Final Report,
                H-D Systems, October 2018, at p. 37.
                ---------------------------------------------------------------------------
                 This paper, and the comments relying on it--specifically that it
                shows that additional technologies can be added to the MY 2018 Camry
                engine to meet future standards--were the subject of considerable
                debate in the rulemaking docket. Toyota provided supplemental comments
                regarding issues Toyota had with the modeling and simulation. These
                included a detailed discussion on why HCR2 is not a reasonable model of
                the 2018 Toyota Camry engine. Toyota identified other technologies that
                contributed to the overall thermal efficiency of the 2018 Camry
                compared to previous generation.\847\ Toyota stated that the 2018
                Toyota Camry employed numerous technologies like SGDI, cEGR, optimized
                intake system, optimized exhaust system, optimized piston design,
                laser-cladded valve seats, VVT, engine friction reduction, variable oil
                pump, and electric coolant pump, that all contributed to the engine's
                improved efficiency over the previous version.\848\
                ---------------------------------------------------------------------------
                 \847\ NHTSA-2018-0067-12431. Supplemental Comments of Toyota
                Motor North America, Inc. (7/15/19) at 1-2; NHTSA-2018-0067-12376.
                Supplemental Comments of Toyota Motor North America, Inc. (3/25/19)
                at 1.
                 \848\ Hakariya, M., Toda, T., and Sakai, M., ``The New Toyota
                Inline 4-Cylinder 2.5L Gasoline Engine,'' SAE Technical Paper 2017-
                01-1021, 2017, available at https://doi.org/10.4271/2017-01-1021.
                ---------------------------------------------------------------------------
                 In addition, Toyota stated:
                [T]he 2018 Exemplar Vehicle that is based on the baseline 2018
                Toyota Camry was equipped with engine start stop that doesn't exist
                on the production vehicle. Cylinder deactivation was added to the
                2025 exemplar vehicle as a protentional enhancement. We acknowledged
                that adding cylinder deactivation to the Atkinson-cycle engines is
                technically possible and would provide some fuel economy benefits.
                However, the primary function of cylinder deactivation is to reduce
                engine pumping losses which the Atkinson cycle and EGR already
                accomplish. The diminishing return on the cylinder deactivation,
                Atkinson cycle and EGR are further exaggerated by smaller 4-cylinder
                engines.
                 This assessment aligns with the 2015 NAS committee report that
                estimated a 0.7 percent fuel consumption improvement for adoption of
                cylinder deactivation for DOHC and SOHC V6 and V8 engines.\849\ The
                agencies agree with Toyota and the NAS assessment that applying
                cylinder deactivation in small cylinder count engines is subject to
                diminishing returns.
                ---------------------------------------------------------------------------
                 \849\ 2015 NAS at p. 34.
                ---------------------------------------------------------------------------
                 The agencies agree with Toyota that the presence of the advanced
                technologies, in addition to the HCR technology, contributed to the
                performance of the Camry. The analysis already provides benefits for
                the other advanced technologies individually, and risks, if not
                ensures, double counting these benefits if the HCR2 model is used (as
                discussed above and in VI.B). Likely double counting of technology
                effectiveness further supported the agencies' choice not to use the
                HCR2 model for the final rule analysis.
                 The agencies disagree that the approach taken to modeling ADEAC
                technology should similarly apply to modeling the HCR2 engine, or that
                because ADEAC just recently entered the market and was employed in the
                modeling, HCR2 should be as well. As discussed further below, the
                effectiveness estimates for ADEAC were based on extensive discussions
                with suppliers and manufacturers that provided CBI data, and technical
                publications.\850\ The effectiveness estimates provided for ADEAC
                represented the effects of applying a single technology, and not a
                combined estimate for several technologies applied at once. Moreover,
                as commenters noted, ADEAC had recently ``come to market readiness,''
                \851\ compared to the HCR2 technology which cannot be found, as
                modeled, in the market, or even in prototype form. As discussed
                throughout this document, the preferred approach for the NPRM and final
                rule was to isolate the effectiveness improvement attributable to
                specific technologies and apply those through full vehicle simulations
                to capture technology synergies and dis-synergies appropriately.
                ---------------------------------------------------------------------------
                 \850\ Eisazadeh-Far, K. and Younkins, M., ``Fuel Economy Gains
                through Dynamic-Skip-Fire in Spark Ignition Engines,'' SAE Technical
                Paper 2016-01-0672, 2016, doi:10.4271/2016-01-0672.
                 \851\ NHTSA-2018-0067-11985.
                ---------------------------------------------------------------------------
                 The agencies also disagree with ICCT's comment that the agencies
                were simply deferring to stakeholders, or that EPA was simply deferring
                to NHTSA regarding the feasibility of the HCR2 engine. It is reasonable
                to assume that the automobile manufacturers that belong to the Alliance
                employ some engine experts that are qualified to speak on the
                feasibility of an engine. Not just one or two manufacturers objected to
                the HCR2 engine; the Alliance commented on behalf of its members in
                support of the exclusion of the engine from the analysis,\852\ and this
                exclusion was further supported by comments from individual automakers
                as well. Toyota, the automaker cited by several commenters as closest
                to implementing HCR2 technology stated in supplemental comments that
                (1) the HCR2 is not representative of its engine technology; \853\ and
                (2) Toyota believes there are diminishing returns for implementing the
                HCR2 technologies.\854\ The agencies received no comments from
                stakeholders that manufacture engines in support of the HCR2
                technology's feasibility and potential future adoption.
                ---------------------------------------------------------------------------
                 \852\ NHTSA-2018-0067-12073, at 139.
                 \853\ Comment from Toyota NHTSA-2018-0067-12376 (``While the
                agencies' definitions for the different levels of Atkinson
                technology seem to have evolved, the 2018 Camry is clearly not
                equipped with HCR2 technology.'').
                 \854\ Comment from Toyota NHTSA-2018-0067-12376 (``advanced
                cylinder deactivation has not yet been established when packaged
                with an Atkinson-cycle engine. Both technologies play similar roles
                in reducing engine pumping losses which can led to diminishing
                returns when combined.'').
                ---------------------------------------------------------------------------
                 For HCR technology, the agencies carefully considered comments to
                the NPRM and the available data, and concluded it is appropriate to
                include HCR0 and HCR1 engine models for the final rule analysis. The
                engine maps for those technologies provide the best estimates for the
                effectiveness of HCR technology relative to the engine maps for the
                other engine technologies used for the analysis. The agencies have
                reconsidered issues associated with the HCR2 engine models and maps.
                The agencies find that significant technical questions and issues
                remain and the engine maps very likely overstate the feasible amount of
                effectiveness that could be achieved by the represented technologies.
                Therefore, HCR2 technology is not included for the final rule analysis.
                (4) HEV Atkinson Cycle Engines
                 Three types of Atkinson technology were discussed in the previous
                section.
                [[Page 24414]]
                HEV Atkinson cycle engines fall in the first category, operating solely
                or primarily in Atkinson mode, supported by an electric drive.
                 Engine map 26 (Eng26) is the model of the HEV/PHEV Atkinson cycle
                engine used for the NPRM and final rule analysis. The engine was based
                on Argonne's Advanced Mobility Technology Laboratory (AMTL) 2010 Toyota
                Prius test data and published literature.\855\ Argonne's AMTL is
                continuously involved in research and testing of advanced technologies,
                especially in areas of electrification, and has a large existing
                database of test data from advanced technology vehicles.\856\ As a
                result of Argonne's continued research, a 2017 Toyota Prius was
                characterized for an independent project. Argonne updated the HEV
                Atkinson cycle engine using the new Prius data to reflect the 41
                percent thermal efficiency of the new 2017 system.\857\ The
                electrification technology groups that used Eng26 include powersplit
                hybrid vehicles (SHEVPS) and plug-in powersplit hybrid vehicles
                (PHEV20/50).
                ---------------------------------------------------------------------------
                 \855\ ``2010 Toyota Prius.'' http://www.anl.gov/energy-systems/group/downloadable-dynamometer-database/hybrid-electric-vehicles/2010-toyota-prius. Last accessed April, 2018.
                 \856\ ANL AMTL Downloadable Dynamometer Database (D3). https://www.anl.gov/es/downloadable-dynamometer-database. Last accessed Dec.
                05, 2019.
                 \857\ Carney, D. ``Toyota unveils more new gasoline ICEs with
                40% thermal efficiency.'' SAE. April 4, 2018. https://www.sae.org/news/2018/04/toyota-unveils-more-new-gasoline-ices-with-40-thermal-efficiency. Last accessed Dec. 5, 2019.
                ---------------------------------------------------------------------------
                (5) Advanced Cylinder Deactivation Technologies
                 Advanced cylinder deactivation (ADEAC) systems, also known as
                rolling or dynamic cylinder deactivation systems, allow a further
                degree of cylinder deactivation than the base DEAC. ADEAC allows the
                engine to vary the percentage of cylinders deactivated and the sequence
                in which cylinders are deactivated, essentially providing
                ``displacement on demand'' for low load operations.
                 ADEAC systems may be integrated into the valvetrains with moderate
                modifications on OHV engines. However, while the ADEAC operating
                concept remains the same on DOHC engines, the valvetrain hardware
                configuration is very different, and application on DOHC engines is
                projected to be more costly per cylinder due to the valvetrain
                differences.
                 The agencies discussed assumptions and effectiveness for the ADEAC
                package in the NPRM preamble.\858\ The initial review of this
                technology was based on a technical publication that used a MY 2010
                engine design that had incorporated a SOHC VVT basic engine.\859\ Other
                preproduction 8-cylinder OHV prototype vehicles with ADEAC were briefly
                evaluated for this analysis, but no production versions of the
                technology have been studied.\860\ For ADEAC fuel consumption
                effectiveness values, no engine map was available at the time of the
                NPRM analysis. Accordingly, the agencies took the effectiveness values
                as predicted by full vehicle simulations of a DEAC engine with SGDI,
                VVL, and VVT, and added 3 percent or 6 percent respectively for I-4
                engines and V-6 or V-8 engines, and cross-referenced CBI data to
                quality check this approach.
                ---------------------------------------------------------------------------
                 \858\ 83 FR 43038-39.
                 \859\ Wilcutts, M., Switkes, J., Shost, M., and Tripathi, A.,
                ``Design and Benefits of Dynamic Skip Fire Strategies for Cylinder
                Deactivated Engines,'' SAE Int. J. Engines 6(1):278-288, 2013,
                available at https://doi.org/10.4271/2013-01-0359. Eisazadeh-Far, K.
                and Younkins, M., ``Fuel Economy Gains through Dynamic-Skip-Fire in
                Spark Ignition Engines,'' SAE Technical Paper 2016-01-0672, 2016,
                available at https://doi.org/10.4271/2016-01-0672.
                 \860\ EPA, 2018. ``Benchmarking and Characterization of a Full
                Continuous Cylinder Deactivation System.'' Presented at the SAE
                World Congress, April 10-12, 2018. Retrieved from https://www.regulations.gov/document?D=EPA-HQOAR-2018-0283-0029.
                ---------------------------------------------------------------------------
                 The agencies noted two potential approaches to including advanced
                cylinder deactivation in the full-scale Argonne simulation modeling
                analysis for the final rule. First, the agencies proposed using IAV
                Eng25a, which was developed to capture the maximum benefits of advanced
                cylinder deactivation with several constraints that could include
                emissions, cold start, NVH, and durability. Second, the agencies
                proposed using a technique developed by Argonne in coordination with
                NHTSA to split the overall engine data into individual cylinder data
                and compute overall torque and the fuel consumption rate by accounting
                for whether each cylinder is active or inactive. The agencies sought
                comment on using either approach in the final rule analysis to capture
                best the benefits of advanced cylinder deactivation.
                 CARB, ICCT, Meszler Engineering Services, HDS, and UCS provided a
                mixed set of comments on numerous aspects of ADEAC in the NPRM
                analysis.\861\ Broadly, HDS commented on a need to describe ADEAC
                technology better: ``The 2018 analysis also utilized Advanced Cylinder
                Deactivation in its analysis but the package components were not
                completely explained in the PRIA.'' \862\ Other stakeholders provided
                comments on ADEAC adoption features, effectiveness, and cost, which are
                discussed below.
                ---------------------------------------------------------------------------
                 \861\ ICCT Docket # NHTSA-2018-0067-11741 at I-12, Duleep Docket
                # NHTSA-2018-0067-11873 at 108, Meszler Docket # NHTSA-2018-0067-
                11723 at p. 26.
                 \862\ Duleep, K.G., ``Review of the Technology Costs and
                Effectiveness Utilizing in the Proposed SAFE Rule,'' Final Report,
                H-D Systems, October 2018, at p. 17.
                ---------------------------------------------------------------------------
                 The agencies discussed assumptions and effectiveness for the ADEAC
                package in the NPRM preamble.\863\ The initial review of this
                technology was based on a technical publication that used a MY 2010
                engine design incorporating SOHC and VVT.\864\ After determining the
                MY2010 engine design was not representative of the analysis fleet, the
                agencies used effectiveness values based on CBI data. The MY2017
                baseline fleet reflects technology updates such as SGDI and DEAC that
                could adopt ADEAC incrementally in the final rule analysis. The cost
                and effectiveness for ADEAC reflects the baseline engine. The 2015 NAS
                Committee estimated an 0.7 percent fuel consumption improvement for
                adoption of cylinder deactivation for V6s and V8s
                engines.865 866
                ---------------------------------------------------------------------------
                 \863\ 83 FR 43038-39.
                 \864\ Wilcutts, M., Switkes, J., Shost, M., and Tripathi, A.,
                ``Design and Benefits of Dynamic Skip Fire Strategies for Cylinder
                Deactivated Engines,'' SAE Int. J. Engines 6(1):278-288, 2013,
                available at https://doi.org/10.4271/2013-01-0359. Eisazadeh-Far, K.
                and Younkins, M., ``Fuel Economy Gains through Dynamic-Skip-Fire in
                Spark Ignition Engines,'' SAE Technical Paper 2016-01-0672, 2016,
                available at https://doi.org/10.4271/2016-01-0672.
                 \865\ Applied after VVT and VVL.
                 \866\ Applied before VVT and VVL.
                ---------------------------------------------------------------------------
                 The agencies requested comments on alternative methods to estimate
                ADEAC effectiveness but received no comments regarding either approach
                mentioned in the NPRM. For the final rule analysis, the agencies used
                effectiveness values as predicted by full vehicle simulations of a DEAC
                engine with SGDI, VVL, and VVT, and added 3 percent or 6 percent
                respectively for I-4 engines and V-6 or V-8 engines for the naturally
                aspirated engines. Effectiveness for turbocharged engines used 1.5
                percent and 3 percent values, as predicted by full vehicle simulation
                of a TURBOD engine for I4 and V6/V8, respectively. Without sufficient
                data to simulate ADEAC, both the IAV and Argonne methodologies
                described in the NPRM provided questionable estimates for ADEAC. These
                errors would have propagated across other technology combinations in
                the analysis. The estimates used for ADEAC and TURBOD for the final
                rule analysis are also in line with EPA
                [[Page 24415]]
                estimates discussed in their SAE technical publications.\867\
                ---------------------------------------------------------------------------
                 \867\ Kargul, J., Stuhldreher, M., Barba, D., Schenk, C. et al.,
                ``Benchmarking a 2018 Toyota Camry 2.5-Liter Atkinson Cycle Engine
                with Cooled-EGR,'' SAE Int. J. Adv. & Curr. Prac. in Mobility
                1(2):601-638, 2019, https://doi.org/10.4271/2019-01-0249 at pp. 19-
                21.
                ---------------------------------------------------------------------------
                 For the final rule analysis, the agencies used the same
                effectiveness values for ADEAC applied to naturally aspirated engines
                as in the NPRM, and incorporated estimated effectiveness values for
                TURBOAD to represent ADEAC on downsized turbocharged engines.
                (6) Miller Cycle Engines
                 In the proposed rule, the agencies provided two engine maps
                representative of Miller cycle and Eboost engines with 48V battery
                systems. The Miller cycle engine (Eng23b) and Miller cycle engine with
                Eboost (Eng23c) specifications were provided in the PRIA but were not
                used in the NPRM analysis,\868\ although the agencies sought comment on
                the specifications used for the modeling.
                ---------------------------------------------------------------------------
                 \868\ NPRM PRIA at p. 307-09.
                ---------------------------------------------------------------------------
                 Roush on behalf of CARB, ICCT, Meszler Engineering on behalf NRDC,
                HDS, and UCS, commented that the agencies did not consider the
                combination of turbocharging and Miller cycle.\869\ Specifically, Roush
                argued that the agencies' omission of an engine that utilizes a
                combination of turbocharging and Miller cycle was unreasonable because
                it is already in production, specifically on the VW 2.0L EA888 Gen3B--
                DI. Roush stated this omission would limit the effectiveness for
                turbocharged engines and cause the adoption of more expensive
                solutions, thereby overstating the cost to achieve target fuel economy
                levels. Similarly, Roush pointed to the omission of an engine that uses
                a variable geometry turbocharger as an error in the agencies' vehicle
                modeling; Roush pointed to VW's EA211 TSI Evo engine available in
                Europe in 2017 as an example of an engine in production that enables
                cost-effective Miller cycle applications.
                ---------------------------------------------------------------------------
                 \869\ NHTSA-2018-0067-11985. HD systems at p, 34; ICCT at p.
                102; NRDC Attachment 2 at p.16.
                ---------------------------------------------------------------------------
                 In response to these comments, the agencies added and used both
                Miller cycle-type engines and Miller cycle engines with electric assist
                for the final rule analysis. Discussed earlier in this section, the
                agencies developed engine maps for additional combinations of
                technologies for the final rule, including engine maps that became
                available after the NPRM analysis was completed but before the NPRM was
                published. For the final rule analysis, the agencies have included a
                Miller cycle engine, Eng23b (VTG), as another available engine
                technology. The specification of this engine was discussed in PRIA
                Chapter 6.3.2.2.20.20.2.2 and the costs are based on the 2015 NAS
                estimates for this technology.
                (7) Variable Compression Ratio Engines
                 Variable compression ratio (VCR) engines work by changing the
                length of the piston stroke of the engine to operate at a more optimal
                compression ratio and improve thermal efficiency over the full range of
                engine operating conditions. Engines using VCR technology are currently
                in production, but appear to be targeted primarily towards limited
                production, high performance and very high BMEP (27-30 bar)
                applications.
                 A few manufacturers and suppliers provided information about VCR
                technologies, and several design concepts were reviewed that could
                achieve a similar functional outcome. In addition to design concept
                differences, intellectual property ownership complicates the ability of
                the agencies to define a VCR hardware system that could be widely
                adopted across the industry.
                 For the NPRM analysis, the agencies provided specifications of a
                VCR engine (Eng26a) in the PRIA for review and comment.\870\ However
                the VCR engine was not used in the NPRM analysis.
                ---------------------------------------------------------------------------
                 \870\ NPRM PRIA at pp. 304-06.
                ---------------------------------------------------------------------------
                 The Alliance commented in support of the exclusion of variable
                compression ratio engines from the analysis, stating that the
                technology is still in early development, and too speculative to be
                included at this time. The Alliance also stated that the technology is
                unlikely to attain significant penetration in the MY 2026 timeframe due
                to intellectual property protection associated with early
                implementations and its likely application primarily to high-
                performance vehicles. The Alliance also cited the technology's price as
                a potential barrier to adoption.\871\ Similarly, Ford commented that:
                ---------------------------------------------------------------------------
                 \871\ NHTSA-2018-0067-12073 (``At least one source also
                indicates a steep price to this technology--``at least $3,000 more
                to produce than a standard 16-valve double-overhead-camshaft four-
                cylinder.'').
                 [VCR technology] is likely to be adopted only for premium/
                limited-market vehicles in the near future. We also agree that
                intellectual property protections on early implementations will
                further inhibit significant fleet penetration. Incorporation of VCR
                requires a new or highly modified engine architecture, necessitating
                major investment from both the engineering and manufacturing
                standpoints. Sharing/commonality across engine families would be
                greatly limited.'' 872 873
                ---------------------------------------------------------------------------
                 \872\ NHTSA-2018-0067-11928.
                 \873\ NHTSA-2018-0067-11928 at p. 9.
                 Similarly, other automakers commented on a confidential basis that
                several main hurdles prevented them from employing VCR engines,
                including the complexity of VCR engines and the associated cost of
                those complex parts.
                 UCS commented that the agencies did not consider VCR engine
                technologies in the NPRM analysis.\874\ They stated that the technology
                was not modeled, nor was it incorporated into the CAFE model. UCS
                argued that Nissan's VC-Turbo engine is part of a strategy to improve
                fuel efficiency for Nissan's luxury vehicles by 30-35 percent over
                previous models, which would be enough to exceed the vehicle's
                regulatory targets without any credits. UCS concluded that given VCR
                technology is being put into production in a high-volume vehicle, there
                is no reason for the agencies to exclude its adoption.
                ---------------------------------------------------------------------------
                 \874\ NHTSA-2018-0067-12039 at p. 6.
                ---------------------------------------------------------------------------
                 The agencies agreed with comments to include VCR engine
                technologies in the final rule analysis and on further technical
                consideration, the agencies have added a VCR engine to the engine
                technologies list manufacturers could adopt. However, the agencies
                limited the adoption of the VCR engine technology to Nissan only. VCR
                engines are complex, costly by design, and synergetic with mainstream
                technologies like downsize turbocharging, making it unlikely that a
                manufacturer that has already started down an incongruent technology
                path would adopt VCR technology.
                (8) Diesel Engines
                 Diesel engines have several characteristics that result in superior
                fuel efficiency over traditional gasoline engines, including reduced
                pumping losses due to lack of (or greatly reduced) throttling, high
                pressure direct injection of fuel, a combustion cycle that operates at
                a higher compression ratio, and a very lean air/fuel mixture relative
                to an equivalent-performance gasoline engine.\875\ However, diesel
                technologies requires additional enablers, such as a NOX
                adsorption catalyst system or a urea/ammonia selective catalytic
                reduction system, for control of NOX emissions.
                ---------------------------------------------------------------------------
                 \875\ Diesel cycle is also a four-stroke cycle like the Otto
                Cycle, except in the Intake stroke no fuel is injected and fuel is
                injected late in the compression stroke at higher pressure and
                temperature.
                ---------------------------------------------------------------------------
                 For the NPRM, the agencies modeled one diesel engine, represented
                by
                [[Page 24416]]
                Eng17,\876\ which was termed ``ADSL'' in the CAFE modeling. DSLI, a
                more advanced diesel engine, was modeled using a 4.5 percent fixed
                effectiveness improvements over ADSL.
                ---------------------------------------------------------------------------
                 \876\ Docket ID NHTSA-2018-0067-1972. NPRM PRIA at p. 295.
                ---------------------------------------------------------------------------
                 CARB commented that diesel technologies are essentially locked out
                of being selected in the CAFE model because of the high cost.\877\ They
                state that diesel technology is only selected in rare instances.
                ---------------------------------------------------------------------------
                 \877\ Docket ID NHTSA-2018-0067-11873. CARB at 108.
                ---------------------------------------------------------------------------
                 The agencies agree that diesel technology is rarely selected. The
                technologies required to meet diesel emissions standards are costlier
                compared to gasoline technologies, particularly in the rulemaking
                timeframe. For example, the 2015 NAS report determined that in the
                current market, ``vehicles with diesel engines are priced an average of
                more than $4,000 more than comparably equipped gasoline vehicles.''
                \878\ Furthermore, the NAS report stated that the ``Carbon Penalty''
                makes it harder for manufactures to meet CO2 standards
                because of the higher carbon density in the diesel fuel compared to
                gasoline that results in higher CO2 per gallon.\879\ In
                addition, the market for diesel vehicles has stagnated at around 1
                percent for many years after it peaked at 5.9 percent in 1981,
                according to the EPA Trends Report.\880\ The agencies believe that the
                modeled cost of diesel engines appropriately prevents their widespread
                adoption in the analysis.
                ---------------------------------------------------------------------------
                 \878\ 2015 NAS at 123-24.
                 \879\ 2015 NAS Findings 3.3 and 3.4 at p. 120.
                 \880\ EPA, ``The 2018 EPA Automotive Trends Report.'' March
                2019. EPA-420-R-19-002. https://nepis.epa.gov/Exe/ZyPDF.cgi/P100W5C2.PDF?Dockey=P100W5C2.PDF at pp. 5 & 6. Last accessed
                December 16, 2019.
                ---------------------------------------------------------------------------
                 UCS commented that the agencies restricted cylinder deactivation
                technologies to only naturally aspirated gasoline engines.\881\ In
                response to this and other comments, the agencies have allowed diesel
                engines to adopt ADEAC for this final rule analysis. These engines were
                designated as DSLIAD to represent diesel engines with ADEAC, and were
                modeled using a 7.5 percent fixed effectiveness improvement on top of
                DSLI. This effectiveness improvement of ADEAC on diesel engines is
                based on the review of technical publications discussed earlier in
                Section VI.C.1.c)(5).
                ---------------------------------------------------------------------------
                 \881\ Docket ID NHTSA-2018-0067-12039, at p. 3.
                ---------------------------------------------------------------------------
                (9) Alternative Fuel Engines
                 CNG engines use compressed natural gas as a fuel source. The fuel
                storage and supply systems for these engines differ tremendously from
                gasoline, diesel, and flex fuel vehicles. CNG engines were a baseline-
                only technology and were not applied to any vehicle that was not
                already CNG-based in NHTSA's analysis, per EPCA/EISA's restrictions on
                considering dedicated alternative fueled vehicles to set fuel economy
                standards.882 883 However, for the EPA program the agencies
                allowed any vehicle to adopt CNG engines. The NPRM MY 2016 analysis
                fleet did not include any dedicated CNG vehicles to simulate in the
                CAFE Model.
                ---------------------------------------------------------------------------
                 \882\ NHTSA's provisions for dedicated alternative fuel vehicles
                in 49 U.S.C. 32905(a) state that the fuel economy of any dedicated
                automobile manufactured after 1992 shall be measured based on the
                fuel content of the alternative fuel used to operate the automobile.
                A gallon of liquid alternative fuel used to operate a dedicated
                automobile is deemed to contain 0.15 gallon of fuel. Under EPCA, for
                dedicated alternative fuel vehicles, there are no limits or phase-
                out for this special fuel economy calculation, unlike for duel-
                fueled vehicles, as discussed below.
                 \883\ EPA's provisions for dedicated alternative fuel vehicles
                that are able to run on compressed natural gas (CNG) currently are
                eligible for an advanced technology multiplier credit for MYs 2017-
                2021.
                ---------------------------------------------------------------------------
                 In addition, for the NPRM and this final rule analysis, NHTSA
                modified the CAFE model to include the specific provisions related to
                AFVs under the CO2 standards. In particular, the CAFE model
                now carries a full representation of the production multipliers related
                to electric vehicles, fuel cell vehicles, plug-in hybrids, and CNG
                vehicles, all of which vary by year through MY 2021.
                (10) Emerging Gasoline Engine Technologies
                 Manufacturers, suppliers, and researchers continue to create a
                diverse set of fuel economy technologies, some of which are still in
                the early stages of the development and commercialization process. Due
                to uncertainties in the cost and capabilities of emerging technologies,
                some new and pre-production technologies are not a part of the CAFE
                model simulation. As discussed throughout this section and in VI.B.3,
                the agencies declined to include technologies in the analysis where the
                agencies did not believe those technologies would be feasible in the
                rulemaking timeframe, or the agencies did not have appropriate data
                upon which to generate an estimate of how effective the technology is
                that could be applied across the ten vehicle classes. Evaluating and
                benchmarking promising fuel economy technologies as they enter
                production-intent stages of development continues to be a priority as
                commercial development matures.
                 UCS and ICCT commented that the agencies should consider novel
                engine designs.\884\ Specifically, ICCT stated that the agencies should
                consider a more advanced HCR technology called HCCI (similar to Mazda's
                Skyactiv-X) by estimating efficiency and cost to EPA's process that
                assigned effectiveness estimates using LPM. They stated that ``the
                agencies developed estimates for ADEAC in the NPRM and the associated
                modeling even without conclusive and independently verifiable
                effectiveness.''
                ---------------------------------------------------------------------------
                 \884\ ICCT, Full Comments Summary. Docket No. NHTSA-2018-0067-
                117411, at I-17 to I-19.
                 UCS, Comment. Docket No. NHTSA-2018-0067-12039, at pp. 6 & 7.
                ---------------------------------------------------------------------------
                 In response to comments, a number of technologies were added for
                the final rule analysis, and adoption features were refined
                accordingly, as discussed further in Section VI.C.1.e). New engine
                technologies and combinations include Atkinson engine technology
                allowed with P2 HEV, new high compression ratio engine (HCR0), variable
                compression ratio engine, variable geometry turbo engine, variable
                geometry turbo with electric assist engine, diesel with advanced
                cylinder deactivation engine, turbo with cylinder deactivation engine,
                diesel with manual transmission, diesel with start-stop, and PHEV-turbo
                with 20 mile range, and PHEV-turbo with 50 mile range.
                 The agencies also disagree with ICCT's comment that because ADEAC
                was developed without ``conclusive and independently verifiable
                effectiveness'' estimates, and as such the agencies should allow HCCI
                technology as well. First, conclusive estimates for ADEAC effectiveness
                were based on CBI data from both manufacturers and suppliers, technical
                publications, and engineering judgement. The references can be reviewed
                in the previous Section VI.C.1.c)(5) Advanced Cylinder Deactivation
                Technologies. In addition, the agencies benchmarked the first prototype
                vehicle equipped with skip-fire, and discussed potential application of
                it for other engines. A similar level of data has not been made
                available for HCCI engine technologies.
                 The agencies also believe that the technology associated with Mazda
                SkyActiv-X has been mischaracterized by ICCT and other commenters, and
                declined to include a specific representation of the SkyActiv-X family
                of technologies in the analysis for two reasons. The engine known as
                Skyactiv-X is characterized by Mazda as a unique spark plug controlled
                compression ignition (SPCCI) technology, 2-liter displacement, 4-
                cylinder engine with mechanical compression ratio of 16.3:1 operating
                on 95 RON fuel (91 AKI) with
                [[Page 24417]]
                a mild hybrid system.\885\ The NPRM and this final rule analysis may
                not have the exact technology combination associated with this vehicle,
                but the analysis does include technologies that are representative of
                them, that could enable the benefits employed by the Mazda engine. A
                mild hybrid system is available for adoption in both the NPRM and this
                final rule analysis.
                ---------------------------------------------------------------------------
                 \885\ Mazda Press Release. ``Revolutionary Mazda Skyactiv-x
                engine details confirmed sales start.'' May 6, 2019. https://www.mazda-press.com/eu/news/2019/revolutionary-mazda-skyactiv-x-engine-details-confirmed-as-sales-start/. Last accessed Dec, 11,
                2019.
                ---------------------------------------------------------------------------
                 Also, the effectiveness associated with this engine was from
                European test cycles and cannot be compared for U.S. application.
                European compliance tests are significantly different than those in the
                U.S., especially when it comes to fuel type and test cycles. Any
                effectiveness data provided for this engine or any non-U.S. engine
                cannot be used for U.S. vehicle application without an adjustment for
                fuel and emissions. For example, the higher-octane fuel used in Europe
                enables engines to operate at higher compression ratios across wider
                areas of engine operation.
                 The agencies further believe that with the technology additions for
                the final rule discussed in previous sections, the analysis reasonably
                represents the suite of engine technologies that could be available in
                the rulemaking time frame. Manufacturers, suppliers, and researchers
                continue to create a diverse set of fuel economy technologies. However,
                due to the uncertainties in the cost, manufacturing, and intellectual
                property concerns like those identified by commenters, the agencies did
                not consider prototype technologies in the final rule analysis.
                (11) Engine Lubrication and Friction Reduction Technologies
                 Manufacturers have already widely adopted both lubrication and
                friction reduction technologies. Previous agency analysis considered
                these improvements in combination as Improved Low Friction Lubricants
                and Engine Friction Reduction (LUBEFR). The NPRM analysis included
                advanced engine maps that already assume application of low-friction
                lubricants and engine friction reduction technologies, and therefore
                additional levels of friction reduction were not considered. Low-
                friction lubricants including low viscosity and advanced low-friction
                lubricant oils are now available, and widely used. Manufacturers may
                make engine changes and conduct durability testing to accommodate the
                lubricants. The level of low-friction lubricants exceeded 85 percent
                penetration in the MY 2016 fleet.\886\ Reduction of engine friction can
                be achieved through low-tension piston rings, roller cam followers,
                improved material coatings, more optimal thermal management, piston
                surface treatments, and other improvements in the design of engine
                components and subsystems that improve efficient engine operation.
                ---------------------------------------------------------------------------
                 \886\ NPRM CAFE Model Market Data file.
                ---------------------------------------------------------------------------
                 Meszler Engineering on behalf of NRDC commented that ``the NPRM
                CAFE model no longer considers advanced lubricants and evolutionary
                friction reduction (LUBEFR) to be adoptable. As a result, no fuel
                efficiency improvement credits are available. Engine friction reduction
                is an ongoing evolutionary process that should generate benefits on the
                order of 5 percent or so increase in fuel economy over a multiyear
                forecast period, with costs totaling approximately $100. Moreover, the
                technology is a benefit of ongoing industry research and evolutionary
                engine improvements so that it is easily `adoptable' and deployed
                throughout the fleet. Accordingly, NHTSA should revise the NPRM CAFE
                model to reinstate the ability to adopt evolutionary friction reduction
                technology.'' \887\
                ---------------------------------------------------------------------------
                 \887\ Meszler Engineering. Docket ID NHTSA-2018-0067-11723, at
                p. 32.
                ---------------------------------------------------------------------------
                 The agencies disagree with Meszler that a five percent fuel economy
                improvement attributable to lubricants and evolutionary friction
                reduction is continuously feasible. The MY 2017 baseline vehicles have
                incorporated many technologies like low viscosity engine oil,
                integrated exhaust manifold for faster oil warmup, and internal
                component friction reduction.\888\ \889\ \890\ The LUB and EFR
                technologies are a legacy of the existing rulemaking work going back to
                the 2010 CAFE and CO2 rule for MY 2012 to MY 2016.\891\ The
                agencies believe that many of these technologies have been incorporated
                in many of the engines in the baseline fleet, and therefore the engine
                maps used for the NPRM and final rule analysis incorporated them as
                well. Furthermore, manufactures have raised concerns over issues with
                further decreasing oil viscosity; specifically, manufacturers have
                articulated concerns that damage caused by low speed pre-ignition
                (LSPI) \892\ can damage an engine.\893\ \894\ \895\
                ---------------------------------------------------------------------------
                 \888\ Wards Auto. ``Infiniti's Brilliantly Downsized V-6 Turbo
                Shines.'' July 11, 2017. Available at https://www.wardsauto.com/print/engines/infiniti-s-brilliantly-downsized-v-6-turbo-shines.
                Last accessed Dec. 11, 2019. Nissan Motor Corp. ``Mirror Bore
                Coating.'' Available at https://www.nissan-global.com/EN/TECHNOLOGY/OVERVIEW/mirror_bore_coating.html. Last accessed Dec 11, 2019.
                 \889\ Toyota's 2AR-FE I4 and 2GR-FE V6 use 0-W20.
                 \890\ Audi Media Center. ``Efficiency and driving pleasure:
                innovative V engines at Audi.'' Available at https://www.audi-mediacenter.com/en/techday-on-combustion-engine-technology-8738/efficiency-and-driving-pleasure-innovative-v-engines-at-audi-8748.
                Last accessed Dec.11, 2019.
                 \891\ 75 FR 25373.
                 \892\ LSPI is an abnormal combustion event in which the fuel-air
                mixture ignites before intended, causing excessive pressures inside
                the engine's cylinders. In mild cases, this can cause engine noise,
                but when severe enough, LSPI can cause engine damage. There are
                several factors that contribute to LSPI, of which lubricating oil
                has been observed to be one.
                 \893\ Motor Magazine. ``Will ILSAC GF-6 Ever Be Approved?'' Nov,
                20, 2018. Available at http://newsletter.motor.com/2018/20181120/!ID_Infineum_ILSAC_GF-6.html. Last accessed Dec 11, 2019.
                 \894\ Chevron. ``Low Speed Pre-ignition.'' Available at https://www.oronite.com/about/news/low-speed-pre-ignition.aspx. Last
                accessed Dec. 11, 2019.
                 \895\ Elliott, I., Sztenderowicz, M., Sinha, K., Takeuchi, Y. et
                al., ``Understanding Low Speed Pre-Ignition Phenomena across Turbo-
                Charged GDI Engines and Impact on Future Engine Oil Design.'' SAE
                Technical Paper 2015-01-2028, 2015, available at https://doi.org/10.4271/2015-01-2028.
                ---------------------------------------------------------------------------
                 In response to the comment that engine friction reduction
                technology is evolutionary technology, the agencies introduced one
                level of friction reduction (EFR) for the final rule analysis. The
                agencies estimated a 1.4 percent effectiveness for this type of
                technology based on the 2015 NAS report assessment of further
                improvements in lubrication and friction.\896\
                ---------------------------------------------------------------------------
                 \896\ 2015 NAS at pp. 28 & 29.
                ---------------------------------------------------------------------------
                d) How the Agencies Assign Engine Technologies to the Baseline Fleet
                [[Page 24418]]
                 Manufacturers have made significant improvements in fuel economy
                and CO2 emissions reductions since the MY 2012 rulemaking
                analysis.\897\ \898\ The agencies expended substantial effort to update
                the analysis fleet from the MY 2016 representative fleet used for the
                NPRM to a MY 2017 analysis fleet used for this final rulemaking to
                capture the technologies manufacturers have used to increase their
                fleet's fuel economy and CO2 emissions performance. Detailed
                discussion of the model year 2017 fleet development and application can
                be found in VI.B.1. The agencies extensively updated the new MY 2017
                fleet engine technologies using available manufacturer final model year
                CAFE compliance submissions to the agencies, as well as manufacturer
                press release specifications, agency-sponsored vehicle benchmarking
                studies, review of available technical publications, and through
                manufacturer CBI.\899\
                ---------------------------------------------------------------------------
                 \897\ EPA. ``2018 EPA Automotive Trends Report'' 12 pp, 421 K,
                EPA-420-S-19-001, March 2019. https://www.epa.gov/automotive-trends/download-automotive-trends-report#Full%20Report last accessed Feb.
                12, 2020
                 \898\ FOTW #1108, Nov 18, 2019: Fuel Economy Guide Shows the
                Number of Conventional Gasoline Vehicle Models Achieving 45 miles
                per gallon or Greater is Increasing. DOE VTO. Available at https://www.energy.gov/eere/vehicles/articles/fotw-1108-november-18-2019-fuel-economy-guide-shows-number-conventional. Last accessed Nov 18,
                2019.
                 \899\ NPRM CAFE Market Data file.
                ---------------------------------------------------------------------------
                 The data for each manufacturer was used to determine which
                platforms shared engines and to establish the leader-follower
                relationships between vehicles. Within each manufacturer's fleet,
                engines were assigned unique identification designations based on
                configuration, and technologies applied, along with other
                characteristics. The data were also used to identify the most similar
                engine among the IAV engine maps, as discussed in Section VI.C.1.
                 Just like the real-world vehicle variants, the CAFE model considers
                differences between each vehicle like base performance and higher
                performance levels. For example, the 2017 Ford F150 has many variants
                with different types of engines like the 2.7L turbocharged V6, 3.3L
                naturally-aspirated V6, 3.5L turbocharged V6, and 5L naturally-
                aspirated V8. In contrast to the LPM, the CAFE model rosters each
                variant level and powertrain application individually. This variation
                is accounted for as engine technologies are assigned in the analysis
                fleet.
                 As a result of new information available since publication of the
                NPRM and comments received to the NPRM, the agencies included
                additional engine technologies in the compliance analysis, expanding
                the total number of engine technologies available from 16 to 23. This
                expansion is a direct result of comments received to the NPRM and
                further enables the agencies' capabilities to accurately and,
                realistically, characterize the technologies present on an engine found
                in the analysis fleet. This collection of technologies represents the
                best available information the agencies have, at the time of this
                action, regarding both currently available engine technologies and
                engine technologies that could be feasible for application to the U.S.
                fleet during the rulemaking timeframe. The agencies believe this effort
                has yielded the most technology-rich and accurate analysis fleet
                utilized by the CAFE model to date.
                 In some cases, however, it was necessary for the agencies to
                substitute an engine map that closely represented an engine technology
                that were effectively the same, or, based on engineering judgement,
                were the best available proxy at the time of the analysis. For example,
                many manufacturers offer their own proprietary VVT engine technologies
                and so the agencies assigned the same engine map for all of these VVT
                in the baseline fleet. The CAFE model uses compliance CAFE and
                CO2 values for baseline vehicles and so it's not as relevant
                to have exact technology assignment type as it more important to
                provide the advanced vehicle have adopted to date. For further
                discussion of this see section VI.A.3 Fuel-Savings Technologies. This
                substitution was necessary, in some cases, where an ``exact-match''
                engine map was not available for application to a specific vehicle and/
                or vehicle specific engine application. The agencies leveraged a series
                of engine operating characteristic maps developed by industry suppliers
                and, in some cases, the agencies themselves, to assign the closest
                baseline engine map for the analysis.
                 As discussed in Section VI.C.1.b), these engine maps provide
                operational characteristics such as horsepower, torque, or efficiency
                at a specified point in an engine's operational range. These
                operational maps are developed based on a given set of engine
                characteristics and technologies applied to that engine. Engine maps
                are closely held by vehicle manufacturers and are typically considered
                intellectual property. As such, vehicle manufacturers are not typically
                willing provide the operational maps to the agencies, where it would
                ultimately be in the purview of competitors. In some instances,
                manufacturer engine maps are published in media such as technical
                papers or conference presentation materials. However, these publicly
                available engine maps are, in nearly all instances, void of critical
                information that would enable their use for meaningful simulation and
                modeling.
                 Therefore, the agencies are generally limited to the catalog of
                engine maps they have developed through contracts and, where possible,
                in-house which, in turn, yields the need for sound, engineering
                judgement-based substitution of an engine map as a proxy for an engine
                application in the marketplace. Unfortunately, this is necessary as the
                agencies are unable to fund the development of engines maps for every
                possible engine and technology combination available for sale. However,
                it is important to note the agencies do have a substantial catalog of
                engine maps to leverage and continue to fund the development of new
                maps as new technologies enter the marketplace. Additional information
                on the agencies' catalog of engine maps used for this this final
                rulemaking can be found in Section VI.C.1.b).
                 Some engine technologies are designated in the CAFE Model as
                ``baseline only'' technologies, meaning these are characteristics such
                as engine configuration, architecture, or a technology that is
                considered inherent to the fleet for the given model year, an example
                for the MY 2017 fleet used in this analysis is variable-valve-timing
                (VVT). Beyond the aforementioned configurations and technology, engine
                technologies that can be applied to a future engine and, eventually, to
                a vehicle in the compliance modeling are only available at a vehicle
                redesign. As such, a vehicle will only adopt a new engine according to
                the application schedule defined as a CAFE model input.
                e) Engine Adoption Features
                 Engine adoption features are defined through mechanisms like
                technology path logic or the application of selection logic, refresh
                and redesign cycles, and phase-in capacity limits. Most of the
                technology adoption features from the NPRM have been carried over for
                the final rule analysis. However, the final rule analysis also included
                adoption features for the new technologies incorporated in the final
                rule analysis. For a detailed discussion of CAFE model path logic for
                the final rule analysis, including technology supersession logic and
                technology mutual exclusivity logic, please see Section IV.
                 Figure VI-18 and Figure VI-19 below show the engine technology
                paths used for the NPRM and this final rule analysis, respectively. The
                engine
                [[Page 24419]]
                technology paths have increased to incorporate new advanced
                technologies manufacturers could adopt into their fleet.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.152
                [GRAPHIC] [TIFF OMITTED] TR30AP20.153
                BILLING CODE 4910-59-C
                 Similar to the 2012 final rule for MYs 2017-2025, this final rule
                analysis also considered real-world limits when the defining the rate
                at which technologies can be deployed.\900\ During the rulemaking
                timeframe, manufacturers are expected to go through the normal
                automotive business cycle of redesigning and upgrading their light-duty
                vehicle products. This allows manufacturers the time needed to
                incorporate fuel economy improving and CO2 reducing
                technologies into their normal business cycle. This is important
                because it has the potential to avoid the much higher costs that could
                occur if manufacturers need to add or change technology at times other
                than their scheduled vehicle redesigns. This time period also provides
                manufacturers the opportunity to plan for compliance using a multi-year
                time frame, again consistent with normal business practice.
                ---------------------------------------------------------------------------
                 \900\ 77 FR 62712.
                ---------------------------------------------------------------------------
                 Section II.G.3.a of the NPRM provided substantial discussion of how
                an ``application schedule'' is used by the CAFE model to determine when
                manufacturers are assumed to be able to apply a given technology to a
                vehicle. The NPRM application schedule for engine technologies is
                reproduced in Table VI-43, which shows that all of the
                [[Page 24420]]
                engine technologies may only be applied (for the first time) during
                redesign.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.154
                 For this final rulemaking action, a similar schedule is employed,
                and has been updated with information gathered since the NPRM and
                through comments provided to the agencies.
                 Table VI-44 presents the engine technology application schedule
                used for the final rule CAFE modeling.
                [[Page 24421]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.155
                 Fuel economy improving and CO2 reducing technologies for
                vehicle applications vary widely in function, cost, effectiveness, and
                availability. Some of these attributes, like cost and availability,
                vary from year to year. New technologies often take several years to
                become available across the entire market. The agencies use phase-in
                caps to manage the maximum rate that the CAFE model can apply new
                technologies. Phase-in caps are intended to function as a proxy for a
                number of real-world limitations in deploying new technologies in the
                auto industry. These limitations can include but are not limited to,
                engineering resources at the OEM or supplier level, restrictions on
                intellectual property that limit deployment, and/or limitations in
                material or component supply as a market for a new technology develops.
                Without phase-in caps, the model may apply technologies at rates that
                are not representative of what the industry is actually capable of
                producing, which would suggest that more stringent standards might be
                feasible than actually would be. Table VI-45 and Table VI-46 below
                shows the phase-in caps between the NPRM and this final rule analysis,
                respectively.
                 Most engine technologies are available at a rate of 100 percent in
                MY2017 for the final rule analysis. Some advanced technologies that
                have been recently introduced for one or two vehicle models are phased
                in at lower rates. Technologies such as ADEAC and TURBOD are phase in
                at rates that represent manufacturers' adoption capability and
                typically have complementary effectiveness compared to other advanced
                technologies. These lower phase-in caps also represent intellectual
                property and functional performance concerns.
                BILLING CODE 4910-59-P
                [[Page 24422]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.156
                [[Page 24423]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.157
                BILLING CODE 4910-59-C
                 Comments received on engine adoption features were mixed, with
                manufacturers generally supporting the NPRM methodology, and CARB and
                NGOs opposing it. Several manufacturers commented, both in their public
                comments or on a CBI basis, that many of the emerging engine
                technologies had the potential to improve vehicle fuel economy, but
                were technically complex and addressed many of the same issues as other
                existing engine technologies.
                 We agree with manufacturers that broadly, there are technologies
                that, in theory, present large potential effectiveness improvements
                like VCR, ADEAC, and others. However, the agencies believe it is
                important to assure realistic adoption of these technologies into the
                fleet in the rulemaking time frame, so that the rulemaking analysis
                accurately represents the costs and benefits of different regulatory
                alternatives considered. If the agencies were to select stringency
                based on an assumption that an emerging technology would see widespread
                adoption, and then it does not, the benefits of that stringency level
                would not be realized. The agencies have taken steps in the NPRM and
                this final rule analysis to consider the manufacturability and
                feasibility of these technologies for different vehicle types and
                manufacturers. Discussed earlier, the analysis considers these and
                other concerns by accounting for product cadence, and by implementing
                phase-in caps and skips, and by designating technology phase-in and
                phase-out years. Similar to the 2012 final rule, this final rule
                analysis employed these strategies to reflect better the real-world
                considerations faced by manufacturers.
                 EDF commented, referencing EPA's statutory command prescribed in
                Section 202(a) of the Clean Air Act that:
                 EPA's task is thus to identify the major steps necessary for
                `development and application of the requisite technology,' and then
                the respective standard `shall take effect.' These individual
                decisions are highly consequential: As noted above, without changing
                anything else about the agencies' analysis, allowing HCR2 would
                reduce augural compliance costs by $619--or about 30% of the total
                difference between the augural and rollback scenarios. The
                proposal's rejection of these technologies nowhere justifies how the
                (unfounded and cursorily justified) concerns accord with the
                agency's limited discretion under Section 202(a)(2) and duty to
                `press for the development and application of improved technology
                rather than be limited by that which exists today.' If the agency is
                to predict more than the results of merely assembling pre-existing
                components, it must have some leeway to deduce results that are not
                represented by present data.\901\
                ---------------------------------------------------------------------------
                 \901\ NHTSA-2018-0067-12108 at 104.
                 CARB also commented that the CAFE Model prevents manufacturers
                ``from
                [[Page 24424]]
                switching between a turbocharged and HCR pathways under the premise
                that manufacturers either would not develop both or would be committed
                irreversibly to one path or the other. This assumption is not based in
                reality and is not reflective of actual industry practice--
                manufacturers who have pursued turbocharging have also already pursued
                HCR engines for other vehicles in their line-up. For example, General
                Motors (GM) utilizes downsized turbocharging in some vehicles, such as
                the newly designed 2019MY Silverado pick-up and the Malibu sedan which
                has two different turbocharged engine options. GM also has a third
                offering in the Malibu sedan which is an HCR naturally aspirated 1.8L
                equipped with cooled exhaust gas recirculation (CEGR) mated to a hybrid
                electric system.'' \902\
                ---------------------------------------------------------------------------
                 \902\ NHTSA-2018-0067-11873 at 109.
                ---------------------------------------------------------------------------
                 CARB's observation was true for the NPRM analysis, however for the
                final rule analysis the agencies allowed manufacturers to adopt engine
                technologies from alternate tree paths, when incorporating
                electrification technology, see Section VI.C.3.c). The agencies still
                believe that if manufacturers have invested in one type of engine
                technology for their vehicles that they would not transition to another
                technology except in the case of a major vehicle powertrain redesign,
                such as the inclusion of an HEV system. Additional discussion on this
                issue is presented in Section VI.B.1.
                 The following sections discuss adoption features specific to
                individual engine technologies, including comments received and updates
                (or not) for the final rule analysis.
                (1) Basic Engines
                 Most vehicles in the MY 2017 analysis fleet that are DOHC or SOHC/
                OHV spark ignited engines and are not downsized turbocharged engines
                have any two combinations of VVT, VVL, SGDI or DEAC.\903\ For the NPRM,
                only engines with 6-cylinders or more could adopt DEAC and ADEAC.
                ---------------------------------------------------------------------------
                 \903\ EPA. ``2018 EPA Automotive Trends Report'' 12 pp, 421 K,
                EPA-420-S-19-001, March 2019. https://www.epa.gov/automotive-trends/download-automotive-trends-report#Full%20Report (last accessed Feb.
                12, 2020) p. 72.
                ---------------------------------------------------------------------------
                 HDS on behalf of CARB commented that in the NPRM analysis VVL,
                which is cost ineffective compared to other conventional technologies,
                was always included in an adopted technology package.\904\ HDS further
                stated that the ``effectiveness of VVL is even smaller when the
                technology is combined with turbocharged downsized engines.''
                Accordingly, HDS stated that removing VVL from the base pathway would
                save $314 but reduce fuel economy by only 1.4 percent, according to the
                LPM.
                ---------------------------------------------------------------------------
                 \904\ NHTSA-2018-0067-11985 at p.34.
                ---------------------------------------------------------------------------
                 The agencies did not agree with HDS' assessment of the NPRM
                analysis. The agencies do not agree VVL was forced to be adopted in the
                analysis fleet and do not agree with how technology effectiveness
                values compare to LPM estimates. As discussed earlier in the
                effectiveness and modeling section, each engine technology was modeled
                independently and the CAFE model was allowed to adopt the most cost
                effective technology. Therefore, it is inaccurate to state, a
                technology is less effective, especially when comparing LPM.
                Particularly because VVL technologies reduce pumping losses in engines,
                so it is realistic that other technologies, that also reduce pumping
                losses, have synergetic effect. This is specifically true for
                turbocharged engines.
                 ICCT commented that DEAC technology should be available for every
                engine, and should not be limited to 6-cylinder and higher cylinder
                count engines. ICCT and CARB also commented that DEAC should be allowed
                on turbocharged engines. ICCT also commented that ADEAC should be
                widely available as it can be a viable technology application for
                various other powertrain technology combinations.\905\ Furthermore,
                CARB commented ``automakers will combine technologies like
                turbocharging, HCR and DEAC as well as more technologies when they have
                cost-effectiveness synergies.'' \906\
                ---------------------------------------------------------------------------
                 \905\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-13.
                 \906\ CARB at p. 6.
                ---------------------------------------------------------------------------
                 The agencies agree with ICCT that DEAC and ADEAC could be applied
                to additional engine types, including turbocharged engines. However,
                the agencies disagree with ICCT that ADEAC should be widely applied to
                all powertrain technology combinations in this analysis. The agencies
                have updated the final rule analysis to allow DEAC and ADEAC for
                various engine cylinder counts and for turbocharged engines.
                 For the final rule analysis, both DEAC and ADEAC technologies can
                be adopted by any naturally aspirated engine. Similarly, any
                turbocharged engine can also adopt cylinder deactivation technology, as
                characterized by TURBOD and TURBOAD in the CAFE model. In this final
                rule analysis, the agencies distinguished cylinder deactivation
                technologies between naturally aspirated and forced air induction
                systems.
                 For the final rule analysis, the agencies allow any combination of
                VVT, VVL, SGDI and DEAC to be adopted for any engine displacement and
                cylinder count. Figure VI-18 below shows the basic engine paths a
                vehicle could traverse for the final rule analysis. Similar to the
                NPRM, the agencies have not changed the adoption features of the
                technologies shown in Figure VI-18, with one exception. Vehicles that
                are SOHC or DOHC configuration that do not have VVT in the baseline can
                now adopt it.
                 Finally, the agencies disagree with ICCT and CARB that these DEAC,
                ADEAC, TURBOD, and TURBOAD should apply beyond these configurations.
                DEAC's fundamental benefits are driven by reducing pumping losses and
                by enabling the engine to operate in a more thermal efficient region of
                the engine fuel map. Conventional spark-ignited engines control airflow
                into the cylinders via a throttle operated by the driver to provide the
                level of power that is delivered.\907\ In an 8-cylinder engine, when
                driving in light load conditions such as highway driving, there are
                lower engine power requirements. In a throttle controlled system,
                engine pumping losses increase as air flow decreases. A way to reduce
                pumping loss in an engine is by increasing the airflow into the
                cylinders. By deactivating a set of cylinders, the same power output
                can be delivered by a ``smaller'' engine. Many technologies modeled for
                this analysis work to reduce pumping losses, but through other
                mechanisms like VVT, VVL, downsized engines with turbochargers, high
                compression Atkinson mode cycle, and Miller Cycle.\908\ Transmissions
                with a higher number of gears also provide the opportunity to reduce
                pumping work of the engine.\909\
                ---------------------------------------------------------------------------
                 \907\ A throttle is the mechanism by which fluid flow is managed
                by constriction or obstruction. An engine's power can be increased
                or decreased by the restriction of inlet gases, but usually
                decreased.
                 \908\ 2015 NAS at p. 23.
                 \909\ 2015 NAS at p.173.
                ---------------------------------------------------------------------------
                 As discussed earlier, DEAC can reduce pumping losses, so when
                combined with other technologies that also reduce pumping losses, like
                downsized turbocharged engines, the benefits for cylinder deactivation
                are lower than for naturally aspirated engines because downsized
                turbocharged engines already have lower pumping losses due to having a
                downsized engine.\910\
                ---------------------------------------------------------------------------
                 \910\ 2015 NAS at p. 34.
                ---------------------------------------------------------------------------
                [[Page 24425]]
                (2) Turbocharged Downsized Engines
                 About 23 percent of vehicles in the MY 2017 baseline fleet had
                turbocharged engines. For the final rule analysis, the agencies allowed
                any basic engine to adopt turbo engine technology (TURBO1, TURBO2 and
                CEGR1) from the Turbo path similar to the NPRM analysis. This includes
                any combination of VVT, VVL, SGDI and DEAC for both SOHC and DOHC
                configurations. Vehicles that have turbocharged engines in the baseline
                fleet will stay on the turbo engine path to prevent unrealistic engine
                technology change in a short timeframe considered in the rulemaking
                analysis. Turbo path is a mutually exclusive technology in that it
                cannot be adopted for HCR, diesel, ADEAC, CNG and powersplit PHEVs.
                (3) Non-HEV Atkinson Mode Engines
                 The NPRM analysis allowed limited application of HCR engines (HCR1
                and HCR2) to vehicles in the MY 2016 baseline fleet.\911\ As discussed
                above, applying HCR1 or HCR2 technologies to a vehicle resulted in
                overstated effectiveness values relative to the baseline VVT
                engine,\912\ because of differences in how those maps were developed
                compared to the IAV engine maps used for the majority of the technology
                analysis. In an attempt to avoid unrealistic results in the NPRM,
                adoption of HCR1 (Eng24) technology was limited to only manufacturers
                that demonstrated existing use of high compression ratio technology.
                HCR was disallowed for other manufacturers that demonstrated an intent
                to develop other advanced technologies incompatible with HCR
                technology. In addition, the agencies disallowed HCR engines from being
                applied to vehicles with greater performance requirements, like 6- and
                8-cylinder vehicles, because the higher load requirements from these
                vehicles would force the engine to exit the Atkinson mode, where
                maximum efficiency is achieved.
                ---------------------------------------------------------------------------
                 \911\ 83 FR 43037.
                 \912\ 83 FR 43029 Figure II-1--Simulated Technology
                Effectiveness Value.
                ---------------------------------------------------------------------------
                 The Alliance commented in agreement with the application
                restrictions for HCR1 in the NPRM, listing the following
                justifications: ``Packaging and emission constraints associated with
                intricate exhaust manifolds needed to mitigate high load/low
                revolutions per minute knock; Inherent performance limitations of
                Atkinson cycle engines; and Extensive capital and resources required
                for manufacturers to shift to HCR from other established technology
                pathways (e.g., downsized turbocharging).'' \913\ Ford similarly
                commented in support of ``the more restrained application of HCR1 in
                the Proposed Rule, an approach that recognizes the investment,
                packaging, performance and emissions factors that will limit
                penetration of this technology.'' \914\
                ---------------------------------------------------------------------------
                 \913\ NHTSA-2018-0067-12073.
                 \914\ NHTSA-2018-0067-11928.
                ---------------------------------------------------------------------------
                 In contrast, CARB stated that the constraint on HCR1 engines was
                inappropriate and did not reflect reality,\915\ and stated that the
                agencies failed to supply any detailed rationale as to why HCR
                applications were so constrained in the CAFE Model. Specifically, CARB
                took issue with the justification that HCR1 is limited in the CAFE
                model because it is ``not suitable for MY 2016 baseline vehicle models
                that have 8-cylinder engines and in many cases 6-cylinder engines.''
                \916\ CARB stated that ``the HCR1 technology is declared not suitable
                on 207 of the 288 engines cumulatively used by all of industry
                including over 50 percent of the 4 cylinder engines and nearly 90
                percent of the 6 cylinder engines instead of only being restricted from
                8 cylinder and `in many cases 6 cylinder engines.' '' CARB also stated
                that the implied rationale for not allowing HCR1 to be applied to 6-
                and 8-cylinder engines because trucks or larger vehicles could not
                utilize it is unreasonable, as the Toyota Tacoma used a 3.5L V6 HCR
                Atkinson-like engine since MY 2016. CARB stated that the Toyota Tacoma
                was properly assigned a HCR1 engine in the MY 2016 analysis fleet file,
                but the engine was disallowed from other Toyota V6 engines utilized in
                vehicles like the Sienna minivan and 4Runner SUV. CARB commented that
                ``[i]f the intended rationale is that HCR engines will have
                insufficient low end torque to satisfy truck-like towing demands, it
                would be inappropriate to restrict the engine from minivan and SUV
                applications which have a lower tow rating and lower expected towing
                demands.'' Finally, CARB stated that the HCR1 package restrictions were
                inappropriate, as there was no mechanism in the CAFE model to represent
                appropriately the MY 2019 Dodge Ram 1500 5.7L V8 that uses ``a higher
                compression ratio than earlier versions and using its VVT system to
                reduce pumping losses via delayed, or late, intake valve closing--
                resulting in an HCR-like engine with an over-expanded or Atkinson
                cycle.''
                ---------------------------------------------------------------------------
                 \915\ NHTSA-2018-0067-11873.
                 \916\ 83 FR 43038.
                ---------------------------------------------------------------------------
                 Similarly, Meszler Engineering Services, commenting on behalf of
                NRDC, commented that HCR1 appears as a baseline technology on vehicles
                representing about 4 percent of the baseline non-hybrid vehicle market,
                and is subsequently applied to only 23 percent of the market. Meszler
                stated that the ``relative cost effectiveness of the technology is
                perhaps best illustrated by the fact that the market penetration of HCR
                technology on non-hybrid vehicles under the augural standard is modeled
                to be 27 percent of 2032 sales, exactly equal to the baseline
                penetration of 4 percent and the allowable adoption fraction of 23
                percent. In other words, the technology was adopted by every vehicle
                that was not explicitly prohibited (by NHTSA) from doing so.'' EDF
                commented that ``NHTSA has further imposed artificial and unreasonable
                constrains on the use of certain technologies that does not match how
                automakers are applying them in vehicles today,'' stating that HCR1
                represented a technology that had been in the marketplace for many
                years and had been applied by several manufacturers, ``[y]et, even for
                MY 2030 vehicles and beyond, NHTSA only allows the use of HCR1 by about
                30 percent of the U.S. fleet.'' \917\
                ---------------------------------------------------------------------------
                 \917\ NHTSA-2018-0067-12108.
                ---------------------------------------------------------------------------
                 In considering the comments, the agencies agree with commenters
                that the HCR1 engine application was overly limited for the NPRM
                analysis. As a result, the agencies have expanded the availability of
                HCR1 technology for the final rule analysis. The refined adoption
                features for HCR1 are discussed below. The new adoption features do
                maintain considerations for performance neutrality. Comments about how
                the characterization of engine technologies in the analysis fleet
                impacted HCR technology adoption in subsequent model years are
                addressed in Section VI.C.1.d) Baseline Fleet Engine Tech.
                 Regarding HCR2, the Alliance commented in support of ``the decision
                to exclude the speculative HCR2 technology from the analysis.'' \918\
                The Alliance continued, ``[a]s previously documented in Alliance
                comments, the inexplicably high benefits ascribed to this theoretical
                combination of technologies has not been validated by physical
                testing.'' Similarly, Ford stated that ``[t]he effectiveness of the
                `futured' Atkinson package (HCR2) that includes cooled exhaust gas
                recirculation (CEGR) and cylinder deactivation (DEAC) is excessively
                high, primarily due to overly-optimistic efficiencies in the base
                engine map, insufficient accounting of CEGR and DEAC integration
                losses, and no accounting of the impact of 91RON
                [[Page 24426]]
                Tier 3 test fuel. Given the speculative and optimistic modeling of this
                technology combination, Ford supports limiting the use of HCR2
                technology to reference only, as described in the Proposed Rule.''
                \919\
                ---------------------------------------------------------------------------
                 \918\ NHTSA-2018-0067-12073.
                 \919\ NHTSA-2018-0067-11928.
                ---------------------------------------------------------------------------
                 In contrast, several commenters disagreed with the agencies'
                decision to limit the adoption of HCR2 engines, stating that the
                technology was clearly applicable during the rulemaking timeframe, as
                the technology was already being applied by manufacturers, and that the
                technology was cost-effective, as shown by the agencies' own modeling.
                 ICCT commented that ``[i]t is clear that the agencies have
                artificially excluded a known technology that is applicable in the
                timeframe of the rulemaking.'' \920\ ICCT commented that ``[d]espite
                the facts that (as discussed above) the agencies have cost and
                effectiveness data for this technology, many automakers are already
                deploying the HCR1 technology, and the 2018 Camry has already put most
                of the HCR2 technologies into production, the agencies did not allow
                any application of HCR2 by 2025.'' \921\ ICCT concluded that the ``only
                explanations . . . for the agencies' system of omissions and
                constraints are that the agencies have biased the analysis against
                including all the viable technologies by inserting their own artificial
                constraints (either for lack of research, lack of analytical effort, or
                not fully utilizing all the agencies' best analytical tools and data)
                or that the auto industry is providing information that erroneously
                suggests their innovation is far less than what is demonstrated both
                above and in the agencies' own previous analyses.'' ICCT stated that
                ``[t]he great lengths the agencies have gone to artificially impose
                `skip' constraints for HCR in the CAFE modeling system demonstrates
                that the agencies have exerted an explicable and apparently deliberate
                bias towards forcing most of the automaker compliance technology toward
                higher cost, non-HCR turbocharging paths.'' \922\
                ---------------------------------------------------------------------------
                 \920\ NHTSA-2018-0067-11741.
                 \921\ NHTSA-2018-0067-11741.
                 \922\ NHTSA-2018-0067-11741.
                ---------------------------------------------------------------------------
                 Several commenters also stated that HCR should not have been
                restricted because it is clearly a cost-effective technology, citing
                the sensitivity runs conducted that allowed unrestricted HCR
                application in the analysis. For example, ICCT commented that allowing
                HCR2 application across the fleet reduced total per-vehicle cost of
                compliance with the augural standards by $690, which ``shows that the
                agencies intentionally excluded a highly cost-effective technology (by
                their own analysis) in the rulemaking analysis.'' \923\ Similarly, EDF
                performed software modifications of the CAFE model, including allowing
                the use of both HCR1 and HCR2 technology for all manufacturers by MY
                2028. The analysis performed by EDF using their modified version of the
                CAFE model, showed reductions in the per-vehicle compliance cost
                projections by nearly $600.\924\
                ---------------------------------------------------------------------------
                 \923\ NHTSA-2018-0067-11741.
                 \924\ NHTSA-2018-0067-12108.
                ---------------------------------------------------------------------------
                 ICCT concluded that ``[t]he only reasonable and technically valid
                assumption is that HCR be allowed for application to all vehicle
                models' engine redesigns through all the model years of the compliance
                modeling analysis.'' \925\ ICCT stated that ``[f]or the agencies to
                constrain HCR technology for use by other automakers, they have a
                responsibility to demonstrate why each of the other automakers cannot
                adopt this known technology in their fleet.''
                ---------------------------------------------------------------------------
                 \925\ NHTSA-2018-0067-11741.
                ---------------------------------------------------------------------------
                 The agencies agree with commenters' observations about the results
                of the sensitivity runs performed as part of the NPRM analysis.
                However, the agencies also believe the adoption features for HCR1 and
                HCR2 were appropriate for the NPRM analysis. Had the agencies not
                applied adoption features in that way, the agencies would have shown
                unrealistic pathways for compliance for manufacturers that would have
                understated costs and overstated benefits of potential CAFE and
                CO2 standards.
                 The agencies disagree with commenters' statements that HCR has been
                widely available in the automotive market and that the HCR technology
                accordingly should not be limited in the CAFE model. For reasons
                discussed in the NPRM and explained in more detail in Section
                VI.C.1.c)(3), depending on vehicle type and use, Atkinson cycle
                operation may be enabled for low and moderate engine demand conditions,
                whereas Otto cycle operation may be needed for higher load conditions
                to meet performance needs, such as to move more passengers, cargo, or
                for towing. In addition, there may be issues on some platforms to
                package the larger exhaust manifolds needed to enable Atkinson
                operation, particularly with V6 and V8 engines. Manufacturers have
                applied Atkinson technologies in unique ways to meet the needs and
                capabilities of their vehicles to operate using the Atkinson and Otto
                cycles. The agencies agree with comments from stakeholders, including
                Toyota, who observed HCR technology is not suitable for all vehicle
                configurations, and may not meet performance requirements for high-load
                applications. As discussed earlier, the agencies believe the variation
                of technologies can be categorized into three different forms of
                Atkinson engine technologies for this analysis: (1) Atkinson engines,
                (2) Atkinson-mode engines, and (3) Atkinson-enabled engines using
                variable valve timing with late intake closing. Manufacturers typically
                apply one of these technologies and tune that technology for specific
                applications. Some commenters have consistently conflated the
                technologies and asserted the capabilities of all three types of
                Atkinson technologies can be represented by a single engine model. The
                agencies do not agree with stakeholder assertions that a single HCR
                engine map should be applied to every technology class or vehicle
                platform.
                 To reflect better the incremental effectiveness for a low-cost
                version of HCR technology, the agencies added the HCR0 engine for the
                analysis. The specification of this engine was provided in the NPRM
                PRIA as Eng22b. Using this engine improves the estimated incremental
                effectiveness because the incremental engine changes were directly
                specified for the modeling and are relative to the other engine
                technologies in the analysis.\926\ HCR0 is the first engine in the HCR
                path that a manufacturer could adopt. HCR0 represents technology that
                could incrementally be adopted to the VVT engine, increasing
                compression ratio and adding Atkinson cycle capability. The use of the
                HCR0 technology, applied in the final rule analysis, allowed the
                agencies to update HCR adoption features. Once a basic engine adopts
                HCR technology (i.e., HCR0 and HCR1 for the central analysis, or HCR2
                for a sensitivity case) the vehicle will not switch to a different
                engine technology path. For example, if a vehicle had adopted HCR or is
                equipped with HCR technology it is not allowed to adopt turbocharged
                engine technologies. The HCR0 technology appropriately captures the
                benefits of applying transitional Atkinson technologies to conventional
                basic engine technologies. The agencies note that VVT technology valve
                control has late intake valve closing under some operating conditions
                to take some advantage of Atkinson cycle-like operation; however, that
                operation is not as extensive as HCR technology and is not coupled with
                a higher
                [[Page 24427]]
                compression ratio as is the case for HCR technologies.
                ---------------------------------------------------------------------------
                 \926\ PRIA 6.3.2.2.21.20.2.1 IAV Engine 22b--High Compression
                Atkinson Cycle Engine at p. 307.
                ---------------------------------------------------------------------------
                 The agencies also allowed all 4-cylinder engines on the basic
                engine path to adopt HCR technology similar to turbocharged
                technologies. This allowed any small and midsize vehicles, including
                small and midsize SUVs, that had any combinations of basic engine path
                technologies to move to the HCR path. However, there are two exceptions
                to this feature, including: (1) When the vehicle is a pickup including
                both standard and performance class; and (2) when the base engine is
                shared with a pickup including both standard and performance class. The
                agencies discussed earlier in the non-HEV Atkinson section why HCR
                technology cannot be applied to all vehicle applications.
                 Finally, engines with advanced engine technology already in the
                baseline vehicle such as turbocharged engines are not allowed to adopt
                HCR technology. The agencies continue to believe this constraint is
                reasonable given the extensive capital resources and stranded capital
                that would be involved if a manufacturer who focused on and invested
                heavily in non-HCR advanced technologies were to abandon those
                technologies abruptly and switch to HCR technologies.\927\ For example,
                Ford has incorporated turbocharged engines across 75 percent to 80
                percent of their fleet in MY2017, and these engines are shared across
                multiple technology classes.\928\ The abovementioned modeling,
                limitation for this analysis assumes that manufacturers will not change
                advanced engine technology applied to a platform due to the high cost
                and lead time required for research and development, and for the
                development and implementation of new manufacturing plants and
                equipment to implement an entirely new powertrain in the rule making
                time frame. For further discussion see Section VI.B.1.
                ---------------------------------------------------------------------------
                 \927\ 83 FR 43038.
                 \928\ The 2018 EPA Automotive Trends Report figure 4.23. at
                p.68.
                ---------------------------------------------------------------------------
                 In response to ICCT's comment that agencies must discuss the
                reasoning for allowing and disallowing HCR technology for each
                individual manufacturer, these updated adoption features now allow more
                manufacturers to adopt HCR engine technology. The agencies no longer
                apply adoption features based on manufacturer, but now base them on
                individual platforms. The agencies believe a manufacturer that has
                already invested in advanced engine technologies for a specific
                platform would face very high costs and incur significant stranded
                capital to switch that platform to another advanced technology. And
                doing so would not be reasonable given the small incremental fuel
                economy improvement that would be gained, for example, for switching
                from advanced turbocharging to HCR technologies. Specifically,
                manufacturers that have invested in turbocharging technology for
                certain platforms, like Honda, Ford, and the German manufacturers,
                would incur unreasonable costs to switch to another advanced technology
                path. However, manufacturers that use turbo technology on one platform
                are not precluded from implementing HCR technology on another of its
                platforms. HCR adoption is still limited for all manufacturers based on
                vehicle performance requirements discussed earlier.
                (4) Advanced Cylinder Deactivation Technology
                 In the NPRM, any basic engine technology could adopt ADEAC.
                Commenters stated that the agencies restricted ADEAC technologies in
                the NPRM analysis to naturally aspirated engines.
                 ICCT provided a broad comment regarding the treatment of advanced
                technologies, including ADEAC, and criticized how the NPRM ``removed
                many technologies that are viable and being actively deployed by the
                auto industry.'' ICCT specifically criticized ``cases where viable
                technology combinations are disallowed'' such as ``turbocharging and
                cylinder deactivation (DEAC).'' \929\
                ---------------------------------------------------------------------------
                 \929\ NHTSA-2018-0067-11741 at p.6.
                ---------------------------------------------------------------------------
                 UCS also commented on how ADEAC technology was applied in the NPRM,
                stating ``While the agencies have acknowledged the existence of dynamic
                cylinder deactivation, they have not appropriately included it as an
                available technology, dramatically limiting its availability.'' UCS
                specifically disagreed with adoption features of the ADEC, noting the
                technology ``is restricted to naturally aspirated, low-compression
                ratio engines--it cannot be combined with turbocharged engines, high
                compression ratio engines, or variable compression ratio engines due to
                pathway exclusivity in the Volpe model.'' \930\ CARB and Meszler
                mirrored these concerns.\931\
                ---------------------------------------------------------------------------
                 \930\ NHTSA-2018-0067-12039 at p.4
                 \931\ NHTSA-2018-0067-12039 at p.4.
                ---------------------------------------------------------------------------
                 The agencies agreed with commenters and in response have allowed
                both naturally aspirated engines and turbocharged engines to adopt
                ADEAC in the final rule analysis. The new Advanced Turbocharging path
                includes TURBOD and TURBOAD, while naturally aspirated engines use the
                same ADEAC engine designation. There is some potential for this type of
                technology to improve fuel economy and reduce CO2 emissions,
                however, the technology provides diminishing returns if it is included
                with engine downsizing or other technologies that already reduce
                pumping losses. Accordingly, once a vehicle has adopted ADEAC, TURBOD,
                or TURBOAD, the agencies did not allow further adoption of other engine
                technologies that reduce pumping losses such as VCR and VTG.
                (5) Miller Cycle Engines
                 Miller cycle engine technologies (VTG and VTGe) are new for this
                final rule analysis, and VTG engines could be applied to any basic and
                turbocharged engine. Discussed earlier, the VTGe technology is enabled
                by the use of a 48V system that presents an improvement from
                traditional turbocharged engines, and accordingly VTGe could only be
                applied with a mild hybrid system.
                (6) Variable Compression Ratio Engines
                 In the NPRM analysis, variable compression ratio (VCR) technology
                was not available for adoption, but the engine map and specifications
                were provided for review. For this final rule analysis, VCR engines are
                included in the analysis and can be applied to basic and turbocharged
                engines, however the technology is limited to Nissan. VCR technology
                requires a complete redesign of the engine, and in MY2020, only two of
                Nissan's models had incorporated this technology. In addition, the
                technology showed lower fuel savings than expected.\932\ The agencies
                do not believe any other manufacturers will invest to develop and
                market this technology in their fleet in the rulemaking time frame.
                ---------------------------------------------------------------------------
                 \932\ VanderWerp, D. ``Why Nissan's Holy-Grail VC-T Engine
                Doesn't Achieve Better Fuel Economy,'' C/D Nov 1, 2018. Available at
                https://www.caranddriver.com/features/a24434937/nissan-new-vc-t-engine-fuel-economy/. Last accessed Dec. 19, 2019.
                ---------------------------------------------------------------------------
                (7) Diesel Engines
                 Diesel engine adoption and features have been carried from the NPRM
                analysis for this final rule analysis for ADSL and DSLI. Any basic
                engine technologies (VVT, VVL, SGDI, and DEAC) can adopt ADSL and DSLI
                engine technologies. New for the final rule analysis is the adoption of
                advanced cylinder deactivation for diesel engines (DSLIAD). Any basic
                engine and diesel engine can adopt this technology in the final rule
                analysis;
                [[Page 24428]]
                however, the agencies have applied a phase in cap and year for this
                technology at 34 percent and MY 2023, respectively. In the agencies'
                engineering judgement, the agencies have concluded that this is a
                rather complex and costly technology to adopt and think that it could
                take significant investment to develop. For more than a decade, diesel
                engine technologies have been used in less than one percent of the
                total light-duty fleet production,\933\ and the investment for this
                cylinder deactivation technologies may not be justifiable.
                ---------------------------------------------------------------------------
                 \933\ The 2018 EPA Automotive Trends Report Table 4.1 at p. 72.
                ---------------------------------------------------------------------------
                (8) Alternative Fuel Engines
                 Adoption features for alternative fueled compressed natural gas
                (CNG) engines have been carried over from the NPRM for this final rule
                analysis. Because CNG is considered an alternative fuel under EPCA/
                EISA, it cannot be adopted during the rulemaking timeframe for NHTSA's
                standard setting analysis. The EPA analysis was modeled separately in
                the CAFE model without such constraints.
                (9) Engine Lubrication and Friction Reduction
                 Finally, new for this analysis is the addition of EFR. The agencies
                allow EFR to apply to any engine technology except for DSLI and DSLIAD.
                DSLI and DSLIAD inherently have incorporated engine friction
                technologies from ADSL. In addition, friction reduction technologies
                that apply to gasoline engines cannot necessarily be applied to diesel
                engines due to the higher temperature and pressure operation in diesel
                engines.
                f) Engine Effectiveness Modeling and Effectiveness Values
                 Figure VI-20 below shows the effectivness estimates from all the
                vehicle types for the NPRM analysis using Autonomie full vehicle
                modeling and simulation.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.158
                 Roush commented that they had observed wide variations in estimated
                incremental effectiveness associated with individual technology
                packages between the 2016 Draft TAR and NPRM analysis.\934\
                ---------------------------------------------------------------------------
                 \934\ NHTSA-2018-0067-11984. Roush at p. 16.
                ---------------------------------------------------------------------------
                 The agencies agree that to predict potential incremental
                improvements in fuel efficiency accurately, it is extremely important
                to understand the nature of the improvements being sought by each
                increment (improved thermodynamics, reduced friction, reduced vehicle
                weight, etc.). The technology modeling
                [[Page 24429]]
                and large scale simulation used for the proposal and updated for the
                final rule does exactly that. In fact, the NPRM and final rule use
                these methods more expansively than any previous CAFE and
                CO2 rulemaking, including the 2016 Draft TAR and 2016 EPA
                Proposed Determination.
                 One commenter stated the effectiveness for ADEAC was overestimated
                for the NPRM, and that data from compliance shows much lower
                effectiveness. The agencies disagree with this comment, as it is
                invalid to compare effectiveness of full vehicle compliance data
                directly to the incremental effectiveness modeled for ADEAC. For
                reasons discussed in Section VI.B.3 data from full vehicle benchmarking
                cannot be used as a comparison for specific technology effectiveness.
                The effectiveness estimated for this technology is in line with test
                data, CBI, and engineering analysis.\935\
                ---------------------------------------------------------------------------
                 \935\ Boha, Stani. ``Benchmarking and Characterization of a Full
                Continuous Cylinder Deactivation System.'' EPA. April 10-12, 2018
                SAEA World Congress. https://www.epa.gov/sites/production/files/2018-10/documents/deact-sae-world-congress-bohac-2018-04.pdf last
                access Feb 12, 2020.
                ---------------------------------------------------------------------------
                 Engine effectiveness estimates remained the same for most
                technologies from the NPRM analysis, with the exception of some
                technologies that had characteristics updated, and the new added engine
                technologies. For the final rule analysis, the agencies used the same
                effectiveness values for ADEAC applied to naturally aspirated engines
                as in the NPRM, and incorporated estimated effectiveness values for
                TURBOAD to represent ADEAC on downsized turbocharged engines.
                 Other technology-specific comments and the agencies' responses are
                provided within the discussion of each technology throughout this
                section, as those comments tended to be predicated on issues
                surrounding the engine maps used to model technologies or technology-
                specific adoption features. For the final rule analysis, the technical
                merits of the substantive comments and any accompanying publications
                and information were carefully considered and discussed in the
                subsections where appropriate.
                 Figure VI-21 below shows the effectivness estimates from compact
                car and midsize car vehicle types for the final rule analysis using
                Autonomie full vehicle modeling and simulation.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.159
                g) Engine Costs
                 Discussed in the PRIA, the agencies spent millions of dollars
                sponsoring research to determine direct manufacturing costs (DMCs) for
                fuel saving technologies since the 2012 rule.\936\ Because a major
                objective of the studies was to consider costs in the rulemaking
                timeframe, the agencies believed that these costs were appropriate to
                use for the NPRM and final rule analysis. Table VI-47 below shows the
                DMC used for IC engine technologies for the NPRM analysis.
                ---------------------------------------------------------------------------
                 \936\ FEV prepared several cost analysis studies for EPA on
                subjects ranging from advanced 8-speed transmissions to belt
                alternator starter, or Start/Stop systems. NHTSA also contracted
                with Electricore, EDAG, and Southwest Research on teardown studies
                evaluating mass reduction and transmissions. The 2015 NAS report on
                fuel economy technologies for light-duty vehicles also evaluated the
                agencies' technology costs developed based on these teardown
                studies, and the technology costs used in this proposal were updated
                accordingly. These studies are discussed in detail in Chapter 6 of
                the RIA accompanying the NPRM proposal.
                ---------------------------------------------------------------------------
                [[Page 24430]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.160
                [[Page 24431]]
                 CARB commented that costs associated with IC engines were not
                excluded from the final costs of BEV vehicles.\937\ CARB continued,
                stating that ``the final costs of BEV vehicles are higher due to the
                inclusion of the base absolute costs, to which the assigned BEV
                incremental cost would be added.''
                ---------------------------------------------------------------------------
                 \937\ NHTSA-2018-0067-11873 at p.122.
                ---------------------------------------------------------------------------
                 The agencies agree with CARB that inclusion of IC engine costs in
                the BEV cost was an error in the analysis. In response to this comment,
                the agencies have developed absolute costs for baseline engines for the
                CAFE model in order to account for appropriate cost of removing engines
                from BEVs. In the final rule analysis, once a vehicle adopts BEV
                technology, the costs associated with powertrain systems are removed.
                Due to the extensive variations in engine technologies in real world
                production, the agencies relied on discrete publication costs and
                historical studies to assign costs for base engines.938 939
                For this final rule analysis, the agencies have included these costs
                for base engines shown in Table VI-48.
                ---------------------------------------------------------------------------
                 \938\ FEV P311732-02 Oct13, 2015 at p. 259.
                 \939\ UBS Limited. ``UBS Evidence Lab Electric Car Teardown--
                Disruption ahead?'' May 18, 2017.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.161
                BILLING CODE 4910-59-C
                 Commenters compared engine cost data from the NPRM to other
                sources, in many cases to support their comments that the technology
                costs used in the NPRM were too high. ICCT commented that the agencies
                did not consider the latest reports on technology cost data, and
                specifically referenced an ICCT-sponsored FEV cost study for the
                European EU6b regulations in MY 2025,\940\ as well as prior EPA cost
                estimates for several engine technologies including SGDI, cEGR, HCR,
                and others, to point out differences in cost.\941\ ICCT also commented
                on the difficulty they had in locating the cost data used in the NPRM,
                stating that ``because the agencies present cost data in so many
                different ways in dozens of different places in the NPRM, impact
                assessment, and supporting data files, the precise agencies' costs are
                obscured and not transparent.'' ICCT stated that ``[w]ithout a clear
                explanation of the methodology, it is unclear precisely how price
                increases are determined, as well as the relationship between
                technology costs, fines, and price increases.'' Despite this claim,
                ICCT was able to provide several pages comparing engine technology
                costs.
                ---------------------------------------------------------------------------
                 \940\ FEV. '' 2025 Passenger Car and Light Commercial Vehicle
                Powertrain Technology Analysis'' September 2015. https://theicct.org/sites/default/files/publications/PV-LCV-Powertrain-Tech-Analysis_FEV-ICCT_2015.pdf.
                 \941\ NHTSA-2018-0067-11741 at p. I-68.
                ---------------------------------------------------------------------------
                 In the NPRM PRIA Chapter 6.3.2.2.20.22, the agencies provided DMCs
                for all engine technologies in 2016 dollars without inclusion of RPE
                and learning for review. In the same chapter, the agencies also
                provided absolute costs that incorporated costs in 2016 dollars, RPE
                and learning data as used by the CAFE model to assess cost
                effectiveness for future MY vehicles. Where appropriate, the agencies
                discussed in the individual technology sections where costs were
                updated for this final rule analysis with the latest data. This also
                includes cost data for new technologies available in the CAFE model for
                the final rule analysis.
                 Some engine costs were carried over from prior rulemakings, but may
                have looked different because they were updated to current dollars
                (2016 for the NPRM and 2018 for the final rule), and for engine
                architecture and cylinder count. In addition, costs were updated based
                on appropriate vehicle class. This was important to consider to
                maintain performance neutrality, as technology effectiveness associated
                with one engine technology type for a vehicle class cannot be used for
                the same engine technology for higher performance vehicle class. This
                affected total costs. For further discussion on the cost-effectiveness
                metric used in the CAFE model, see discussions in the Section VI.A
                Overview of the CAFE model and VI.B.3 Technology Effectiveness Values.
                 The agencies do not believe that the FEV report referenced by ICCT
                is applicable for this analysis for a few reasons. First, the primary
                focus of the FEV study ``is the European Market according to the EU6b
                regulation as well as the consideration of emissions under both the
                NEDC and WLTP test procedures.'' This final rule analysis specifically
                considered the U.S. automotive market during the rulemaking timeframe
                based on U.S.-specific regulatory test cycles. Accordingly, the costs
                reflect incremental technology effectiveness for achieving improvements
                as measured through U.S. regulatory test methods. The agencies had
                discussed these test cycles and methods further in Section VI.B.3
                Technology Effectiveness Values.
                 Second, FEV did not conduct original teardown studies for this
                report, as indicated by project tasks, but rather used engineering
                judgement and external studies in assessing incremental costs.\942\ The
                FEV report did not provide sources for each individual cost and it is
                unclear how costs in many scenarios were developed since no teardowns
                were used. Note that for this final rule analysis, the agencies have
                used previously conducted FEV cost teardown studies and the referenced
                2015 NAS costs that referenced FEV teardowns. The agencies are not
                concluding that FEV is an unreliable source. The agencies preferred to
                specifically identify incremental costs of adding technology to account
                appropriately for the costs of those technologies in the analysis.
                ---------------------------------------------------------------------------
                 \942\ FEV EU Costs Tasks: ``Definition of reference hardware or
                description made by experience of development and design engineers
                as well as additional research as base for cost analysis (no
                purchase of hardware)''.
                ---------------------------------------------------------------------------
                 Finally, the cost for different vehicle classes identified by the
                FEV study does not line up with the vehicle classes discussed in the
                NPRM and this final rule analysis. FEV stated specifically, ``the
                configuration of the vehicles has not been optimized for the US market
                and may not be representative of this market.''\943\ The agencies have
                discussed the importance of aligning the CAFE vehicle models with the
                U.S. market earlier in Section VI.B.3
                [[Page 24432]]
                Technology Effectiveness Values and Section VI.C.1.d) Baseline Fleet.
                All of these factors make it difficult to compare directly the
                agencies' estimates and estimates presented in the FEV report cited by
                ICCT in their comments.
                ---------------------------------------------------------------------------
                 \943\ Id. at p.141.
                ---------------------------------------------------------------------------
                 HDS provided a variety of costs and effectiveness comparisons
                between the NPRM and previous 2012 final rule and the 2016 Draft
                TAR.\944\ Specifically, HDS stated that the data presented in the 2016
                TAR indicated a $60 per CO2/mile reduction for most
                conventional engine technologies.
                ---------------------------------------------------------------------------
                 \944\ Duleep, K.G., ``Review of the Technology Costs and
                Effectiveness Utilizing in the Proposed SAFE Rule,'' Final Report,
                H-D Systems, October 2018, at p. 18-19.
                ---------------------------------------------------------------------------
                 Although the comparison was technically sound, there are
                significant differences between the Draft TAR and NPRM analyses that
                clearly account for the differences in engine cost. First, the NPRM
                analysis used the MY 2016 fleet as a starting point to model
                manufacturers' potential responses to CAFE and CO2
                standards, whereas the 2012 final rule and Draft TAR used older
                baseline fleets. Vehicles in the MY 2016 fleet already included more
                advanced technologies than their predecessors in prior MY fleets, which
                would make it more expensive for vehicles that have already adopted
                advanced technologies to adopt more advanced technology. Second, the
                agencies refined the engine modeling from previous analysis to the NPRM
                to account for engine configurations and cylinder count more precisely.
                For the final rule analysis, the same approach was taken to account
                appropriately for costs for different type engine designs and
                configurations.
                 Aside from these updates, engine costs were carried over from the
                NPRM analysis, except for newly added technologies, where costs were
                obtained from various sources such as NAS studies, technical
                publications, and CBI data. Finally, the cost estimates have been
                updated to account for dollar year (updated from 2016 dollars to 2018
                dollars), and learning rate.
                (1) Basic Engines
                 DMCs used for the final rule analysis for basic engine technologies
                were the same as NPRM costs. Table VI-49 below shows the basic engine
                DMC used for this final rule analysis.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.162
                [GRAPHIC] [TIFF OMITTED] TR30AP20.163
                [[Page 24433]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.164
                BILLING CODE 4910-59-C
                (2) Turbocharged Downsized Engines
                 DMCs used for the final rule analysis for the turbocharged engine
                technologies were the same as NPRM costs. When these technologies are
                applied to V6 and V8 non-turbocharged engines, the incremental I4 and
                V6 turbocharged costs are applied, respectively. Table VI-52 below
                shows the DMC used for turbochared technologies for FRM analysis in
                2018 dollars.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.165
                [GRAPHIC] [TIFF OMITTED] TR30AP20.166
                [[Page 24434]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.167
                BILLING CODE 4910-59-C
                (3) Non-HEV Atkinson and Atkinson Engines
                 DMCs used for the final rule analysis for HCR0 and HCR1 were based
                on HCR1 and HCR2 from NPRM, respectively. Discussed in Section
                VI.C.1.c).(3), the agencies aligned the cost of HCR technologies to
                align with 2015 NAS effectiveness and costs.
                 Stakeholders commented on the costs of HCR technology compared to
                previous analysis. ICCT compared the NPRM costs to EPA's Proposed
                Determination costs, stating that ``[t]his is a clear case where the
                agencies appear to have not used the best available data from EPA which
                has extensively analyzed this technology and its associated cost, nor
                have the agencies justified how they have increased the associated
                costs, apparently by a factor of three.'' Similarly, Roush Industries
                commenting on behalf of CARB stated that the costs for implementing HCR
                technology were 5-6 times the 2016 Draft TAR estimated costs, which are
                ``extremely high'' and ``will significantly overstate the incremental
                cost and bias technology pathways.''\945\ HDS also commented that the
                costs for HCR technology were higher than the costs from the 2016 Draft
                TAR, and speculated that was due to ``the bulky exhaust system used in
                the Mazda ATK1 engine, which apart from being expensive also requires
                the vehicle to be modified to accommodate the exhaust system.''\946\
                HDS cited the 2018 Camry as an example of a vehicle that does not use
                the same exhaust system, but stated the sources of the new cost data
                were not documented in the PRIA. ICCT stated that ``[t]he agencies
                should reinstate the better justified and more deeply analyzed original
                Proposed Determination HCR cost numbers from EPA for this rulemaking.''
                ---------------------------------------------------------------------------
                 \945\ NHTSA-2018-0067-11984.
                 \946\ NHTSA-2018-0067-11985.
                ---------------------------------------------------------------------------
                 The NPRM analysis and the final rule analysis used the same DMCs
                established by the 2015 NAS report for the Atkinson cycle technologies.
                However, because there are many various engine configurations in the
                market, the agencies do not use the same fixed costs that were set for
                each type of vehicle described in the 2015 NAS report, such as pickup
                and sedan. The agencies have expanded costs by taking into account the
                type of technology in the baseline, like SGDI, and the configuration of
                the engine, such as SOHC versus DOHC. In addition, the cost used in the
                NPRM also included updated dollar year, learning rate, and RPE.
                Although EPA also used costs from the 2015 NAS report for the Proposed
                Determination analysis, they used a different approach to account for
                components.\947\ For the final rule analysis the agencies continued to
                use the same DMC for HCR technologies. Table VI-55 below shows HCR DMCs
                used for the final rule analysis in 2018 dollars.
                ---------------------------------------------------------------------------
                 \947\ EPA PD TSD at 2-307 to 2-308 ``Note that the NAS costs
                include the costs of gasoline direct injection (shown as ``DI'' in
                the NAS report row header). EPA has removed those costs (using the
                NAS reported values) since EPA accounts for those costs separately
                rather than including them in the Atkinson-2 costs. Note also that
                EPA always includes costs for direct injection, along with variable
                valve timing and other costs, when building an Atkinson-2 package.''
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.168
                [[Page 24435]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.169
                [GRAPHIC] [TIFF OMITTED] TR30AP20.170
                [[Page 24436]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.171
                [[Page 24437]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.172
                [[Page 24438]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.173
                [[Page 24439]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.174
                [[Page 24440]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.175
                BILLING CODE 4910-59-C
                (4) Advanced Cylinder Deactivation Technologies
                 DMCs used for the final rule analysis for the advanced cylinder
                deactivation technologies were the same as NPRM costs.
                 Roush commented that in the NPRM analysis, the agencies did not
                properly
                [[Page 24441]]
                consider the ``very cost-effective benefits of skip-fire technology,''
                referred to in the analysis as ADEAC. Roush stated that ``due to
                extremely high estimated cost ($1,250.00 in MY2016), the benefits of
                this technology will likely not be chosen in any reasonable technology
                pathway. If included, the predicted cost for that pathway will be
                overestimated by $750-$1,000.''\948\ Similarly, Meszler commented on
                the cost for the ADEAC system stating ``advanced cylinder deactivation
                paths are assumed (by NHTSA) to be expensive, and are selected only in
                rare instances.'' \949\ ICCT also stated ``The agencies estimated a
                greatly exaggerated cost of advanced cylinder deactivation for that
                level of the technology.'' \950\
                ---------------------------------------------------------------------------
                 \948\ Roush at p.13.
                 \949\ Meszler Comments, Attachment 2, NHTSA Docket No. NHTSA-
                2018-0067-11723.
                 \950\ ICCT comments, NHTSA-2018-0067-11741, Page I-71.
                ---------------------------------------------------------------------------
                 The agencies do not agree with the commenter's statement that the
                analysis did not consider ADEAC as a cost effective technology or that
                the agencies overestimated costs for the technology. The agencies
                considered the most up to date information and data for the NPRM and
                final rule analysis.\951\ The agencies rely on the CAFE model to
                determine technology cost effectiveness, and if the technology was cost
                effective for a manufacturer to adopt, then the model would apply it to
                a manufacturer's vehicle. The adoption of ADEAC was applied to vehicles
                with corresponding technology combinations to reflect appropriate cost
                and effectiveness, as discussed in the paragraph above. The purpose of
                ADEAC is to reduce pumping losses, but if the engine has been
                downsized, or has already incorporated technologies that also reduce
                pumping loss, then it is likely the ADEAC has reached a point of
                diminishing return. As far as the agencies are aware, Roush did not
                provide alternative DMCs for ADEAC technology. Table VI-58 below shows
                the examples of advanced cylinder deactivation DMC used for both
                naturally aspirated and turbocharged engines for the final rule
                analysis in 2018$.
                ---------------------------------------------------------------------------
                 \951\ Boha, Stani. ``Benchmarking and Characterization of a Full
                Continuous Cylinder Deactivation System.'' EPA. April 10-12, 2018
                SAEA World Congress. https://www.epa.gov/sites/production/files/2018-10/documents/deact-sae-world-congress-bohac-2018-04.pdf. (last
                accessed Feb 12, 2020).
                 CARB. ``Tula Technology's Dynamic Skip Fire.'' September 28,
                2016. CARB_2016 Tula ppt skipfire_NHTSA-2018-0067-11985.pdf
                ---------------------------------------------------------------------------
                (5) Miller Cycle Engines
                 The agencies estimated costs for Miller cycle engines with VTG from
                2016 ICCT-sponsored FEV technology cost assessment report. The agencies
                considered costs from 2015 NAS study that referenced a NESCCAF 2004
                report,952 953 but believed that the reference material from
                the ICCT report had more updated cost estimates for this technology
                that represented what was discussed in the NPRM and modeled in the
                final rule analysis.
                ---------------------------------------------------------------------------
                 \952\ ``Reducing Greenhouse Gas Emissions from Light-Duty Motor
                Vehicles.'' NESCCAF. September 23, 2004 Report. Available at https://www.nesccaf.org/documents/rpt040923ghglightduty.pdf/. Last accessed
                Dec. 22, 2019.
                 \953\ ``VGT gasoline turbo, charge air cooler, piston upgrade,
                piston cooling, steel crankshaft, cooling system upsize, plumbing,
                rings, pressure sensor & bearing upgrade. Excludes any needed
                increase in transmission torque capacity or modifications to
                aftertreatment system.'' NESCCAF Report comment (2004).
                ---------------------------------------------------------------------------
                 NAS estimated the incremental cost for VTG as $525 in 2010$, but
                this cost assumes many of the traditional turbocharged components and
                adds VVT, VVL and SGDI. In addition, VTG (Eng23b) and VTGe (Eng23c)
                engines both have similar modeled BMEP levels and a cooled EGR system
                to CEGR1 (Eng14), implying that the components such as cooling systems
                and piping will have similar costs.
                 The NAS template to calculating the final DMCs for the Miller cycle
                engines for the different engine configuration is the $525 (2010$) plus
                cost of cEGR1 minus cost of VVT, VVL, and SGDI. The agencies estimated
                the cost for electrically-assisted variable supercharger VTGe (Eng23c)
                engines based on the 2015 NAS study that uses a cost of $1050 (2010$)
                plus the cost of the mild hybrid battery. For the final rule analysis,
                the total costs for these technologies are shown below.
                BILLING CODE 4910-59-P
                [[Page 24442]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.176
                [[Page 24443]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.177
                [[Page 24444]]
                BILLING CODE 4910-59-C
                (6) Variable Compression Ratio Engines
                 DMCs used for the final rule analysis for the VCR engines were
                based on the 2015 NAS report.\954\ The 2015 NAS reported cost for VCR
                in MY2025 used a naturally aspirated engine; however, for this final
                rule analysis the agencies have added cEGR and other engine
                technologies to the engine. Total costs were updated to reflect 2018
                dollars and MY2017 learning rate which is based on the NPRM ADEAC
                learning rate. Table VI-67 below shows examples of VCR DMCs used for
                this this final rule analysis in 2018 dollars.
                ---------------------------------------------------------------------------
                 \954\ 2015 NAS at p. 93.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24445]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.178
                [[Page 24446]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.179
                [[Page 24447]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.180
                BILLING CODE 4910-59-C
                (7) Diesel Engines
                 DMCs used for the final rule analysis for diesel engine
                technologies were the same as the NPRM analysis. For DSLIAD
                technologies, the agencies have added the incremental cost of ADEAC to
                DSLI.
                [[Page 24448]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.181
                [GRAPHIC] [TIFF OMITTED] TR30AP20.182
                (8) Alternative Fuel Engines
                 DMCs used for the final rule analysis for CNG engine technologies
                were the same as the NPRM analysis.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.183
                [GRAPHIC] [TIFF OMITTED] TR30AP20.184
                [[Page 24449]]
                (9) Engine Lubrication and Friction Reduction Technologies
                 EFR costs used for the final rule analysis are based on the 2015
                NAS assessment for low friction lubrication and engine friction
                reduction level 2 (LUB2_EFR2). The 2015 NAS report provided estimates
                of $51 (I4 DOHC), and $72 (V6 SOHC and DOHC) for midsize cars, in 2015
                dollars, relative to level 1 engine friction reduction (EFR1), which
                costs about $12 per cylinder. For this analysis, EFR technologies DMCs
                are estimated to be $14.05 per cylinder in 2016 dollars. Total costs
                were updated to reflect 2018 dollars and MY 2017 learning rate. Table
                VI-74 shows the EFR DMC used for the final rule analysis in 2018
                dollars.
                [[Page 24450]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.185
                [[Page 24451]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.186
                [[Page 24452]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.187
                BILLING CODE 4910-59-C
                [[Page 24453]]
                2. Transmission Paths
                 Transmissions transmit torque from the engine to the wheels.
                Transmissions primarily use two mechanisms to improve fuel efficiency:
                (1) A higher gear count, as more gears allow the engine to operate
                longer at higher efficiency speed-load points; and (2) improvements in
                friction or shifting efficiency (e.g., improved gears, bearings, seals,
                and other components), which reduce parasitic losses.
                 There are two major categories of transmission types modeled in the
                analysis: Automatic and manual. Automatic transmissions automatically
                select and shift between transmission gears for the driver during
                vehicle operation. The automatic transmission category is further
                subdivided into four subcategories: Traditional automatic
                transmissions, dual clutch transmissions, continuously variable
                transmissions, and direct drive transmissions. Manual transmissions
                require direct control by the driver to select and shift between gears
                during vehicle operation.
                 Conventional planetary gear automatic transmissions (AT) are the
                most popular transmission.\955\ ATs typically contain three or four
                planetary gear sets that provide the various gear ratios. Gear ratios
                are selected by activating solenoids which engage or release multiple
                clutches and brakes as needed. ATs with gear counts ranging from five
                speeds to ten speeds were considered in the NPRM and final rule
                analysis.\956\
                ---------------------------------------------------------------------------
                 \955\ ``The 2018 EPA Automotive Trends Report,'' https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends, Accessed Aug 23, 2019.
                 \956\ Specifically, the agencies considered five-speed automatic
                transmissions (AT5), six-speed automatic transmissions (AT6), seven-
                speed automatic transmission (AT7), eight-speed automatic
                transmissions (AT8), nine-speed automatic transmissions (AT9), and
                ten-speed automatic transmissions (AT10).
                ---------------------------------------------------------------------------
                 ATs are packaged with torque converters, which provide a fluid
                coupling between the engine and the driveline, and provide a
                significant increase in launch torque. When transmitting torque through
                this fluid coupling, energy is lost due to the churning fluid. These
                losses can be eliminated by engaging the torque convertor clutch to
                directly connect the engine and transmission (``lockup'').
                 Conventional continuously variable transmissions (CVT) consist of
                two cone-shaped pulleys, connected with a belt or chain. Moving the
                pulley halves allows the belt to ride inward or outward radially on
                each pulley, effectively changing the speed ratio between the pulleys.
                This ratio change is smooth and continuous, unlike the step changes of
                other transmission varieties. CVTs were not initially chosen in the
                fleet modeling for the 2012 rulemaking analysis for MYs 2017 and later
                because of the predicted low effectiveness associated with CVTs (due to
                the high internal losses and narrow ratio spans of CVTs in the fleet at
                that time).\957\ However, improvements in CVTs in the current fleet
                have increased their effectiveness, leading to increased adoption rates
                in the fleet. In its 2015 report, the NAS recommended CVTs be added to
                the list of considered technologies. The agencies included CVT
                technology for the NPRM and this final rule analyses.
                ---------------------------------------------------------------------------
                 \957\ Morihiro, S., ``Fuel Economy Improvement by
                Transmission,'' presented at the CTI Symposium 8th International
                2014 Automotive Transmissions, HEV and EV Drives.
                ---------------------------------------------------------------------------
                 Dual clutch transmissions (DCT), like automatic transmissions,
                automate shift and launch functions. DCTs use separate clutches for
                even-numbered and odd-numbered gears, allowing the next gear needed to
                be pre-selected, resulting in faster shifting. The use of multiple
                clutches in place of a torque converter result in lower parasitic
                losses than ATs. However, DCTs are seeing limited penetration in the
                fleet, and because of the low penetration rate, only two DCTs were
                considered in the analysis.
                 Direct drive (DD) transmissions are a direct connection between the
                wheels and a drive motor. In a DD transmission, the ratio between wheel
                speed and motor speed remains constant. A DD transmission is only used
                in battery electric vehicles, and in the NPRM the agencies provided the
                specification for comments.\958\
                ---------------------------------------------------------------------------
                 \958\ NHTSA-2018-0067-0003. ANL Autonomie Summary of Main
                Component Assumptions. Aug 21, 2018. NHTSA-2018-0067-0007. Islam, E.
                S, Moawad, A., Kim, N, Rousseau, A. ``A Detailed Vehicle Simulation
                Process To Support CAFE Standards 04262018--Report'' ANL Autonomie
                Documentation. Aug 21, 2018.Aug 21, 2018 NHTSA-2018-0067-0004. ANL
                Autonomie Data Dictionary. Aug 21, 2018.
                ---------------------------------------------------------------------------
                 Manual transmissions (MT) are transmissions that require direct
                control by the driver to operate the clutch and shift between gears.
                Manual transmissions have seen a significant reduction in application
                by automakers over recent years. As a result of the reduced market
                presence, only three variants are used in the analysis.
                a) Transmission Modeling in the CAFE Model
                 The NPRM analysis modeled pathways for applying improved technology
                for each of the transmission categories and subcategories, except for
                the direct drive, which was only available in the battery electric
                vehicles. The MT and DCT pathways only included increasing gear counts
                (e.g., 5-speed manual transmission, 6-speed manual transmission, and 7-
                speed manual transmission) as improved technologies.
                 The traditional ATs and CVTs included both increased gear counts
                and high efficiency gearbox (HEG) technology improvements as options.
                HEG improvements for transmissions represent incremental advancement in
                technology that improves efficiency, such as: Reduced friction seals,
                bearings and clutches, super finishing of gearbox parts, and improved
                lubrication. All these advancements are aimed at reducing frictional
                and other parasitic loads in transmissions to improve efficiency. Three
                levels of HEG improvements are considered in this analysis, based on
                2015 NAS recommendations and based on CBI data.\959\ HEG efficiency
                improvements were applied to ATs and CVTs, as those transmissions
                inherently have higher friction and parasitic loads related to
                hydraulic control systems and greater component complexity, compared to
                MTs and DCTs.
                ---------------------------------------------------------------------------
                 \959\ 2015 NAS Report, at 191.
                ---------------------------------------------------------------------------
                 In total, 18 unique transmission technology combinations were
                simulated, using explicit input values for gear ratios, gear
                efficiencies, gear spans, shift logic, and transmission
                architecture.\960\ \961\ Table VI-77 shows a list of the multi-gear
                transmissions used for the NPRM.\962\
                ---------------------------------------------------------------------------
                 \960\ See PRIA Chapter 6.3.
                 \961\ Ehsan, I.S., Moawad, A., Kim, N., & Rousseau, A., ``A
                Detailed Vehicle Simulation Process To Support CAFE Standards.''
                ANL/ESD-18/6. Energy Systems Division, Argonne National Laboratory.
                2018.
                 \962\ The NPRM and final rule also included a direct drive
                transmission (single ratio) for BEVs.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24454]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.188
                 The technologies that made up the four transmission/level paths
                defined by the modeling system for the NPRM analysis are shown in
                Figure VI-22. Each vehicle model in the analysis fleet is assigned an
                initial transmission type and level that most closely matches its
                configuration and characteristics. The baseline-level technologies
                (AT5, MT5 and CVT) appear in gray boxes and are only used to represent
                the initial configuration of a vehicle's transmission in the analysis
                fleet. Because there are only a few manual transmissions with less than
                five forward gears in the analysis fleet, for simplicity, all manual
                transmissions with five forward gears or fewer were designated MT5 for
                the analysis. Similarly, all automatic transmissions with five forward
                gears or fewer have been assigned the AT5 technology. For the NPRM
                analysis, the agencies included a 7-speed automatic and a 9-speed
                automatic to account for effectiveness of those transmissions in the
                analysis fleet. These two transmissions were not available for adoption
                but were available as initial configurations, and appear in gray boxes
                in Figure VI-22.
                [[Page 24455]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.189
                BILLING CODE 4910-59-C
                 The model generally may apply any of the more efficient
                transmission technologies that are contained within the pathway of the
                baseline vehicle initial transmission configuration. The model
                prohibits manual transmissions from becoming automatic transmissions.
                Automatic transmissions may become CVT level 2 after progressing though
                the 6-speed automatic, as shown in Figure VI-22. While the structure of
                the model could allow automatic transmissions to consider applying a
                DCT, the market data file was used to preclude the application of DCTs
                to automatic transmission vehicles, as discussed
                [[Page 24456]]
                further in Section VI.C.2.c) Transmission Adoption Features, below.
                 The model does not attempt to simulate ``reversion'' to less
                advanced transmission technologies, such as replacing a 6-speed AT with
                a DCT and then replacing that DCT with a 10-speed AT. The agencies
                invited comment on whether the model should be modified to simulate
                ``reversion'' and, if so, how this possible behavior might be
                practicably simulated. Richard Rykowski, supporting comments from the
                Environmental Defense Fund (EDF), broadly discussed the concept of
                reversion in the CAFE model, and included an example relating to the
                transmission technology paths.\963\ Mr. Rykowski stated that it is
                ``possible that the model could add a 10-speed transmission to a
                vehicle with a very basic engine'' and then as the simulation
                progressed and ``the manufacturer required greater fuel or
                CO2 emission control, the Volpe Model might move to a TURBO1
                or HCR engine'' and the vehicle would no longer need the 10-speed
                transmission to meet standards, and a 6-speed or 8-speed transmission
                might be more cost effective.
                ---------------------------------------------------------------------------
                 \963\ Comments from Environmental Defense Fund, Attachment B,
                NPRM Docket No. NHTSA-2018-0067-12108, at 70.
                ---------------------------------------------------------------------------
                 The scenario discussed by Mr. Rykowski is very unlikely. The CAFE
                model cost optimization algorithm considers both current and future
                standard requirements when selecting current MY technologies. The
                algorithm will look multiple years into the future and compare multiple
                potential technology paths going forward for the most cost-effective
                path. For a more detailed discussion on the cost optimization algorithm
                see Section VI.A.4, Compliance Simulation.
                 Regarding the types of transmission technologies modeled, Meszler
                Engineering Services provided a comment criticizing the limited number
                of manual transmission model options and the limited technology paths
                available to vehicles with manual transmissions.\964\ The agencies do
                not agree with Meszler Engineering Service's assessment. The manual
                transmission path includes three model options and allows for the
                vehicles to receive electrification in the form of SS12V and BISG
                technologies. The agencies believe the technology paths dedicated to
                manual transmission was appropriate for vehicles that typically
                represent manufacturers' specialty performance cars, such as the Subaru
                STI or BMW M-series, that comprise an overall fleet share of less than
                2 percent.
                ---------------------------------------------------------------------------
                 \964\ Comments from Meszler Engineering Services,
                Attachment2_CAFE Model Tech Issues, Docket No. NHTSA-2018-0067-
                11723, at 33.
                ---------------------------------------------------------------------------
                 Commenters also discussed potential missing transmission
                technologies in the NPRM analysis. ICCT stated that the agencies failed
                to consider transmission warm-up technologies, which are available in
                3.7 million new vehicles in the MY 2016 fleet, that are being deployed
                due to regulatory test-cycle benefits and off-cycle credits.\965\ In
                addition, the Fiat Chrysler Automobiles (FCA) also expressed concern
                over the lack of inclusion of thermal bypass devices in the modeling of
                transmission technologies.\966\
                ---------------------------------------------------------------------------
                 \965\ Comments from ICCT, NPRM Docket No. NHTSA-2018-0067-11741
                full comments, at I-28.
                 \966\ Comments from Fiat Chrysler Automobiles, Attachment 1,
                NPRM Docket No. NHTSA-2018-0067-11943, at 97.
                ---------------------------------------------------------------------------
                 The agencies agree with parts of ICCT's and the FCAs comments and
                disagree with other parts. The agencies do agree with ICCT and the Auto
                Alliance that the analysis should consider the off-cycle benefits of
                transmission warm-up technology. For the final rule analysis, the
                agencies applied off-cycle technologies in the CAFE model. For the
                final rule analysis, the agencies applied off-cycle technologies at the
                maximum menu regulatory value of 10 g/mile for all manufacturers by MY
                2023. The modeled adoption included benefits of transmission warm-up as
                a menu item. The modeling of off-cycle technologies is further
                discussed in Section VI.C.8. The agencies disagree with ICCT and the
                Auto Alliance comments that transmission warm-up technologies were not
                included in the NPRM on-cycle analysis. For the NPRM, and for the final
                rule, the HEG level 2 technology package includes rapid transmission
                oil warm-up technology.\967\ The inclusion of the HEG2 technology
                package in AT and CVT models accounts for impacts of this technology to
                performance on the standard test-cycle.
                ---------------------------------------------------------------------------
                 \967\ 2015 NAS Report, at 191.
                ---------------------------------------------------------------------------
                 For the final rule analysis the transmission model paths are shown
                in Figure VI-23. For the final rule analysis, the baseline-only
                technologies (MT5, AT5, AT7L2, AT9L2, and CVT) are grayed and are only
                used to signify initial vehicle transmission configurations. For
                simplicity, all manual transmissions with five forward gears or fewer
                are assigned the MT5 technology in the analysis fleet. Similarly, all
                automatic transmissions with five forward gears or fewer are assigned
                the AT5 technology.
                BILLING CODE 4910-59-P
                [[Page 24457]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.190
                 Since the Manual Transmission path terminates with MT7, the system
                assumes that all manual transmissions with seven or more gears are
                mapped to the MT7 technology. Moreover, all dual-clutch (DCT) or auto-
                manual (AMT) transmissions with five or six forward gears are mapped to
                the DCT6 technology, and all DCTs or AMTs with seven or more forward
                gears are mapped to DCT8.
                 For the final rule analysis, the naming convention for the
                transmission technology models was updated to identify better the
                technologies represented in each transmission. Although the
                technologies in each transmission configuration were described in the
                NPRM, there appears to have been confusion among some commenters about
                the technology content of some transmission configurations. Some
                commenters compared the NPRM AT10 to the NPRM AT8, and commented on
                unexpected differences in effectiveness relative to the differences in
                transmission gear count.\968\ For the given example, the NPRM AT8
                represented a baseline 8-speed automatic transmission, with level 1 HEG
                technology applied, and the NPRM AT10 represented a 10-speed automatic
                transmission with level 2 HEG technology applied. A direct comparison
                of gear count would occur by comparing the NPRM AT8L2 to the NPRM AT10.
                The updated naming convention identifies the transmission technology
                type, gear count and HEG technology level. Table VI-78 shows the final
                rule names for transmission models compared to the names used for the
                NPRM analysis.
                ---------------------------------------------------------------------------
                 \968\ Comments from CARB, Attachment 2018-10-26 FINAL CARB
                Detailed Comments on SAFE, NPRM Docket No. NHTSA-2018-0067 at 110-
                13.
                ---------------------------------------------------------------------------
                [[Page 24458]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.191
                BILLING CODE 4910-59-C
                b) Transmission Analysis Fleet Assignments
                 The agencies discussed in the NPRM the process for developing the
                2016 analysis fleet, including how the agencies weighed using
                confidential business information versus publicly-releasable sources,
                the use of compliance data, and decision to use a 2016 analysis fleet
                over other alternatives.\969\ As discussed above, this final rule
                analysis used the 2017 vehicle fleet as the analysis fleet input, and
                the agencies followed largely the same process for assigning initial
                transmission assignments as in the NPRM.
                ---------------------------------------------------------------------------
                 \969\ 83 FR 43003.
                ---------------------------------------------------------------------------
                 For the 2017 analysis fleet, transmission data was gathered from
                the manufacturer final model year CAFE compliance submissions to the
                agencies as well as manufacturer press releases. The data for each
                manufacturer was used to determine which platforms shared transmissions
                and to establish the leader-follower relationships between vehicles.
                Within each manufacturer fleet, transmissions were assigned unique
                identification designations based on technology type, drive type, gear
                count, and technology version. The data were also used to identify the
                most similar transmission among the Autonomie transmission models, as
                discussed further below.
                 The transmission characteristics of vehicles in the analysis fleet
                show manufacturers use transmissions that are the same or similar on
                multiple vehicle models. Manufacturers have told the agencies they do
                this to control component complexity and associated costs for
                development, manufacturing, assembly, and service. Both the NPRM and
                final rule analyses account for this sharing. To identify common
                transmissions, the agencies considered the transmission type (manual,
                automatic, dual-clutch, continuously variable), number of gears, and
                vehicle architecture (front-wheel-drive, rear-wheel-drive, all-wheel-
                drive based on a front-wheel-drive platform, or all-wheel-drive based
                on a rear-wheel-drive platform). If multiple vehicle models shared
                these attributes, the transmissions were treated as single group for
                the analysis. Vehicles in the analysis fleet with the same transmission
                configuration adopted transmission technology together.
                 For ATs and CVTs, the identification of the most similar Autonomie
                transmission model required additional steps beyond just assigning gear
                count for ATs, or just assigning the CVT model. A review of the age of
                the transmission design, relative performance versus previous designs,
                and technologies incorporated was conducted, and the information
                obtained was used to assign a HEG level. Engineering judgment was used
                to compare the technologies and performance improvements reported
                versus descriptions of HEG technology discussed in the NAS report.\970\
                ---------------------------------------------------------------------------
                 \970\ 2015 NAS Report, at 191.
                ---------------------------------------------------------------------------
                 In addition, no automatic transmissions in the 2017 analysis fleet
                were determined to be initially at a HEG Level 3. However, all 7-speed
                automatic transmissions, all 9-speed automatic transmissions, all 10-
                speed automatic transmissions and some 8-speed automatic transmissions
                were found to be advanced transmissions operating at a Level 2 HEG
                equivalence. All other transmissions were assigned at the minimum
                level.
                c) Transmission Adoption Features
                 The agencies included several transmission adoption features in the
                NPRM that have been carried over for the final rule analysis. For a
                detailed discussion of path logic applied in the final rule analysis,
                including technology supersession logic and technology mutual
                exclusivity logic, please see FRM CAFE Model Documentation Section
                S4.5, Technology Constraints
                [[Page 24459]]
                (Supersession and Mutual Exclusivity).\971\
                ---------------------------------------------------------------------------
                 \971\ Available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
                ---------------------------------------------------------------------------
                (1) Automatic Transmissions
                 Automatic transmission technology adoption is defined by path logic
                and technology availability. The transmission path precludes adoption
                of other transmission types once a platform progresses past an AT6.
                This restriction is used to avoid the significant level of stranded
                capital that could result from adopting a completely different
                transmission type shortly after adopting an advanced transmission,
                which would occur if a different transmission type was adopted after
                AT6 in the rulemaking timeframe. Stranded capital is discussed in more
                detail in Section VI.B.4.c), Stranded Capital Costs. In addition, any
                automatic transmissions that use HEG3 technology cannot be phased in
                until the 2020 model year. The technology phase-in year is based on the
                estimated availability of HEG3 technology from the NAS (2015) report
                and confidential data obtained from OEM's and suppliers. Finally, all
                P2HEVs are paired with an AT8 transmission, which is also discussed
                further in Section VI.C.3.c).
                 One commenter expressed concern that all P2HEVs were paired with an
                AT8 transmission, and argued that the full slate of transmission
                technology should be available for adoption with that powertrain
                technology.\972\ The commenter correctly observed a limit of
                transmission technologies for use only with the P2HEV technology
                option; all other HEV based technology options did not have this
                limitation.
                ---------------------------------------------------------------------------
                 \972\ Comments from Meszler Engineering Services, Attachment 2,
                NPRM Docket No. NHTSA-2018-0067-11723 at 32.
                ---------------------------------------------------------------------------
                 The agencies disagree that a greater variety of transmission
                technologies are necessary to model the P2HEV technology reasonably.
                The P2HEV demonstrated limited response to transmission technologies
                beyond the AT8L2, and access to those technologies were limited to
                reflect the diminishing returns anticipated for higher gear counts used
                in conjunction with the P2 system, and trends in industry.\973\
                Adopting P2HEV to a conventional vehicle provides a significant fuel
                consumption improvement, agnostic of transmission type, based on the
                agencies' full vehicle simulation results.
                ---------------------------------------------------------------------------
                 \973\ Greimel, H. ``ZF CEO--We're not chasing 10-speeds,''
                Automotive News, November 23, 2014, http://www.autonews.com/article/20141123/OEM10/311249990/zf-ceo:-were-not-chasing-10-speeds.
                ---------------------------------------------------------------------------
                (2) Continuously Variable Transmissions
                 Application of CVTs in the NPRM and final rule analysis was not
                allowed for high torque vehicle applications. The launch, acceleration,
                and ratio variation characteristics of powertrains with CVTs may be
                significantly different than ATs leading to potential consumer
                acceptance issues and/or complaints. Several manufacturers have told
                the agencies that they employ strategies that mimic AT shifting under
                some conditions to address these issues. Some manufacturers have also
                encountered significant engineering challenges in employing CVTs for
                use in high torque or high load applications.
                 In addition, the CVT adoption was limited by technology path logic.
                CVTs cannot be adopted by vehicles that do not start with a CVT or by
                vehicles beyond the AT6 in the baseline fleet which have a greater
                number of gear ratios and therefore increased ability to operate the
                engine at a highly efficient speed and load. Once on the CVT path the
                platform is only allowed to apply improved CVT technologies. This
                restriction is used to avoid the significant level of stranded capital
                that could result from adopting a completely different transmission
                type shortly after adopting an advanced transmission, which would occur
                if a different transmission type was adopted in the rulemaking
                timeframe. Stranded capital is discussed in more detail in Section
                VI.B.4.c), Stranded Capital Costs.
                 The Alliance commented that the analysis ``appropriately restricts
                the application of CVT technology on larger vehicles.'' \974\ The
                agencies concurred with the Alliance's observations and thus the
                limitations on CVT application were continued in the final rule
                analysis.
                ---------------------------------------------------------------------------
                 \974\ Comments from Auto Alliance, Attachment 1, NHTSA-2018-
                0067-12073, at 142.
                ---------------------------------------------------------------------------
                (3) Dual Clutch Transmission
                 For DCTs, while the structure of the model could allow automatic
                transmissions to consider applying a DCT, the market data file was used
                to preclude the application of DCTs to vehicles that had already
                adopted an automatic transmission with six or more gears (e.g., AT6
                through AT10). The model allows baseline vehicles that have DCTs to
                apply an improved DCT (if opportunities to do so exist), and allows
                vehicles with an AT5 to consider DCTs. This was done to ensure vehicle
                functionality is maintained as technologies are applied, and accounts
                for consumer acceptance issues related to the drivability and launch
                performance tradeoffs. These issues with DCTs resulted in a low
                relative adoption rate over the last decade.\975\ It also is broadly
                consistent with manufacturers' technology choices.
                ---------------------------------------------------------------------------
                 \975\ ``The 2018 EPA Automotive Trends Report,'' Page 60, figure
                4.18, https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends, Accessed Aug 23, 2019.
                ---------------------------------------------------------------------------
                (4) Manual Transmissions
                 Manual transmission technology adoption in the CAFE model remained
                unchanged from the NPRM and is only limited by the technology path
                limits discussed above. Manual transmissions cannot be adopted by
                vehicles that do not start with a manual transmission in the analysis
                fleet. Vehicles with manual transmissions cannot receive an alternate
                transmission technology, and may only progress to more advanced manual
                transmissions. These restrictions are in recognition of the low
                customer demand for manual transmissions.\976\
                ---------------------------------------------------------------------------
                 \976\ ``The 2018 EPA Automotive Trends Report,'' https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends, Accessed Aug 23, 2019.
                ---------------------------------------------------------------------------
                d) Transmission Effectiveness Modeling and Resulting Effectiveness
                Values
                 For the NPRM and final rule analysis, full vehicle simulation was
                used to understand how transmissions work within the full vehicle
                system to improve fuel economy, and how changes to the transmission
                subsystem influence the performance of the full vehicle system.
                 The Autonomie tool models transmissions as a sequence of mechanical
                torque gains. The torque and speed are multiplied and divided,
                respectively, by the current ratio for the selected operating
                condition. Furthermore, torque losses corresponding to the torque/speed
                operating point are subtracted from the torque input. Torque losses are
                defined based on a three-dimensional efficiency lookup table that has
                as inputs: Input shaft rotational speed, input shaft torque, and
                operating condition.\977\
                ---------------------------------------------------------------------------
                 \977\ Detailed discussion of transmission modeling can be found
                in the ANL Model Documentation at Chapter 4 and Chapter 5.
                ---------------------------------------------------------------------------
                 The general transmission models are populated with characteristics
                data to model specific transmissions. Characteristics data are
                typically provided in the form of tabulated data for transmission gear
                ratios, maps for transmission efficiency, and maps for torque converter
                performance, as applicable. The quantity of data needed
                [[Page 24460]]
                depends on the transmission technology being modeled. The
                characteristics data for these models was collected from peer-reviewed
                sources, transmission and vehicle testing programs, results from
                simulating current and future transmission configurations, and
                confidential data obtained from OEMs and suppliers.\978\
                ---------------------------------------------------------------------------
                 \978\ Downloadable Dynamometer Database.: https://www.anl.gov/energy-systems/group/downloadable-dynamometer-database, Kim, N.,
                Rousseau, N., Lohse-Bush, H., ``Advanced Automatic Transmission
                Model Validation Using Dynamometer Test Data,'' SAE 2014-01-1778,
                SAE World Congress, Detroit, April 2014. Kim, N., Lohse-Bush, H.,
                Rousseau, A., ``Development of a model of the dual clutch
                transmission in Autonomie and validation with dynamometer test
                data,'' International Journal of Automotive Technologies, March
                2014, Volume 15, Issue 2, pp 263-271.
                ---------------------------------------------------------------------------
                 The level of HEG improvement applied to a given transmission was
                modeled by improvements made to the efficiency map of the transmission.
                As an example, the 8-speed automatic transmission models show how a
                model can be incrementally improved with the addition of the HEG
                enhancement. The AT8 is the model of a baseline transmission developed
                from a transmission characterization report.\979\ The AT8L2 has the
                same gear ratios as the AT8, however the gear efficiency map has been
                improved to represent application of the HEG level 2 technologies. The
                AT8L3 models the application of HEG level 3 technologies using the same
                principle, further improving the gear efficiency map over the AT8L2
                improvements.
                ---------------------------------------------------------------------------
                 \979\ See PRIA Section 6.3.3.2
                ---------------------------------------------------------------------------
                 The NPRM and final rule analysis, using the Autonomie tool,
                comprehensively simulated each of the 18 transmission technologies.
                Each transmission was modeled with explicit gear ratios, gear
                efficiencies, gear spans, adaptive shift logic, and transmission
                architecture individually for each of the ten vehicle types. The NPRM
                and final rule analysis clearly showed the specific contributions to
                effectiveness provided by each transmission technology combination and
                the associated cost. This provided greater transparency for public
                review and comment.
                 The implementation of the full vehicle simulation approach used in
                the NPRM analysis, and carried forward to the final rule analysis,
                clearly defines the contribution of individual transmission
                technologies and separates those contributions from other technologies.
                This modeling approach comports with the National Academy of Science
                2015 recommendation to use full vehicle modeling supported by
                application of collected improvements at the sub-model level.\980\ The
                approach allows the isolation of technology effects in the analysis
                which contributes to an accurate cost assessment.
                ---------------------------------------------------------------------------
                 \980\ 2015 NAS Report, at 292.
                ---------------------------------------------------------------------------
                 This approach was supported by the Auto Alliance, who commented in
                support of the agencies' explicit and transparent modeling of the cost
                and effectiveness for each of the transmission technologies. The
                Alliance contrasted the NPRM approach with the transmission modeling
                methodology used in the Proposed Determination--which they strongly
                objected to--which had lumped together fundamentally different
                transmission technologies into bundles with identical cost and
                efficiencies, ``making it impossible to fully comprehend the
                rationale'' for the Proposed Determination's high effectiveness
                estimates.\981\
                ---------------------------------------------------------------------------
                 \981\ Comments from Alliance of Automobile Manufacturers, NHTSA-
                2018-0067-12073, at 142.
                ---------------------------------------------------------------------------
                 However, other stakeholders were not supportive of the modeling
                approach used in the NPRM. The Union of Concerned Scientists (UCS)
                thought a level of abstraction was necessary to account for
                unpredictability in the market, such as the failure of the dual-clutch
                transmission to reach widespread use as anticipated in the agencies
                2012 analysis for MYs 2017 and later. UCS thought that keeping the
                transmission technology generalized would avoid the pitfalls of
                potentially picking the wrong technology leader, but would still
                predict the general trend of behavior, stating that ``[i]ncidentally,
                this is an example of why we supported EPA's move to a more generic
                representation of transmissions in its OMEGA modeling.'' \982\
                ---------------------------------------------------------------------------
                 \982\ Comments from Union of Concerned Scientists, NHTSA-2018-
                0067-12039, at 20-21.
                ---------------------------------------------------------------------------
                 The agencies disagree with UCS's suggestion to generalize the
                transmission technology groupings for the analysis. By grouping the
                technologies into overly broad, generic categories, the analysis loses
                accuracy on the costs and the effectiveness for specific systems. The
                OMEGA model used general transmission categories, asked for by UCS's
                comments, as part of the CO2 analysis in the Draft TAR and
                in the Proposed Determination, and the assumptions and limitations were
                acknowledged at the time.983 984 One assumption used by the
                OMEGA model approach was ``[t]he incremental effectiveness and cost for
                all automated transmissions are based on data from conventional
                automatics.'' \985\ In response, the Alliance observed that the
                transmission groups used ``do not recognize unique efficiencies of
                different transmission technologies.'' \986\ At the time EPA stated
                ``the potential effectiveness gains between TRX levels, while arising
                from different technology packages within each transmission type, will
                be very similar among the transmission types.'' \987\ However, as shown
                in Table VI-81 and Table VI-82, there are nontrivial differences in the
                costs of different transmission technologies.
                ---------------------------------------------------------------------------
                 \983\ ``Midterm Evaluation of Light duty Vehicle Greenhouse Gas
                Emission Standards and Corporate Average Fuel Economy Standards for
                Model Years 2022-2025,'' Paragraph 5.3.4.2.1, EPA-420-D-16-900, July
                2016.
                 \984\ ``Proposed Determination on the Appropriateness of the
                Model Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions
                Standards under the Midterm Evaluation, Technical Support
                Document,'' Pages 2-328--2-329, EPA-420-R-16-021, November 2016.
                 \985\ ``Proposed Determination on the Appropriateness of the
                Model Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions
                Standards under the Midterm Evaluation, Technical Support
                Document,'' Pages 2-327, EPA-420-R-16-021, November 2016.
                 \986\ ``Proposed Determination on the Appropriateness of the
                Model Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions
                Standards under the Midterm Evaluation, Technical Support
                Document,'' Pages 2-329, EPA-420-R-16-021, November 2016.
                 \987\ ``Proposed Determination on the Appropriateness of the
                Model Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions
                Standards under the Midterm Evaluation, Technical Support
                Document,'' Pages 2-329, EPA-420-R-16-021, November 2016.
                ---------------------------------------------------------------------------
                 The approach used in the NPRM analysis and this final rule analysis
                is an evolution of the approach used for the Proposed Determination
                model, and avoids the issue described above. The NPRM and final rule
                analyses reduce the span of transmission technology groupings, with the
                intent to provide an increase in fidelity and precision for cost and
                performance, as was requested by stakeholders such as the Auto
                Alliance, while including tools to mitigate market effects, which
                addresses other concerns such as those expressed by UCS. In the
                analysis for the final rule the transmissions are grouped by technology
                type (AT, DCT, CVT, etc.) and gear count (5,6,7, etc.). The level of
                HEG technology applied as a separate factor further subdivided the
                transmission groups. Defining technology adoption features addresses
                the potential for market forces, such as those that affected the sales
                of DCTs, and supports the narrower technology groupings. Technology
                adoption features are defined through market research, historic and
                current fleet composition analysis, and dialogue with manufacturers.
                 Commenters also provided general comments regarding the values of
                effectiveness for advanced transmissions used for the NPRM
                [[Page 24461]]
                analysis versus values used for the Draft TAR. For example, CARB noted
                a ``2 percent-3 percent lower efficiency assumed for advanced 8- and 9-
                speed transmissions relative to the data EPA itself previously
                developed with back to back testing on FCA vehicles,'' \988\ with
                similar concerns expressed by other commenters.\989\ Meszler
                Engineering Services wondered ``why the AT10 technology was being so
                widely adopted when its associated benefits appeared negligible for a
                particular vehicle'' and noted ``[t]he wide ranging effectiveness
                estimates were unexpected.'' \990\ Senator Tom Carper also noted ``the
                most advanced eight speed transmission technology are assigned
                unrealistically low fuel efficiency effectiveness values for some
                vehicle types.'' \991\
                ---------------------------------------------------------------------------
                 \988\ Comments from CARB, Attachment 2018-10-26 FINAL CARB
                Detailed Comments on SAFE, NPRM Docket No. NHTSA-2018-0067-11873, at
                110-113.
                 \989\ Comments from Roush Industries, Attachment 1, NPRM Docket
                No. NHTSA-2018-0067-11984, at 5; Comments from CARB, Attachment HDS
                Final Report, NPRM Docket No. NHTSA-2018-0067-11985, at 26, 47.
                 \990\ Comments from Meszler Engineering Services, Attachment 2,
                NPRM Docket No. NHTSA-2018-0067-11723, at 5-6.
                 \991\ Comments from Senator Tom Carper, Attachment 1, NPRM
                Docket No. NHTSA-2018-0067-11910, at 4.
                ---------------------------------------------------------------------------
                 The Auto Alliance also provided comments with regards to the larger
                variation of effectiveness values that were of concern to commenters
                such as Meszler Engineering Services and Senator Tom Carper. The Auto
                Alliance acknowledged that the use of full vehicle simulation, with
                more details, results in greater diversity of results. The comment
                stated, ``Over an entire fleet, a more reasonable expectation is that
                there will be some vehicles with higher fuel economy than expected for
                a given technology set and some vehicles with a lower fuel economy than
                expected for a given technology set. As discussed above, these
                differences arise for a variety of reasons, and cannot simply be
                attributed to ``less than optimal technology integration.'' \992\
                ---------------------------------------------------------------------------
                 \992\ Comments from Alliance of Automobile Manufacturers,
                Attachment 1, NPRM Docket No NHTSA-2018-0067-12385, at 9.
                ---------------------------------------------------------------------------
                 The Auto Alliance also specifically commented on the FCA vehicle
                study used to support CARB's comment and used to generate the TAR
                analysis values. The Auto Alliance pointed out that the vehicles used
                in the study had other technology differences, however the study still
                ``proceeds to compare the fuel economy of these variants to assert
                support for its own estimate of transmission effectiveness. This
                comparison neglects that the 2.4L engines in these variants are not the
                same and that the variant with the nine-speed transmission was a
                redesigned vehicle.'' The Alliance concluded, therefore, that ``the
                Chrysler 200 comparison provided by H-D Systems does not compare a
                transmission change in isolation from other changes that impact fuel
                economy and likely overestimates the benefits associated with the
                transmission change.'' The Auto Alliance summarized the analysis of the
                study by noting that ``[s]uch differences also impact fuel economy,
                confounding an analysis which purports to compare the fuel economy
                benefits associated directly with the transmission.'' \993\
                ---------------------------------------------------------------------------
                 \993\ Comments from Alliance of Automobile Manufacturers,
                Attachment 1, NPRM Docket No NHTSA-2018-0067-12385, at 27-28.
                ---------------------------------------------------------------------------
                 The agencies agree with the Auto Alliance assessment of the 8- and
                9-speed FCA vehicles, and have based analysis inputs on alternate
                information sources.\994\ However, the observations by commenters of a
                wider range of values for the NPRM effectiveness when compared to the
                Draft TAR compliance analyses are a direct result of the improvements
                in modeling approach. As discussed above the NPRM compliance analysis
                increased the number of transmission technology paths considered by
                further subdividing the technology groupings. The change resulted in a
                wider range of effectiveness, as the specific transmission technologies
                are paired across all the configurations of vehicle technologies. In
                addition to this greater range, there were also specific effectiveness
                issues identified for some of the transmission technologies, which are
                addressed in the sections below.
                ---------------------------------------------------------------------------
                 \994\ See Data discussed in PRIA Section 6.3.3.2. and Kim, N.,
                Rousseau, N., Lohse-Bush, H. ``Advanced Automatic Transmission Model
                Validation Using Dynamometer Test Data,'' SAE 2014-01-1778, SAE
                World Congress, Detroit, April 2014. Kim, N., Lohse-Bush, H.,
                Rousseau, A. ``Development of a model of the dual clutch
                transmission in Autonomie and validation with dynamometer test
                data,'' International Journal of Automotive Technologies, March
                2014, Volume 15, Issue 2, pp 263-271.
                ---------------------------------------------------------------------------
                 Commenters may also be observing, with comments like ``advanced
                transmissions have low effectiveness with some vehicles types,'' an
                expected effect when an advanced transmission is coupled to an advanced
                engine. The National Academy of Science, in their 2015 report, noted
                that ``as engines incorporate new technologies to improve fuel
                consumption, including variable valve timing and lift, direct
                injection, and turbocharging and downsizing, the benefits of increasing
                transmission ratios or switching to a CVT diminish.'' \995\ This is not
                to say that transmissions are not an important technology going
                forward, but rather a recognition that advanced engines have larger
                ``islands'' of low fuel consumption that rely less on the transmission
                to improve the overall efficiency of the vehicle. Thus, effectiveness
                percentages reported for transmissions paired with unimproved engines
                would be expected to be reduced when the same transmission is paired
                with a more advanced engine.
                ---------------------------------------------------------------------------
                 \995\ 2015 NAS Report, at 175.
                ---------------------------------------------------------------------------
                 Commenters also expressed concern for the transmission gear set and
                final drive values used for the NPRM analysis, or, more specifically,
                that the gear ratios were held constant across applications. Roush
                commented that ``all transmissions with a given number of ratios (8-
                speed, 10-speed) maintain the same individual step ratios'' and that
                this would lead to ``powertrain inefficiencies and under-predict
                potential fuel economy benefits.'' \996\ CARB, quoting a report from
                its contractor, noted that ``the final drive ratio was kept constant as
                powertrains were changed and that transmission gear ratios were not
                optimized,'' and suggested that manufacturers forgoing improvements
                from gear ratio or final drive ratio changes is unrealistic and results
                in an underestimation of the benefits from advanced transmissions.\997\
                ---------------------------------------------------------------------------
                 \996\ Comments from Roush Industries, Attachment 1, NPRM Docket
                No. NHTSA-2018-0067-11984, at 14-15.
                 \997\ Comments from CARB, Attachment 1, NPRM Docket No. NHTSA-
                2018-0067-11873, at 110.
                ---------------------------------------------------------------------------
                 However, the Auto Alliance stated that ``[m]anufacturers share
                major technologies such as transmissions and engines across multiple
                vehicle models and platforms.'' The Auto Alliance also supported the
                agencies' approach of not including final drive ratio changes,
                particularly when only minor system changes are incurred. The Auto
                Alliance continued further stating that ``[i]n the case of passenger
                cars, the final drive ratio is frequently the same across multiple
                models that use the same transmission.'' \998\
                ---------------------------------------------------------------------------
                 \998\ Comments from Auto Alliance, Attachment 1, NPRM Docket No.
                NHTSA-2018-0067-12073, at 142.
                ---------------------------------------------------------------------------
                 The agencies disagree with Roush, Duleep, and CARB's assessment. It
                is an observable practice in industry to use a common gear set across
                multiple platforms and applications. The most recent example is the GM
                10L90, a 10-speed automatic transmission that used the same gear set in
                both pick-up truck
                [[Page 24462]]
                and passenger car applications.\999\ Optimization of performance is
                achieved through shift control logic rather than customized hardware
                for each vehicle line. The use of a single gear set for each
                transmission technology also supports the overall analysis approach.
                The level of technology performance modeled must reasonably represent a
                typical level of performance representative of the industry range of
                performance. If the systems were over-optimized for the agencies'
                modeling, such as applying a unique gear set for each individual
                vehicle configuration, the analysis would likely over-predict the
                reasonably achievable fuel economy improvement for the technology.
                Over-prediction would be exaggerated when applied under real-world
                large-scale manufacturing constraints necessary to achieve the
                estimated costs for the transmission technologies. Accordingly, the
                agencies used the NPRM approach for the final rule analysis.
                ---------------------------------------------------------------------------
                 \999\ ``GM Global Propulsion Systems--USA Information Guide
                Model Year 2018'' (PDF). General Motors Powertrain. Retrieved 26
                September 2019. https://www.gmpowertrain.com/assets/docs/2018R_F3F_Information_Guide_031918.pdf.
                ---------------------------------------------------------------------------
                 In response to comments related to the effectiveness of micro-HEV
                systems, which are discussed in Section VI.C.3.d)(2)(a), and comments
                related to the effectiveness of diesel engines, which are discussed in
                Section VI.C.1.c)(8), the agencies took a close look at NPRM
                effectiveness results. Two issues were identified related to the
                interaction between Autonomie transmission models and other Autonomie
                powertrain technology models. First, a logic issue was found in a
                transmission control subroutine and, second, there was an issue with a
                sub-model input. While these items were caused by issues in the
                transmission model sub-systems, the effects manifested in the
                effectiveness of the micro-HEV systems and the diesel engine systems.
                Autonomie uses a gearbox transient sub-model to control the simulated
                state of powertrain components during a transmission event, such as
                shifting or vehicle starting and stopping. The simulated powertrain
                component states include conditions such as clutch engagement, or
                engine operation mode. A detailed discussion of the Autonomie control
                model can be found FRM Argonne Model Documentation file at Section 4.4.
                Different versions of the sub-model are used for micro-HEV technologies
                (12VSS and ISG) than for conventional drivetrains, mild-HEV or Strong-
                HEV systems.
                 An issue was found in the control logic used in the micro-HEV
                version related to the sequence of powertrain component modes during
                shifting events for automatic transmissions, regenerative braking
                events for automatic transmissions, and stop start events for manual
                transmissions. While these issues reduced the effectiveness of the
                micro-HEV technology in the Argonne modeling results, they had very
                minimal effect on the overall NPRM Analysis. The control logic issue
                was resolved for the final rule analysis. There also was an issue with
                the gearbox transient sub-model used for micro HEVs that impacted
                calculation of the CVT best efficiency operating ratio targets under
                low torque conditions. This resulted in some negative effectiveness
                values for certain CVT technology combinations, but had very minimal
                effect on the overall NPRM results. This software item was also
                resolved for the final rule analysis.
                 As discussed in the Autonomie model documentation, FRM Argonne
                Model Documentation file at Section 4, the full vehicle model is
                created from a network of subsystem models. The subsystems all interact
                through data connections transferring outputs from one subsystem model
                to the inputs of another. An issue was identified with the definition
                of the connection between the gearbox transient sub-model for DCT's
                with diesel engines, which impacted the values provided to the diesel
                control model. This caused reduced effectiveness values for the diesel
                engines with DCTs in the Argonne modeling results, however it had very
                minimal effect on the overall NPRM analysis. The data connection issue
                was resolved for the final rule analysis.
                 Lastly, the agencies received several comments on transmission
                shifting logic, which are addressed in the following section.
                (1) Shift Logic
                 Transmission shifting logic has a significant impact on vehicle
                energy consumption and was modeled in Autonomie to maximize the
                powertrain efficiency while maintaining acceptable drive quality. The
                logic used in the Autonomie full vehicle modeling relied on two
                components: (1) The shifting controller, which provides the logic to
                select appropriate gears during simulation; and (2) the shifting
                initializer, an algorithm that defines shifting maps (i.e., values of
                the parameters of the shifting controller) specific to the selected set
                of modeled vehicle characteristics and modeled powertrain
                components.\1000\
                ---------------------------------------------------------------------------
                 \1000\ See FRM ANL Model Documentation file at Paragraph 4.4.5.
                ---------------------------------------------------------------------------
                (a) Shifting Controller
                 The shift controller is the logic that governs shifting behavior
                during simulated operation. The shift controller performance was
                informed by inputs from the model. The inputs included: Specific engine
                or transmission used, and instantaneous conditions in the simulation.
                Instantaneous conditions included values such as vehicle speed, driver
                demand and a shifting map unique to the full vehicle
                configuration.\1001\ The shift controller logic was consistently
                applied for all vehicles simulated.
                ---------------------------------------------------------------------------
                 \1001\ See FRM ANL Model Documentation file at Paragraph 4.4.5.
                ---------------------------------------------------------------------------
                 Although no comments were received specifically on shift control
                logic, the agencies tracked several effectiveness concerns identified
                by commenters back to how the agencies modeled some transmissions
                paired with turbocharged engines. Meszler Engineering Services
                discussed an unexpected range of effectiveness observed for
                transmissions when coupled to different engine technologies, and
                concluded that ``[m]oreover, the variation across technology
                combinations is markedly different.'' \1002\ Senator Carper's comments
                mirrored Meszler's, noting that ``the more expensive version of an
                engine technology (TURBO2), which would be expected to be more fuel-
                efficient, was instead assigned a negative fuel-efficiency value for
                some types of vehicles.'' \1003\ The Senator also observed the same
                phenomenon for cooled exhaust gas recirculation (CEGR I), which ``was
                assigned a fuel-efficiency effectiveness of at or near zero.''
                Similarly, UCS noted that ``many simulations of improved transmissions
                and turbocharged engines show little incremental improvement over less
                complex technologies.'' \1004\
                ---------------------------------------------------------------------------
                 \1002\ Comments from Meszler Engineering Services, Attachment 2,
                NPRM Docket No. NHTSA-2018-0067-11723, at 5-6.
                 \1003\ Comments from Senator Tom Carper, Attachment 1, NPRM
                Docket No. NHTSA-2018-0067-11910, at 4.
                 \1004\ Comments from UCS, Attachment 1, NPRM Docket No. NHTSA-
                2018-0067-12039, at 32.
                ---------------------------------------------------------------------------
                 In response to the comments, the agencies conducted an in-depth
                review of these technology combinations. The agencies determined the
                minimum lugging speed for turbocharged engines, which controls the
                minimum engine speed allowed before down-shifting, caused the observed
                behavior. The issue was isolated to some combinations of advanced
                transmissions and
                [[Page 24463]]
                turbocharged engines. For the final rule analysis, a modification was
                made to the shift controller logic of transmissions coupled to
                turbocharged engines. Specifically, the minimum lugging speed allowed
                for turbocharged engines was increased in the shift controller. An
                increase in lugging speed increases the minimum speed at which the
                shift controller will allow the engine to operate before down-shifting,
                resulting in increased operation in better efficiency regions of the
                engine map.\1005\ The updated lugging speeds are based on Argonne
                benchmarking data of the 2017 F150.\1006\ The updated values are shown
                in Table VI-79, the lugging speeds for naturally aspirated engines are
                shown as reference and remain unchanged from the NPRM.
                ---------------------------------------------------------------------------
                 \1005\ See FRM ANL Model Documentation at Paragraph 4.4.5.1, for
                more details on lugging speed.
                 \1006\ NHTSA Benchmarking, ``Laboratory Testing of a 2017 Ford
                F-150 3.5 V6 EcoBoost with a 10-speed transmission.'' DOT HS 812
                520.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.192
                (b) Shift Initializer
                 As defined above, the shifting initializer is an algorithm that
                defines shifting maps (i.e., values of the parameters of the shifting
                controller) specific to the selected set of modeled vehicle
                characteristics and modeled powertrain components.
                 Commenters stated that the model did not customize shifting maps
                for each transmission application. Roush Industries commented, ``[t]he
                2018 PRIA analysis assumes that all transmissions with a given number
                of ratios maintain the same individual step ratios and shift maps.''
                \1007\ Roush also commented that the effectiveness of transmissions
                were understated due to inaccurate transmission maps or ``the lack of
                vehicle system optimization and calibration.'' \1008\ UCS stated that
                the ``transmission shift strategy does not deploy gear-skipping or
                other more modern control strategies.'' \1009\ HDS provided similar
                comments to Roush, observing that the Autonomie models ``do not
                optimize engine efficiency after most changes in tractive load because
                the model employs fixed shift points, gear ratios, and axle ratios.''
                \1010\ Finally, CARB expressed that ``[f]or the Autonomie modeling, a
                fixed final drive ratio was utilized and, presumably, a fixed shift
                logic based on the selected transmission.'' \1011\
                ---------------------------------------------------------------------------
                 \1007\ Comments from Roush Industries, Attachment 1, NPRM Docket
                No. NHTSA-2018-0067-11984, at 14-15.
                 \1008\ Comments from Roush Industries, Attachment 1, NPRM Docket
                No. NHTSA-2018-0067-11984, at 5.
                 \1009\ Comments from UCS, Attachment 1, NPRM Docket No. NHTSA-
                2018-0067-12039, at 23.
                 \1010\ Comments from K. Gopal Duleep, Attachment 1, NPRM Docket
                No. NHTSA-2018-0067-12395, at 4-5.
                 \1011\ Comments from CARB, Attachment 2018-10-26 FINAL CARB
                Detailed Comments on SAFE, NPRM Docket No. NHTSA-2018-0067-11873, at
                185.
                ---------------------------------------------------------------------------
                 The commenters seem to conflate the practice in the analysis of
                using the same gear sets across vehicle configuration with using the
                same shift maps. As commenters stated, they assumed the same maps were
                applied across vehicle models. However, the shift initializer routine
                was run for every unique Autonomie full vehicle model configuration and
                generated customized shifting maps. The algorithms' optimization was
                designed to balance minimization of energy consumption and vehicle
                performance.\1012\ This balance was necessary to achieve the best fuel
                efficiency while maintaining customer acceptability by meeting
                performance neutrality requirements, as discussed in Performance
                Neutrality, Section VI.B.3.a)(6).
                ---------------------------------------------------------------------------
                 \1012\ See FRM ANL Model Documentation at Paragraph 4.4.5.2.
                ---------------------------------------------------------------------------
                 While discussing shift logic, commenters also expressed concern
                about the capturing of fuel efficiency losses associated with shifting
                events. Roush stated, ``[t]he 2018 PRIA transmission modeling does not
                accurately capture the losses and FE penalty associated with a shift
                event.'' \1013\ The agencies disagree with this statement. While losses
                associated with a shifting event are not modeled as a single factor,
                the mechanisms that cause the loss are appropriately incorporated in
                the Autonomie transmission models.
                ---------------------------------------------------------------------------
                 \1013\ Comments from Roush Industries, Attachment 1, NPRM Docket
                No. NHTSA-2018-0067-11984, at 14-15.
                ---------------------------------------------------------------------------
                [[Page 24464]]
                The automatic transmission models have an associated torque converter
                model.\1014\ The torque converter model is designed to simulate the
                inertial and torque loads imposed on an engine because of shift events.
                Other clutch-based transmission models, MTs and DCTs, apply a general
                loss of efficiency across transmission efficiency maps to account for
                losses due to shift events.
                ---------------------------------------------------------------------------
                 \1014\ See FRM ANL Model Documentation at Paragraph 4.5 and
                Paragraph 5.4.
                ---------------------------------------------------------------------------
                (2) Transmission Effectiveness Values
                 The NPRM technology effectiveness modeling results showed that the
                effectiveness of a technology often varies with the type of vehicle and
                the other technologies that are on the vehicle. Figure VI-24 shows the
                range of effectiveness for each transmission technology across the
                range of vehicle types and technology combinations in the NPRM
                analysis. The data reflect the change in effectiveness for applying
                each transmission technology by itself while all other technologies are
                held unchanged. The effectiveness improvement range is over a 5-speed
                automatic transmission.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.193
                (a) Automatic Transmissions
                 Regarding AT effectiveness values, commenters pointed out the
                unusually high level of effectiveness displayed by the AT6L2
                transmission. ICCT and UCS both specifically expressed concern with the
                effectiveness of the AT6L2 compared to other advanced
                transmissions.1015 1016 The performance of the AT6L2 was
                central to ICCT's analysis of the NPRM inputs, which highlighted the
                AT6L2 models' performance, showing the cost versus effectiveness of the
                AT6L2 outperformed more advanced transmission options.\1017\
                ---------------------------------------------------------------------------
                 \1015\ Comments from International Council on Clean
                Transportation, Attachment 3, NPRM Docket No. NHTSA-2018-0067-11741,
                at I-26, I-64 (`` ``However, the impact of adding level 2
                transmission efficiency technologies varies wildly and produces
                absurd results. A 6-speed AT6L2 Is modeled as much more efficient
                (12.0% improvement) than a comparable 8-speed AT8L2 (9.1%) and even
                slightly more efficient than a comparable 10-speed AT10L2
                (11.5%).'')%).''.
                 \1016\ Comments from Union of Concerned Scientists, Attachment
                1, NPRM Docket No. NHTSA-2018-0067-12039, at 32. (``[I]n the NPRM
                analysis, 0 percent of vehicles had an AT6L2 transmission while 52.4
                percent adopted AT10L2 transmissions, even though the latter
                supplies virtually identical modeled efficiency.'').
                 \1017\ Comments from International Council on Clean
                Transportation, Attachment 3, NPRM Docket No. NHTSA-2018-0067-11741,
                at I-64--I-65.
                ---------------------------------------------------------------------------
                 Evaluation of the AT6L2 transmission model in response to these
                comments revealed an overestimated efficiency map was developed for the
                NPRM model. The high level of efficiency assigned to the transmission
                surpassed benchmarked advanced transmissions.\1018\ To address the
                issue, the agencies replaced the effectiveness values of the AT6L2
                model for the final rule analysis with AT7L2 effectiveness values.
                ---------------------------------------------------------------------------
                 \1018\ See PRIA Section 6.3.3.2. Sources of Transmission
                Effectiveness Data.
                ---------------------------------------------------------------------------
                 The updated estimate of effectiveness is supported by values shown
                in the NAS 2015 analysis.\1019\ The study estimated the difference in
                effectiveness between a 6-speed automatic transmission and a 7-speed
                automatic transmission of approximately the same technology level to be
                0.8 percent. The difference is reduced further when application of high
                efficiency gear box technology ranges of effectiveness is applied.
                Because the 7-speed automatic transmission and the advanced 6-speed
                automatic transmission technologies are parallel on the technology
                tree, the agencies felt using the same effectiveness value was
                reasonable and appropriate.
                ---------------------------------------------------------------------------
                 \1019\ 2015 NAS Report, at page 189.
                ---------------------------------------------------------------------------
                 Commenters also pointed out a lack of skip-shift logic used in the
                NPRM analysis, and an increase in the shift busyness observed for the
                high gear count transmissions. Roush commented on the NPRM analysis
                ``not incorporating the concept of `Skip shifting' which is important
                for reducing shift busyness and increasing FE especially in vehicles
                equipped with transmission with a large number of ratios (8-10).''
                \1020\ Both CARB and UCS repeated similar concerns.\1021\
                ---------------------------------------------------------------------------
                 \1020\ Comments from Roush Industries, Attachment 1, NPRM Docket
                No. NHTSA-2018-0067-11984 at 14-15.
                 \1021\ Comments from CARB, Attachment 2018-10-26 FINAL CARB
                Detailed Comments on SAFE, NPRM Docket No. NHTSA-2018-0067-11873, at
                110-113 (``Rogers found that the modeling did not consider `skip-
                shifting' where a transmission can upshift or downshift in a non-
                sequential manner''). Comments from UCS, Attachment 1, NPRM Docket
                No. NHTSA-2018-0067-12039, at 23 ``including that ANL's transmission
                shift strategy does not deploy gear-skipping'').''.
                ---------------------------------------------------------------------------
                 After consideration of the comments and re-evaluation of the NPRM
                results, the agencies concurred with the commenters. The lack of skip-
                shift logic and increased shift busyness can result in lower overall
                efficiency and decreased consumer acceptance. For the final rule
                analysis, a skip-shift logic was applied to the 10 speed automatic
                transmissions. The logic was based on the baseline 2017 Ford F150 10-
                speed transmission benchmarking performed by Argonne.\1022\ The
                introduction of the skip-shift logic impacted effectiveness and reduced
                the number of shifts by 23 percent for the 10-speed automatic
                transmission over the UDDS cycle.\1023\
                ---------------------------------------------------------------------------
                 \1022\ NHTSA Benchmarking, ``Laboratory Testing of a 2017 Ford
                F-150 3.5 V6 EcoBoost with a 10-speed transmission.'' DOT HS 812
                520.
                 \1023\ See FRM ANL Model Documentation file at Paragraph
                4.4.5.5. This update reduced the number of shift events from 231 to
                178.
                ---------------------------------------------------------------------------
                 In the NPRM analysis, transmission gear spans increased as the
                number of
                [[Page 24465]]
                gears increased.\1024\ However, to address further the comments related
                to optimization, the gear span of the AT10L3 was increased over the
                AT10L2, based on gear span data for the Honda 2018 10-speed
                transmission.\1025\ The AT10L3 span was increased to 10.10 in the final
                rule analysis from 7.34 in the NPRM analysis. However, the efficiency
                map for the AT10L3 remained the same for the final rule analysis.\1026\
                ---------------------------------------------------------------------------
                 \1024\ See FRM ANL Model Documentation file at 5.3.2.1.
                 \1025\ Sugino, S., SAE Internation Presentation., ``ALL-NEW
                HONDA 10-SPEED FWD TRANSMISSION.'' November 2017. ``2018 Honda
                Odyssey Press Kit--Overview.'' internet: Honda News, https://hondanews.com/en-US/releases/2018-honda-odyssey-press-kit-overview.
                Last accessed October 8, 2019.
                 \1026\ See FRM ANL Model Documentation file at 5.3.4.1.
                ---------------------------------------------------------------------------
                 Finally, in the agencies' review of NPRM model inputs, a weight
                discrepancy for the AT10 transmissions was identified. The weight
                assigned to the AT10 transmission in the NPRM analysis was too high.
                The weights were corrected for the final rule analysis. The AT10
                transmission weights were reduced by 20-45 kg, depending upon vehicle
                type.\1027\
                ---------------------------------------------------------------------------
                 \1027\ See FRIA VI.C.2.d.2.
                ---------------------------------------------------------------------------
                 The AT effectiveness values used for the final rule analysis can be
                seen in Figure VI-25. For automatic transmission technologies, the
                effectiveness improvement range is relative to a 5-speed automatic
                transmission. The new effectiveness values are a result of the
                aforementioned changes implemented to address comments. To summarize,
                the changes included an adjustment to the modeled effectiveness of the
                AT6L2, the use of skip-shift logic on the 10-speed transmissions, and
                the increase of the AT10L2 gear span.
                 Figure VI-25 shows the automatic transmission's effectiveness
                increases progressively in a logical order and behaves in an expected
                manner. Gains in effectiveness can be observed increasing as gear count
                increases, and as HEG levels increase. The effects of diminishing
                returns can be observed as gear count reaches higher levels, and
                effectiveness effects for increased gear count are reduced. This agrees
                with observed data reported by the NAS and industry
                stakeholders.1028 1029
                ---------------------------------------------------------------------------
                 \1028\ 2015 NAS Report, at 175.
                 \1029\ Greimel, H., ``ZF CEO--We're not chasing 10-speeds,''
                Automotive News, November 23, 2014, http://www.autonews.com/article/20141123/OEM10/311249990/zf-ceo:-were-not-chasing-10-speeds.
                ---------------------------------------------------------------------------
                (b) Continuously Variable Transmissions
                 For CVTs, the agencies also identified a discrepancy with the NPRM
                CVT weights. The weight assigned to the CVT class during the NPRM
                analysis was incorrect. Corrected values were assigned for the final
                rule analysis. The CVT weights were reduced by 9-10 kg based on vehicle
                type.\1030\
                ---------------------------------------------------------------------------
                 \1030\ See FRIA VI.C.2.d.2.
                ---------------------------------------------------------------------------
                 The CVT effectiveness values used for the final rule analysis can
                be seen in Figure VI-26, shown as an effectiveness improvement over a
                5-speed automatic transmission. The effectiveness values were not
                changed significantly from the values used in the NPRM analysis.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.194
                (c) Dual Clutch Transmissions
                 The DCT effectiveness values used for the final rule analysis can
                be seen in Figure VI-27, shown as an effectiveness improvement over a
                5-speed automatic transmission. The effectiveness values were not
                changed significantly from the values used in the NPRM analysis.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.195
                [[Page 24466]]
                (d) Manual Transmission
                 The MT effectiveness values used for the final rule analysis can be
                seen in Figure VI-28, shown as an effectiveness improvement over a 5-
                speed manual transmission. The effectiveness values were not changed
                significantly from the values used in the NPRM analysis.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.196
                e) Transmission Costs
                 For the NPRM, the transmission technology costs used as inputs for
                the CAFE model were retail price equivalent costs with learning curves
                applied. For a complete discussion on how the retail price equivalent
                and learning effects were applied to direct manufacturing costs see
                Section VI.B.4.b), Indirect Costs, and Section VI.B.4.d), Cost
                Learning. The direct manufacturing costs for the transmission
                technologies used in the NPRM were derived from technical sources and
                manufacturer's CBI.\1031\
                ---------------------------------------------------------------------------
                 \1031\ See PRIA Section 6.3.7.3.
                ---------------------------------------------------------------------------
                 Table VI-80 below shows the relative costs of the transmissions
                used in the NPRM analysis including learning and retail price
                equivalent.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.197
                BILLING CODE 4910-59-C
                (1) Automatic Transmissions
                 Several comments were received on technology costs, or cost
                effectiveness. Meszler Engineering Services noted that ``AT10L2 (level
                2 ten-speed automatic) transmission technology is another example of an
                end-of-path technology with very poor cost effectiveness relative to
                other transmission options.'' \1032\ A cost analysis by ICCT also
                showed relative costs of transmission technologies may not be in line
                with the modeled effectiveness.\1033\
                ---------------------------------------------------------------------------
                 \1032\ Comments from Meszler Engineering Services, Attachment 2,
                NPRM Docket No. NHTSA-2018-0067-11723, at 33.
                 \1033\ Comments from International Council on Clean
                Transportation, Attachment 3, NPRM Docket No. NHTSA-2018-0067-11741,
                at I-64.
                ---------------------------------------------------------------------------
                 The agencies conducted a review of transmission costs in response
                to the comments. For the final rule analysis, adjustments were made to
                costs of the AT6L2, AT7L2, AT9L2, AT10L2, and the AT10L3. The costs
                were adjusted based on reviewing the recommended relative costs
                discussed in the NAS 2015 report. Table VI-81 shows the cost for the
                automatic transmissions in the final rule analysis.
                 The direct manufacturing cost (DMC) estimate for the AT6 is drawn
                from Table 5.7 of the NAS report. The DMC estimate for the AT6L2 is
                based on the cost of the AT6 with HEG level 2 technology costs applied.
                This cost change is applied in accordance with the effectiveness
                adjustment made for the AT6L2.
                 A DMC estimate for the AT7 was drawn from Table 5.9 of the NAS
                report and was based on the cost of a system already equipped with HEG
                technology. The DMC estimate was given in 2007 dollars and relative to
                an AT5/AT4. The new DMC replaces the DMC from the NPRM, which did not
                account for the HEG technology.
                 The DMC for the AT9 technology was drawn from Table 8A.2a of the
                NAS (2015) report and per the NPRM description of the technology made
                relative to the AT8L2. The AT9 is assumed to have at least the level 2
                HEG technology applied. The NPRM analysis assumed the AT9 cost was only
                relative
                [[Page 24467]]
                to the AT8 and did not account for the cost of the HEG technology.
                 The DMC for the AT10 technologies was drawn from Table 8A.2a of the
                NAS report and per the NPRM description of the technology made relative
                to the AT8L2. The AT10L2 is assumed to have at least the level 2 HEG
                technology applied. The AT10L3 has the HEG3 technology applied. The
                NPRM analysis assumed the AT10 costs were only relative to the AT8 and
                did not account for the cost of the HEG technology.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.198
                (2) Continuously Variable Transmissions
                 No adjustments were made to the NPRM costs of the CVT technologies
                for the final rule analysis. Table VI-82 shows the cost for the CVTs in
                the final rule analysis.
                [[Page 24468]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.199
                (3) Dual Clutch Transmissions
                 The agencies received one comment on cost learning over time for
                DCT technologies. Roush Industries ``believes that the [actual]
                learning factors for such systems are significantly better than those
                estimated by either the 2018 PRIA or the 2016 Draft TAR.'' Roush stated
                that ``eight-speed DCTs (DCT8) are currently in production (MY2018),
                with quantities increasing significantly,''\1034\ but provided no
                specific supporting data.
                ---------------------------------------------------------------------------
                 \1034\ Comments from Roush Industries, Attachment 1, NPRM Docket
                No. NHTSA-2018-0067-11984, at 14-15.
                ---------------------------------------------------------------------------
                 The current learning curve for the DCT technologies was established
                based on recommendations from the NAS 2015 report and on CBI data
                collected from manufacturers and suppliers. Since Roush did not supply
                any data to support its comment, the agencies decided it was reasonable
                to make no change to the DCT learning curve for the final rule
                analysis. Table VI-83 shows the cost for the DCTs in the final rule
                analysis.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.200
                (4) Manual Transmissions
                 No adjustments were made to the NPRM costs of the manual
                transmission technologies for the final rule analysis. Table VI-84
                shows the cost for the MTs in the final rule analysis.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.201
                [[Page 24469]]
                BILLING CODE 4910-59-C
                3. Electric Paths
                 The electric paths include a large set of technologies that share
                the common element of using electrical power for certain vehicle
                functions that were traditionally powered mechanically by engine power.
                Electrification technologies thus can range from electrification of
                specific accessories (for example, electric power steering to reduce
                engine loads by eliminating parasitic loss) to electrification of the
                entire powertrain (as in the case of a battery electric vehicle).
                 Electrified vehicles are considered, for this analysis, to mean
                vehicles with a fully or partly electrified powertrain. These include
                several electrified vehicle categories, including: Battery electric
                vehicles (BEVs), which have an all-electric powertrain and use only
                batteries for propulsion energy; plug-in hybrid electric vehicles
                (PHEVs), which have a primarily electric powertrain and use a
                combination of batteries and an engine for propulsion energy; and
                hybrid electric vehicles (HEVs), which use electrical components and a
                battery to manage power flows and assist the engine for improved
                efficiency and/or performance. HEVs are further divided into strong
                hybrids (including P2 and power-split hybrids) that provide strong
                electrical assist and in many cases, can support a limited amount of
                all-electric propulsion, and mild hybrids (such as belt integrated
                starter generator (BISG) hybrids, crankshaft integrated starter
                generator (CISG) hybrids, and 48V mild hybrids) that typically provide
                only engine on/off with minimum electrical assist.
                 Fuel cell electric vehicles (FCEVs) are also another form of
                electrified vehicle having a fully electric powertrain, and are
                distinguished by the use of a fuel cell system rather than grid power
                as the primary energy source.
                 The factors that influence the cost and effectiveness of
                electrification technologies are their components. These include:
                Energy storage components such as battery packs; propulsion components
                such as electric motors; and power electronics components, such as
                inverters and controllers, that process and route electric power
                between the energy storage and propulsion components. For the purpose
                of this analysis, these components are divided into battery components
                and non-battery components.
                 Battery components strongly influence the cost of electrified
                vehicles.\1035\ Because developments in battery technology may apply to
                more than one category of electrified vehicles, they are discussed
                collectively in Section VI.C.3.e). That section details battery-related
                topics that directly affect the specification and costing of batteries
                for all types of electrified vehicles considered in this analysis.
                ---------------------------------------------------------------------------
                 \1035\ Battery costs are not necessarily a strong influence on
                fuel Cell Electric Vehicles, where the cost of the fuel cell
                technology has a larger influence.
                ---------------------------------------------------------------------------
                 Non-battery components also have an influence on both the cost and
                effectiveness of electrified vehicles. The selection and configuration
                of non-battery technologies distinguish the different architecture
                among electrified vehicles. Non-battery components largely consist of
                propulsion components and power electronics.
                 Propulsion components typically include one or more electric
                machines (an umbrella term that includes what are commonly known as
                motors, generators, and motor/generators). Depending on how they are
                employed in the design of a vehicle, electric machines commonly act as
                motors to provide propulsion, and/or act as generators to enable
                regenerative braking and conversion of mechanical energy to electrical
                energy for storage in the battery.
                 ``Power electronics'' refers to the various components that control
                or route power between the battery system and the propulsion
                components, and includes components such as: Motor controllers, which
                issue complex commands to control torque and speed of the propulsion
                components precisely; inverters and rectifiers, which convert and
                manage DC and AC power flows between the battery and the propulsion
                components; onboard battery chargers, for charging the BEV or PHEV
                battery from AC line power; and DC-to-DC converters that are sometimes
                needed to allow DC components of different voltages to work together.
                 Onboard chargers are charging devices permanently installed in
                electrified vehicles to allow charging from grid electrical power.
                Onboard chargers travel with the vehicle and are distinct from
                stationary charging equipment. Level 1 charging refers to charging
                powered by a standard household 110-120V AC power outlet. Level 2
                charging refers to charging at 220-240V AC power.
                 The agencies included a more extensive overview of charging
                technology and the state of charging infrastructure in the NPRM and
                PRIA, however, this was purely qualitative because charging was not
                accounted for in any respect in the NPRM analysis. The Alliance
                commented that ``[w]hile the costs of installing chargers and charger
                convenience were not taken into account within the Volpe model . . .
                these factors will continue to have an impact on the overall
                penetration of electrification technologies that the market will be
                willing to accept.'' \1036\ In contrast, the National Coalition for
                Advanced Transportation (NCAT) commented that the qualitative
                discussion overstated the risks and understated the benefits of
                electric vehicle charging.\1037\ Specifically, NCAT took issue with the
                characterization of potential risks of charging to the electric grid,
                stating that ``the PRIA's focus on worst case hypotheticals does not
                reflect the current capabilities of the grid, nor the dynamic nature of
                EV charging to mitigate any potential negative impacts. In both in the
                short-term and long-term, the impact of EVs with respect to the
                electric grid would have a net-positive impact to society, including
                the EV owners and utility customers broadly.'' NCAT also commented that
                ``[w]hile substantial investments in EV infrastructure have and will be
                made, the costs and benefits to consumers must be put into the
                appropriate context.'' NCAT cited two studies for the proposition that
                the average lifetime distribution electric vehicle infrastructure
                impact is about $80-$90 per electric vehicle sold, with the adoption of
                time of use rates and assuming a diversity of charging rates. NCAT also
                cited the California Public Utilities Commission 2016-2017 Electric
                Vehicle Load Research Report in support of their statement that the
                additional service and distribution system upgrades due to additional
                plug-in electric vehicle load is minimal, as ``of the approximately
                275,000 [electric] vehicles estimated to be on the road as of October
                2017 in the service areas of California's three investor-owned
                utilities, only 460, or 0.16 percent required a service line or
                distribution system upgrade solely to support the plug-in electric
                vehicle load at their residential charging location.''\1038\
                ---------------------------------------------------------------------------
                 \1036\ NHTSA-2018-0067-12073.
                 \1037\ NHTSA-2018-0067-11969.
                 \1038\ Citing Joint IOU Electric Vehicle Load Research Report
                (December 29, 2017), pp. 1-2, 12, available at http://www.cpuc.ca.gov/zev/ (2016-2017 Load Research Report).
                ---------------------------------------------------------------------------
                 The agencies agree that adding electric vehicle infrastructure will
                require additional costs, and information about what that cost is and
                how it can or should be accounted for
                [[Page 24470]]
                in the analysis is helpful for commenters to submit in order to put
                those considerations in the appropriate context. For this final rule,
                the agencies did not incorporate any costs related to electric vehicle
                charging infrastructure in the technology compliance analysis because
                those costs are separate from the costs that manufacturers and
                consumers would directly incur from a manufacturer transitioning part
                of their fleet to plug-in electric vehicles and consumers paying for
                those vehicles, even though local electric ratepayers will in all
                likelihood pay higher rates to upgrade local power grids to accommodate
                any widespread adoption of electrified vehicles. Accordingly, this
                means that the actual costs associated with electrified vehicles have
                been underestimated for the final rule analysis. The agencies did
                refine the estimates for the value of refueling time for electric
                vehicles, and that topic is discussed in Section VI.D.1.b)(11)(b). The
                agencies will continue to explore whether and how charging
                infrastructure should be incorporated into the analysis for future
                actions.
                 The following sections discuss vehicle electrification issues that
                were accounted for in the analysis, including the agencies'
                characterizations of electric vehicle technology, additional electric
                vehicle configurations added for the final rule analysis per
                commenters' requests, and the sources and methods used to develop
                battery and non-battery components, which were also refined for this
                final rule.
                a) Electrification Modeling in the CAFE Model
                 A set of technologies was chosen to represent the spectrum of
                electrification methods observed in the baseline fleet and that the
                agencies believed could be applied to vehicles in the rulemaking
                timeframe. Each technology was placed in a specific electrification
                pathway, grouping and defining the progression of related technologies.
                In the NPRM analysis, a total of eleven electrification technologies
                were contained in four electrification pathways. In consideration of
                comments received, the electrification technologies and associated
                pathways were modified for the final rule analysis, resulting in a
                total of eighteen variants of electrification technologies. Each of
                these NPRM and final rule technologies, and the electrification
                pathways they belong to, are detailed below. Operational modes of
                electrified vehicles are further described in the Argonne Model
                Documentation for the final rule.
                (1) Electrification Technologies
                (a) Electric Improvements
                 The electrification of power steering (EPS) and other accessories
                (IACC) have the potential of reducing fuel consumption by facilitating
                power-saving control strategies that avoid parasitic loss of engine
                power. These accessories traditionally are directly coupled to and
                driven by the conventional combustion engine; any time the engine is
                running some energy is continuously consumed by each accessory, even
                when it is not needed. By decoupling these accessories from the engine
                and instead driving them ``on-demand'' with electric motors, a more
                energy-efficient control strategy can be employed to reduce fuel
                consumption. EPS and IACC are discussed in detail in Section VI.C.7,
                Other Vehicle Technologies.
                (b) Micro Hybrid
                 12-volt stop-start (SS12V), sometimes referred to as start-stop,
                idle-stop or 12-volt micro hybrid, is the most basic hybrid system that
                facilitates idle-stop capability. In this system, the integrated
                starter generator is coupled to the internal combustion (IC) engine.
                When the vehicle comes to an idle-stop the IC engine completely shuts
                off and, with the help of 12-volt battery, the engine cranks and starts
                again in response to throttle to move the vehicle, or release of the
                brake pedal. The 12-volt battery used for the start-stop system is an
                improved unit capable of higher power, increased life cycle, and
                capable of minimizing voltage drop on restart. This technology is
                beneficial to reduce fuel consumption and emissions when the vehicle
                frequently stops, such as in city driving conditions or in stop and go
                traffic, and can be applied to all vehicle technology classes.
                 (c) Mild Hybrids
                 The belt integrated starter generator (BISG) and crank integrated
                starter generator (CISG), sometimes referred to as mild hybrid systems,
                provide idle-stop capability and use a higher voltage battery with
                increased energy capacity over typical automotive batteries. The higher
                voltage allows the use of a smaller, more powerful and efficient
                electric motor/generator, which replaces the standard alternator. In
                BISG systems, the motor/generator is coupled to the engine via belt
                (similar to a standard alternator), while the CISG integrates it to the
                crankshaft between the engine and transmission; both of these systems
                allow the engine to be automatically turned off as soon as the vehicle
                comes to a full stop. In addition, these motor/generators can recover
                braking energy while the vehicle slows down (regenerative braking) and
                in turn can propel the vehicle at the beginning of launch, allowing the
                engine to be restarted later. Some limited electric assist is also
                provided during acceleration to improve engine efficiency. The CISG
                system has a higher efficiency, but also higher cost than the BISG.
                 The agencies received limited high-level comments on CISG systems,
                with CARB stating that CISG systems are generally considered more
                capable and more efficient relative to BISG systems because they do not
                have the same belt-related constraints including maximum torque
                limitations, load restrictions on the front crank to avoid uneven
                crankshaft bearing wear, and mechanical energy transfer losses.\1039\
                CARB also noted that the decision to implement a CISG system is
                typically made early in the design process because doing so often
                requires an engine block casting change. CARB stated that the current
                high costs and larger dimensions, compared to BISGs, will likely delay
                major market penetration of CISG systems until beyond the MY 2025
                timeframe.
                ---------------------------------------------------------------------------
                 \1039\ Roush Industries on behalf of California Air Resources
                Board, Rogers_Final_Final_NPRM_10.26.2018, Docket No. NHTSA-2018-
                0067-11984, at 15.
                ---------------------------------------------------------------------------
                 For the final rule analysis, the agencies did not include CISG
                systems. The effectiveness of CISG systems were similar to the BISG,
                and the high cost of the CISG caused it to be applied infrequently.
                Other packaging and integration issues make it difficult for most
                vehicles to adopt CISG technology. Typically, a manufacturer would have
                to modify the flywheel housing to allow the installation of an electric
                motor, which must also fit where the system is mounted between the
                transmission and the engine block. Space in that part of the vehicle
                also comes at a premium because other components such as exhaust
                systems and piping systems must also be housed in the same area. In the
                final rule analysis, all vehicles previously considered to possess CISG
                technology were instead assigned a BISG system.
                 (d) Strong Hybrids
                 A hybrid vehicle is a vehicle that combines two or more sources of
                propulsion energy, where one uses a consumable fuel (like gasoline),
                and one is rechargeable (during operation, or by another energy
                source). Hybrids reduce fuel consumption through three major
                mechanisms, including (1) potential engine downsizing, (2) optimizing
                the performance of the engine to operate at
                [[Page 24471]]
                the most efficient operating point and under some conditions storing
                excess energy such as by charging the battery, and (3) capturing energy
                during braking and some decelerations that might otherwise be lost to
                the braking system and using the stored energy to provide launch
                assist, coasting, and propulsion during stop and go traffic conditions.
                The effectiveness of the hybrid systems depends on how the above
                factors are balanced, taking into account complementary equipment and
                vehicle application. For some performance vehicles, the hybrid
                technologies are used for performance improvement without any engine
                downsizing.
                 The NPRM analysis evaluated the following strong hybrid vehicles:
                Hybrids with ``P2'' parallel drivetrain architecture (SHEVP2),\1040\
                and hybrids with power-split architecture (SHEVPS). The parallel hybrid
                drivetrain, although enhanced by the electric portion, remains
                fundamentally similar to a conventional powertrain. In contrast, the
                power-split hybrid drivetrain is novel and considerably different than
                a conventional powertrain. Although these hybrid architectures are
                quite different, both types provide start-stop or idle-stop
                functionality, regenerative braking capability, and vehicle launch
                assist. A SHEVPS has a higher potential for fuel economy improvement
                than a SHEVP2, although its cost is also higher.
                ---------------------------------------------------------------------------
                 \1040\ Depending on the location of electric machine (motor with
                or without inverter), the parallel hybrid technologies are
                classified as P0-motor located at the primary side of the engine,
                P1-motor located at the flywheel side of the engine, P2-motor
                located between engine and transmission, P3-motor located at the
                transmission output, and P4-motor located on the axle.
                ---------------------------------------------------------------------------
                 Power-split hybrid (SHEVPS) is a hybrid electric drive system that
                replaces the traditional transmission with a single planetary gear set
                (the power-split device) and a motor/generator. This motor/generator
                uses the engine either to charge the battery or to supply additional
                power to the drive motor. A second, more powerful motor/generator is
                permanently connected to the vehicle's final drive and always turns
                with the wheels. The planetary gear splits engine power between the
                first motor/generator and the drive motor either to charge the battery
                or to supply power to the wheels. During vehicle launch, or when the
                battery state of charge (SOC) is high, the engine, which is not as
                efficient as the electric drive, is turned off and the electric machine
                propels the vehicle. During normal driving, the engine output is used
                both to propel the vehicle and to generate electricity. The electricity
                generated can be stored in the battery and/or used to drive the
                electric machine. During heavy acceleration, both the engine and
                electric machine (by consuming battery energy) work together to propel
                the vehicle. When braking, the electric machine acts as a generator to
                convert the kinetic energy of the vehicle into electricity to charge
                the battery.
                 The Autonomie simulations assumed all SHEVPS' used an Atkinson
                cycle engine (Eng26). Therefore, all vehicles equipped with SHEVPS
                technology in the CAFE model simulations were assumed to have Atkinson
                cycle engines. This Atkinson cycle engine with high compression ratio
                is optimized for efficiency, rather than performance. Accordingly,
                SHEVPS technology as modeled in this analysis was not suitable for
                large vehicles that must handle high loads.\1041\ Further discussion of
                Atkinson engines and their capabilities is discussed in Section VI.C.1
                Engine Paths.
                ---------------------------------------------------------------------------
                 \1041\ Kapadia, J., Kok, D., Jennings, M., Kuang, M. et al.,
                ``Powersplit or Parallel--Selecting the Right Hybrid Architecture,''
                SAE Int. J. Alt. Power. 6(1):68-76, 2017, https://doi.org/10.4271/2017-01-1154.
                ---------------------------------------------------------------------------
                 P2 parallel hybrids (SHEVP2) are a type of hybrid vehicle that uses
                a transmission-integrated electric motor placed between the engine and
                a gearbox or CVT, with a clutch that allows decoupling of the motor/
                transmission from the engine. Although similar to the configuration of
                the CISG system discussed previously, a P2 hybrid would typically be
                equipped with a larger electric machine and battery in comparison to
                the CISG. Disengaging the clutch allows all-electric operation and more
                efficient brake-energy recovery. Engaging the clutch allows efficient
                coupling of the engine and electric motor and, when combined with a
                transmission, reduces gear-train losses relative to power-split or 2-
                mode hybrid systems. P2 hybrid systems typically rely on the internal
                combustion engine to deliver high, sustained power levels. Only low and
                medium power demands are allowed for electric-only mode.
                 In the NPRM CAFE modeling, the SHEVP2 system represented a hybrid
                system paired with an existing engine on a given vehicle, while the
                SHEVPS removed and replaced the previous engine with an Atkinson cycle
                engine. The agencies explained that while many vehicles may use HCR1
                engines as part of a hybrid powertrain, HCR1 engines may not be
                suitable for some vehicles, such as high performance vehicles or
                vehicles designed to carry or tow large loads (this is further
                discussed in Section VI.C.1, Engine Paths). Many manufacturers may
                prefer turbocharged engines (with high specific power output) for P2
                hybrid systems, in order to maintain performance. Accordingly, in the
                NPRM analysis, to satisfy power demands, many SHEVP2 systems were
                paired with non-HCR powertrains.
                 ICCT and Meszler Engineering Services commented that as a result of
                NPRM CAFE model constraints, low-cost, HCR engines were too
                infrequently paired with SHEVP2 technology. These commenters claimed
                that frequent pairing of SHEVP2 with downsized turbocharged engines
                resulted in higher cost and lower effectiveness for these strong
                hybrids.1042 1043
                ---------------------------------------------------------------------------
                 \1042\ Meszler Engineering Services, Attachment 2, Docket No.
                NHTSA-2018-0067-11723, at 15.
                 \1043\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-25.
                ---------------------------------------------------------------------------
                 In consideration of these comments, the final rule analysis
                includes additional strong hybrids (P2HCR0, P2HCR1, and P2HCR2\1044\)
                that use HCR engines in a P2 parallel hybrid system. The SHEVP2
                technology allows the engine type to be inherited from the outgoing
                engine; this is unchanged from the NPRM and provides a good solution
                for vehicles that need to undergo hybridization but require other
                engine technologies (such as turbocharging) to meet performance
                requirements. In addition, this final rule analysis allows any
                conventional engine technology to go to P2HCR strong hybrid technology
                within the set performance requirements. This is further discussed in
                the Section VI.C.3.c), Electrification Adoption Features.
                ---------------------------------------------------------------------------
                 \1044\ P2HCR2 was included in simulations used for sensitivity
                studies, but was excluded in the central analysis simulations for
                reasons surrounding the HCR2 engine, as discussed in Section VI.C.1.
                ---------------------------------------------------------------------------
                (e) Plug-In Hybrids
                 Plug-in hybrid electric vehicles (PHEV) are hybrid electric
                vehicles with the means to charge their battery packs from an outside
                source of electricity (usually the electric grid). These vehicles have
                larger battery packs with more energy storage and a greater capability
                to be discharged than other non-plug-in hybrid electric vehicles. PHEVs
                also generally use a control system that allows the battery pack to be
                substantially depleted under electric-only or blended mechanical/
                electric operation and batteries that can be cycled in charge-
                sustaining operation at a lower state of charge than is typical of
                other hybrid electric vehicles. These vehicles generally have a greater
                all-electric range than the typical SHEVs discussed above. In the NPRM
                analysis,
                [[Page 24472]]
                PHEVs with two all-electric ranges--a 30 mile and a 50 mile all-
                electric range (AER)--were included as technologies that vehicles could
                adopt. The PHEV30 represented a ``blended-type'' plug-in hybrid, which
                can operate in all-electric (engine off) mode only at light loads and
                low speeds, and must blend electric machine and engine power together
                to propel the vehicle at medium or high loads and speeds. The PHEV50
                represented an extended range electric vehicle (EREV), which is capable
                of travelling in all-electric mode even at higher speeds and loads.
                 Unlike other alternative fuel systems that require specific
                infrastructure for refueling or recharging (e.g., hydrogen vehicles or
                rapidly charged battery electric vehicles), PHEV batteries can be
                charged using existing infrastructure, although widespread adoption may
                require upgrades to electrical power distribution systems.\1045\ PHEVs
                are considerably more expensive than conventional vehicles and more
                expensive than SHEVPS technologies because of larger battery packs and
                charging systems capable of connecting to the electric grid.
                ---------------------------------------------------------------------------
                 \1045\ See above for a discussion of electrical vehicle
                infrastructure.
                ---------------------------------------------------------------------------
                 Commenters, such as CARB, stated that in the NPRM analysis the PHEV
                motors were oversized and overpowered, and that model-built PHEV30s
                have excessive battery pack size and electric range when compared to
                actual production vehicles.\1046\ In response to such comments, the
                agencies, in collaboration with Argonne, conducted further market study
                to confirm CARB's observations and determined that replacing PHEV30
                (with a nominal 30 mile AER) with PHEV20 (with a nominal 20 mile AER)
                would more closely characterize the PHEVs actually in production.\1047\
                The agencies therefore elected to replace PHEV30 with PHEV20 in the
                final rule.
                ---------------------------------------------------------------------------
                 \1046\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 150, 153.
                 \1047\ ``ANL response on NPRM comments (PHEV sizing)-
                181112.pptx,'' available in Docket No. NHTSA-2018-0067.
                ---------------------------------------------------------------------------
                 The final rule also includes four additional types of plug-in
                hybrids; two additional plug-in hybrids were added to allow the use of
                turbocharged engines (PHEV20T, PHEV50T), and two additional plug-in
                hybrids were added to provide maximum efficiency by utilizing an
                Atkinson cycle engine (PHEV20H, PHEV50H).
                 In practice, many PHEVs recently introduced in the marketplace use
                turbocharged engines in the PHEV system, and this is particularly true
                for PHEVs produced by European manufacturers and for other PHEV
                performance vehicle applications. However, the NPRM Autonomie
                simulations (and thus all the CAFE model simulations) assumed all PHEVs
                used a naturally aspirated, Atkinson cycle engine. The agencies
                determined through continued marketplace observation that PHEV vehicles
                should indeed be allowed to adopt or retain turbocharged engines. Also,
                BorgWarner commented that modeling of PHEVs should include turbocharged
                engines, since these engines can be downsized to reduce vehicle mass
                and fit into smaller engine compartments, and offer efficiency and
                performance advantages especially when paired with a higher expansion
                ratio.\1048\ Thus, in addition to the PHEV20 and PHEV30, the final rule
                analysis included PHEV20T and PHEV50T variations which are,
                respectively, 20 and 50 mile all electric range PHEVs with turbocharged
                engines.
                ---------------------------------------------------------------------------
                 \1048\ BorgWarner, Attachment 2, Docket No. NHTSA-2018-0067-
                11873, at 150,153.
                ---------------------------------------------------------------------------
                 This final rule also added PHEV20H and PHEV50H, although
                effectively these are not used by the model simulations. These plug-in
                types represent 20 and 50 mile all electric range plug-in hybrids that
                use particularly efficient high-compression, Atkinson cycle engines.
                These were added with the intent to provide PHEVs with a maximum level
                of fuel economy at a lower cost. However, they proved to be too similar
                to existing plug-in technology choices and were thus assigned identical
                characteristics as the PHEV20 and PHEV50. In this final rule analysis,
                PHEV20 and PHEV50 sizing were updated and so the similarities in
                performance between different engines converged. For further discussion
                on PHEV sizing, see Section VI.C.3.d), Electrification Effectiveness
                Modeling and resulting Effectiveness values.\1049\ The PHEV20H and
                PHEV50H technologies are still considered by the CAFE model but they
                remain as ``placeholders'' for potential incorporation in future
                analyses.
                ---------------------------------------------------------------------------
                 \1049\ This final rule analysis used Atkinson Engine for PHEVPS
                electrified vehicles. The components such as electric motor and
                engine power in these hybrid systems were sized in ways to meet
                vehicle class performance characteristics and efficiency. And after
                these vehicle components were sized, the Atkinson engines in these
                vehicles were operating in similar efficiency as HCR engines as the
                full vehicle modeling and simulation. As discussed in PO 06 C.1.c.1
                Non-HEV Atkinson Engine Modes, power-split hybrid-based Atkinson
                engines attempt to operate in the most efficient regions while using
                electric motors to meet deficiencies in performance. And so, PHEV20H
                and PHEV50H HCR engines compared to PHEV20 and PHEV50 Atkinsons
                engines would have be sized to operate in the most efficiency
                regions and the thermal efficiency between these two set of
                combinations would have had similar efficiency for this analysis.
                ---------------------------------------------------------------------------
                (f) Battery Electric Vehicles
                 Electric vehicles (EVs), or battery electric vehicles (BEVs) are
                equipped with all-electric drive and with systems powered by energy-
                optimized batteries charged primarily from grid electricity. The range
                of a battery electric vehicle depends on the vehicle's class and the
                battery pack size. The NPRM analysis included BEVs with a range of 200
                miles.
                 Following the NPRM, the agencies conducted continued market
                analysis of production BEVs, and observed a growing number of vehicles
                with nominal ranges above 200 miles. CARB also commented that certain
                BEVs modeled as BEV200 in the NPRM in fact had ``well over 200 miles of
                range.'' \1050\ The agencies thus concluded that a 300-mile-range
                BEV300 should be included in the final rule to represent better these
                higher-range electric vehicles as well as a potential future range
                alternative more comparable to IC engines. The agencies still believe
                that, in the rulemaking timeframe, BEV300 will be the most cost
                effective extended range BEVs that could be available for adoption.
                Longer-range electric vehicles could have been modeled in the analysis,
                but the compliance simulation would likely not have selected the
                longer-range vehicle if lower-range vehicles were still available. This
                is because the CAFE model only applies technologies until a
                manufacturer meets its CAFE or CO2 standard, and the BEV200
                and BEV300 vehicles operate functionally the same in helping a
                manufacturer towards meeting its compliance obligations. The only
                difference between these vehicles is cost. As discussed further in
                Section VI.C.3.c), the agencies used phase-in caps to control expected
                BEV200 and BEV300 penetration based on the current trend and future
                assumption that consumers will transition towards longer-range electric
                vehicles.
                ---------------------------------------------------------------------------
                 \1050\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 147.
                ---------------------------------------------------------------------------
                (g) Fuel Cell Vehicles
                 Fuel cell electric vehicles (FCEVs or FCVs) utilize a full electric
                drive platform but consume hydrogen fuel to generate electricity in an
                onboard fuel cell. Fuel cells are electrochemical devices that directly
                convert reactants (hydrogen and oxygen via air) into electricity, with
                the potential of achieving more than twice the efficiency of
                conventional internal combustion engines. High pressure gaseous
                hydrogen storage tanks are used by most
                [[Page 24473]]
                automakers for FCEVs. These high-pressure tanks are similar to those
                used for compressed gas storage in more than 10 million CNG vehicles
                worldwide, except that they are designed to operate at a higher
                pressure (350 bar or 700 bar vs. 250 bar for CNG), and to contain the
                very small, and very flammable, gaseous hydrogen molecule. FCEVs are
                currently produced in limited numbers and are available in limited
                geographic areas.
                (2) Electrification Pathways
                 The electrification technologies described above were applied in
                the CAFE model through a number of technological pathways. Three main
                electrification technology pathways were modeled: The Electric
                Improvements Path, the Electrification Path, and the Hybrid/Electric
                Path. These three electrification pathways are evaluated in parallel by
                the CAFE model; the model can consider any of the three right away, and
                does not need to go ``through'' one pathway in order to begin
                evaluating another. Any superseded technology is also disabled whenever
                a succeeding technology is applied to a vehicle, even if a specific
                superseded technology was not previously utilized on that vehicle. As
                previously explained, this requirement exists so that the modeling
                system does not downgrade technologies during analysis.
                 The Electrics Improvements Path defined in the NPRM and final rule
                is shown in Figure VI-29 below, which starts with EPS and progresses to
                IACC. While these two electrified-accessory technologies are mutually
                exclusive, either one can be modularly paired with any other
                technology, including those in the other electrification pathways.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.203
                 The Electrification Path shown in Figure VI-29 allows a
                conventional powertrain to become a micro-hybrid with SS12V, or a mild
                hybrid with BISG, or CISG (which is no longer available for the final
                rule analysis, as discussed previously) technologies. All three of the
                Electrification Path technologies are mutually exclusive with respect
                to all conventional powertrain technologies, as well as technologies
                contained in the Hybrid/Electric path discussed below. The model first
                evaluates SS12V, and then progresses to BISG or CISG (NPRM-only). The
                conventional engine technology CONV is grayed out to indicate that the
                model uses information about the previous conventional (non-
                electrified) powertrain to map properly to simulation results found in
                the vehicle simulation database. Although the adoption of these
                technologies will classify a vehicle as a micro/mild hybrid (MHEV) and
                no longer a conventional (CONV), the vehicle is allowed to retain the
                engine and transmission technologies possessed before entering the
                Electrification Path.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.204
                 The Hybrid/Electric Pathways are shown in Figure VI-30. Both the
                NPRM and final rule Hybrid/Electric paths begin at the ``strong
                hybrid'' technology types, each of which is mutually exclusive of the
                others; once one is chosen, the other is eliminated from future
                selection for that vehicle. The paths then progress into plug-in
                hybrids and then culminate with the mutually exclusive battery electric
                vehicles or fuel cell vehicles. The additional final rule technologies
                described above can be found in the final rule Hybrid/Electric pathway
                on the right side of Figure VI-31, in comparison to the NPRM
                technologies shown on the left
                [[Page 24474]]
                side of the figure.\1051\ The hybrid/electric pathways contains
                multiple ``roots,'' or starting points, which force a vehicle to remain
                within the branches of a chosen root. For example, the final rule
                hybrid/electric pathway has three roots: SHEVP2, SHEVPS, and P2HCR0. If
                a vehicle uses SHEVPS, then SHEVP2 technology and the entire P2HCR0
                through PHEV50H branch will be disabled from further consideration. In
                other words, from one technology in the pathway, a vehicle can only
                move forward along any of the indicated arrows, and never in the
                reverse direction. Also, when using any technology in the Hybrid/
                Electric pathway, with the exception of SHEVP2, all engine and
                transmission technologies as well as the Electrification Path
                technologies shown in Figure VI-31 are prohibited. SHEVP2 is an
                exception because it allows engine technologies previously held by the
                vehicle to be inherited into the parallel hybrid system.
                ---------------------------------------------------------------------------
                 \1051\ Note that the NPRM Hybrid/Electric Path (left side of
                Figure I-3) refers to a portion of the path containing plug-in
                hybrids and electric vehicles as the ``Advanced Hybrid/Electric
                Path.'' For this discussion, we will simply refer to the entire
                collection of these technologies, including the ``Advanced''
                technologies, as the ``Hybrid/Electric Path.''
                [GRAPHIC] [TIFF OMITTED] TR30AP20.205
                b) Electrification Analysis Fleet Assignments
                 Since the 2012 rulemaking, manufacturers have implemented a number
                of powertrain electrification technologies, including 48V mild hybrid,
                strong HEV, PHEV, and BEV powertrains.1052 1053 For the NPRM
                analysis, the agencies identified the specific electrification
                technologies in each vehicle model in the MY 2016 analysis fleet, and
                used those technology levels as the starting point for the regulatory
                analysis. The agencies assigned electrification technology levels based
                on manufacturer-submitted CAFE compliance information, vehicle
                technical specifications released publicly by manufacturers, agency-
                sponsored vehicle benchmarking studies, technical publications, and
                manufacturer CBI.\1054\ For the final rule analysis, the agencies used
                a similar process and data sources to identify the electrification
                technologies in the MY 2017 analysis fleet.\1055\
                ---------------------------------------------------------------------------
                 \1052\ ``The 2018 EPA Automotive Trends Report,'' https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends, Accessed Aug 23, 2019.
                 \1053\ FOTW #1108, Nov 18, 2019: Fuel Economy Guide Shows the
                Number of Conventional Gasoline Vehicle Models Achieving 45 miles
                per gallon or Greater is Increasing. DOE VTO. Available at https://www.energy.gov/eere/vehicles/articles/fotw-1108-november-18-2019-fuel-economy-guide-shows-number-conventional. Last accessed Nov 18,
                2019.
                 \1054\ NPRM Market Data central analysis input file.
                 \1055\ FRM Market Data central analysis input file.
                ---------------------------------------------------------------------------
                 The agencies received comments regarding the application of
                electrification technologies in the MY 2016 analysis fleet. Commenters,
                such as the California Air Resources Board, stated the agencies
                mischaracterized some hybrid technologies, such as power-split and P2
                hybrid architectures.\1056\ Specifically CARB was concerned about the
                ``misclassification of the 2016 Chevrolet Malibu Hybrid as having a P2
                hybrid,'' noting the Malibu shared many of its drivetrain components
                with the 2016 Chevy Volt, a vehicle classified as a power-split HEV.
                ---------------------------------------------------------------------------
                 \1056\ Comments from CARB, Attachment 2, NHTSA Docket No. NHTSA-
                2018-0067-11873, at 136.
                ---------------------------------------------------------------------------
                 BorgWarner stated that the ``modeling should be inclusive of all
                approaches of PHEV and HEV and not be limited only to Atkinson Cycle
                engines,'' suggesting that it was appropriate for the NPRM analysis to
                include turbocharged engines in combination with PHEV and HEV
                technologies.\1057\
                ---------------------------------------------------------------------------
                 \1057\ Comments from BorgWarner, Attachment 1, Appendix, NHTSA
                Docket No. NHTSA-2018-0067-11895, at 10.
                ---------------------------------------------------------------------------
                 The agencies agree with the underlying issue identified by both
                CARB and BorgWarner's comments. In
                [[Page 24475]]
                both cases a limitation of modeling classification, and not a lack of
                academic understanding of HEV systems, is the crux of the issue. In the
                specific case of the 2016 Chevy Malibu, the electrical architecture is
                a power split, however, the vehicle uses a non-Atkinson, basic direct
                injection engine. These characteristics put the Malibu HEV in an
                overlap with the powertrain models used to represent HEV systems in the
                agencies' analysis. If the system had been classified as a PS HEV
                system in the analysis fleet, the engine would have incorrectly been
                modeled as an Atkinson engine, resulting in overestimation of the
                baseline system's level of efficiency and technology applied. The
                overestimation of the baseline fleet model would have limited the
                potential for the baseline system to improve over the timeframe of the
                analysis. With the system classified as the P2 HEV, the engine can be
                accurately modeled while still accounting for the benefits of an HEV
                system. This allowed the platform the full potential for technology and
                efficiency improvement in the analysis.
                 The agencies considered the issues identified in comments and
                reviewed the MY 2017 analysis fleet information to determine what
                changes could improve the final rule analysis. The agencies determined
                that expanding the number of electrification technologies would address
                the CARB and BorgWarner comments, as well as the comments from others
                that are discussed in Section VI.C.3.a)(1) Electrification
                Technologies. The agencies increased the number of unique
                electrification technologies from twelve in the NPRM to eighteen for
                the final rule analysis. The expanded list enabled greater precision in
                the assignment of technologies to the MY 2017 analysis fleet, and
                enabled the agencies to characterize the electrification technologies
                found in the fleet accurately and realistically. The expanded list also
                provided more granularity for the application of technologies for the
                rulemaking analysis. Table VI-85 shows the full list of electrification
                technologies for the final rule analysis.
                 This collection of technologies represents the best available
                information the agencies have, at the time of this action, regarding
                both currently available electrification technologies and
                electrification technologies that could be feasible for application to
                the U.S. fleet during the rulemaking timeframe. The agencies believe
                this effort has yielded the most accurate analysis fleet utilized for
                rulemakings to date.
                 As discussed in the previous section and shown in Figure VI-29,
                Figure VI-30, and Figure VI-31, electrification may be added to
                vehicles as shown on the decision tree pathways. Further application of
                electrification technologies to vehicle platforms was dependent on
                electrification technology already present on vehicles in the MY 2017
                analysis fleet. Electrification may also be predicated on whether a
                vehicle has a dedicated platform that accommodates battery electric
                capability or whether a platform is designed (``package protected'')
                \1058\ to enable the addition of some form(s) of hybridization. The
                agencies' assessment of each existing platform's capability to adopt
                electrification technologies is identified in the CAFE model market
                data input file.\1059\
                ---------------------------------------------------------------------------
                 \1058\ `Package Protected' is an automotive industry term used
                to describe the purposeful design of a vehicle to include space and
                weight allowances for future technology additions.
                 \1059\ FRM Market Data central analysis input file.
                ---------------------------------------------------------------------------
                c) Electrification Adoption Features
                 In the NPRM and final rule analysis, electrification adoption
                features were applied in multiple ways. First, when an electrification
                technology is selected, a path logic is applied that dictates what
                other technologies are either superseded or mutually exclusive to the
                applied technology. For a detailed discussion of path logic for the
                final rule analysis, including technology supersession logic and
                technology mutual exclusivity logic, please see CAFE model
                documentation section. Second, application of the more advanced
                electrification technologies, such as the strong hybrids, plug-in
                hybrids, and full BEVs, result in major changes to the whole
                powertrain. The changes to the powertrain include substitution of
                transmission and engine technologies, and accordingly these
                technologies can only be applied at a vehicle redesign, as shown in
                Table VI-85 below. Finally, some of electrification technologies are
                restricted from application to certain vehicle classifications. These
                restrictions will be discussed under the specific technology sections.
                 The fully-electric technologies, BEV technology and FCV technology,
                qualify as alternative fuel technologies. As a result, these
                technologies are not considered during portions of the agencies'
                analysis. Specifically, the exclusion of dedicated alternative fuel
                technology from NHTSA's analysis of potential fuel economy standards is
                a result of statutory obligations prescribed under EPCA/EISA.\1060\
                However, NHTSA performed two fuel economy analyses, a standard-setting
                analysis that constrained the use of the technologies, and an
                unconstrained analysis that did not exclude the technologies, which
                provides an estimation of real-world environmental impacts used as
                inputs for the Environmental Impact Statement (EIS). The unconstrained
                analysis included the alternative fuel technologies, and used the
                adoption features for BEVs and FCVs discussed below. Further, for
                purposes of analyzing EPA's tailpipe CO2 emissions
                rulemaking pursuant to the Clean Air Act, consideration of these
                technologies is likewise unconstrained. For a detailed discussion of
                the analysis versions and statutory obligations please refer to Section
                VI.A Analytical Approach as Applied to Regulatory Alternatives,
                Overview of Methods and Section VI.A.4 Compliance Simulation.
                ---------------------------------------------------------------------------
                 \1060\ 49 U.S.C. 32902(b)(1). A ``dedicated automobile'' is
                defined in 49 U.S.C. 32901 as ``an automobile that only operates on
                alternative fuel.''
                ---------------------------------------------------------------------------
                 The exclusion of the BEV and FCV technology from the standard-
                setting analysis resulted in a comment from ICCT. ICCT stated, ``the
                agencies prevented their fleet compliance model from allowing battery
                electric vehicles from being applied in their analysis of the Augural
                standards.'' \1061\ The agencies believe this reflects a
                misunderstanding of NHTSA's statutory obligation under EPCA/EISA and
                how the agencies ran the analysis. NHTSA did consider alternative
                fueled vehicles in the unconstrained analysis--but as discussed further
                in Section VIII, is prohibited from considering the availability of
                such technologies when setting maximum feasible standards.
                ---------------------------------------------------------------------------
                 \1061\ Comments from ICCT, Attachment 3, Appendix, NPRM Docket
                No. NHTSA-2018-0067-11741, at 182.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24476]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.206
                BILLING CODE 4910-59-C
                (1) Micro and Mild Hybrid
                 For the NPRM and final rule analysis, the only adoption features
                for the SS12V and BISG technologies were functions of path logic. The
                SS12V and BISG technologies were allowed for consideration in any
                existing vehicle configuration that did not already have a more
                advanced electrification technology applied. Per Table VI-85 above, the
                BISG technology was considered more advanced than the SS12V technology.
                 Meszler Engineering commented that 48V batteries used in
                conjunction with 12 volt systems (what are referred to in the analysis
                as BISG systems) are one example of a ``bolt-on'' technology that can
                be added to a vehicle during a product refresh without causing
                [[Page 24477]]
                production problems or significantly increasing costs.\1062\ Meszler
                Engineering stated that 48V systems do not require reengineering of the
                engine and can be added at any time during a model's lifespan, as shown
                by key suppliers that are expanding production capacity to meet
                customer demand for the technology.\1063\ Meszler Engineering also
                pointed to examples of vehicles that utilize 48V systems, including
                high-volume non-luxury vehicles like the Ram pickup truck, Jeep
                Wrangler, and Ford F-150.\1064\
                ---------------------------------------------------------------------------
                 \1062\ Comments by Meszler Engineering, Attachment 4 CAF[Eacute]
                Model Redesign and Refresh Rates, NHTSA Docket No. NHTSA-2018-0067-
                11723, at 2-4. (citing A.K. Kumawat and A.K. Thakur, A Comprehensive
                Study of Automotive 48V Technology, SSRG International Journal of
                Mechanical Engineering (SSRG-IJME), Vol. 4 (5) (May 2017), available
                at: https://jalopnik.com/everything-you-need-to-know-about-the-upcoming-48-volt-1790364465 (last viewed 10/23/2018)).
                 \1063\ Comments by Meszler Engineering, Attachment 4 CAFE Model
                Redesign and Refresh Rates, NHTSA Docket No. NHTSA-2018-0067-11723,
                at 2-4.
                 \1064\ Comments by Meszler Engineering, Attachment 4 CAFE Model
                Redesign and Refresh Rates, NHTSA Docket No. NHTSA-2018-0067-11723,
                at 2-4.
                ---------------------------------------------------------------------------
                 The agencies disagree with Meszler Engineering's assessment of 48V
                technology as a ``bolt-on'' technology. Although BISG systems represent
                a first step in vehicle electrification, and the number of components
                involved is fewer than most other types of hybrid systems, a BISG
                system still requires engineering and packaging of motors, cooling
                systems, additional wiring harnesses from the 48V battery pack to the
                motors, control systems, and other components incorporated into the
                front engine compartment. Further, the addition of a BISG system
                requires recalibration and validation of numerous engine performance
                parameters, including emissions controls, balancing torque supply to
                the transmission between the BISG system and engine, and noise-
                vibration-harshness controls. In addition, the examples Meszler
                Engineering provided support the agencies' designation of SS12V and
                BISG systems as redesign technologies; the BISG system in the MY 2019
                Ram pickup and in the MY 2018 Jeep Wrangler were introduced during a
                product redesign and not during a mid-cycle product
                refresh.1065 1066 Although Ford has indicated that the F-150
                will include hybrid variants,\1067\ the agencies do not have
                information about specific plans for a 48V system on the F-150. In
                consideration of this information, the agencies maintained the redesign
                schedule for mild hybrids for the final rule analysis.
                ---------------------------------------------------------------------------
                 \1065\ See, e.g., K.C. Colwell, The 2019 Ram 1500 eTorque Brings
                Some Hybrid Tech, If Little Performance Gain, to Pickups, Car and
                Driver (Mar. 14, 2019), available at: https://www.caranddriver.com/reviews/a22815325/2019-ram-1500-etorque-hybrid-pickup-drive/ (``Any
                2019 Ram 1500--the all-new one, not the Ram Classic that is just a
                continuation of the previous generation--can be equipped with a
                motor/generator attached to its engine's crankshaft via a belt that
                is capable of adding torque, cranking the engine in a stop/start
                event, or making electricity with regenerative braking.'').
                 \1066\ See, e.g., Tony Quiroga, The 2018 Jeep Wrangler Hybrid
                Provides Effortless Thrust, Much Improved Fuel Economy, Car and
                Driver (Oct. 15, 2018), available at: https://www.caranddriver.com/reviews/a23746585/2018-jeep-wrangler-unlimited-suv-turbo-four-cylinder-hybrid/ (``Completely redesigned for 2018, the Wrangler is
                even more like a Power Wheels now that it's available with an
                electric motor.'').
                 \1067\ ``Ford to Invest more than $1.45 Billion, Add 3,000 Jobs
                in SE Mich. Plants to Deliver New Pickups, SUVs, EVS, and AVS,''
                Ford Media Center, 17 Dec 2019. https://media.ford.com/content/fordmedia/fna/us/en/news/2019/12/17/ford-invests-adds-jobs-southeast-michigan-plants.html.
                ---------------------------------------------------------------------------
                (2) Strong Hybrids--SHEVP2, SHEVPS, P2HCR0, P2HCR1, P2HCR2
                 NPRM adoption features applied to strong hybrid technologies
                included path logic, powertrain substitution, and vehicle class
                restrictions. For the NPRM analysis technologies on the Hybrid/Electric
                path (SHEVP2 and SHEVPS) were defined as stand-alone and mutually
                exclusive. When the modeling system applies one of those technologies,
                the other one is immediately disabled from future application. Once a
                strong hybrid technology is applied it also supersedes lower
                technologies on the electrification path, allowing future application
                of technology to consider only more advanced forms of electrification.
                 In the NPRM when the SHEVP2 technology or the SHEVPS technology
                were applied, the transmission technology was superseded. Regardless of
                the transmission technology present when the technology was applied,
                the transmission technology was replaced by either the AT6 or DCT6. The
                specific transmission technology selected was based on choosing the
                best cost versus effectiveness.
                 During the NPRM analysis when the SHEVP2 technology was selected
                the engine technology for the platform was maintained. However, the
                engine technology was locked at the current level and could not be
                changed. For the SHEVPS technology the existing engine was replaced
                with an Atkinson cycle engine (Eng26).
                 The SHEVPS was also constrained from application to particular
                vehicle technology classes or vehicles with specific performance
                characteristics in the NPRM. Application of the power-split
                architecture was restricted from high performance vehicles and vehicles
                with a high towing capability requirements.\1068\ These constraints
                prevented application to the pick-up and performance pick-up class of
                vehicles. The constraints also prevented application to any platform
                with a base horsepower rating greater than 400 HP. Additional platforms
                determined to be purpose built as performance platforms were also
                restricted from receiving SHEVPS technology.
                ---------------------------------------------------------------------------
                 \1068\ Kapadia, J., Kok, D., Jennings, M., Kuang, M. et al.,
                ``Powersplit or Parallel--Selecting the Right Hybrid Architecture,''
                SAE Int. J. Alt. Power. 6(1):68-76, 2017, https://doi.org/10.4271/2017-01-1154.
                ---------------------------------------------------------------------------
                 Comments from ICCT criticized the manner in which SHEVP2 technology
                was applied to a platform. ICCT stated ``the benefits of level-2
                transmission efficiency and TURBO2 over TURBO1 are removed when P2
                strong hybrid systems (SHEVP2) are selected on the electrification
                pathway.'' \1069\
                ---------------------------------------------------------------------------
                 \1069\ Comments from ICCT, Attachment 3, 15 page summary and
                full comments appendix, NPRM Docket No. NHTSA-2018-0067-11741, at
                I25.
                ---------------------------------------------------------------------------
                 Additional comments regarding the adoption features of the SHEVP2
                technology were received from Meszler Engineering and ICCT. Meszler
                argued that the locking of engine technologies when a manufacturer
                selects the SHEVP2 technology may preclude the selection of a more
                cost-effective engine technology.\1070\ This concern was echoed by
                ICCT, who also felt the engine technology lock-in artificially
                increased cost for effectiveness on the overall SHEVP2 technology
                packages.\1071\ Both commenters specifically wanted an option for a
                high compression ratio engine technology to be considered in place of
                any advanced engine technology carried into the SHEVP2 technology
                pathway.
                ---------------------------------------------------------------------------
                 \1070\ Comments from Meszler Engineering Services, Attachment 2,
                NPRM Docket No. NHTSA-2018-0067-11723, at 15-16.
                 \1071\ Comments from ICCT, Attachment 3, 15 page summary and
                full comments appendix, NPRM Docket No. NHTSA-2018-0067-11741, at
                I25-I26.
                ---------------------------------------------------------------------------
                 The agencies agreed with the need for maintaining the benefits of a
                higher transmission technology, and for the final rule analysis a AT8L2
                transmission technology replaced the AT6 or DCT6 transmissions for all
                hybrid-electric technologies. The AT8L2 was selected as the optimal
                transmission technology point for HEV systems. The transmission
                technology point was selected based on observed diminishing returns for
                applying advanced transmission technologies to advanced engine/
                powertrains.\1072\
                ---------------------------------------------------------------------------
                 \1072\ 2015 NAS Report--The National Academy of Science, in
                their 2015 report, noted that ``as engines incorporate new
                technologies to improve fuel consumption, the benefits of increasing
                transmission ratios or switching to a CVT diminish.''
                ---------------------------------------------------------------------------
                [[Page 24478]]
                 The agencies also reconsidered engine options for SHEVP2
                technology, and other strong hybrid-electric technologies. The agencies
                agreed with Meszler and ICCT's observation and instituted new P2 engine
                technology options, as discussed above. For the final rule analysis,
                when a platform considered the SHEVP2 option, the platform also
                compared maintaining the current engine technology, or selecting an HCR
                technology. If the SHEVP2 system chooses to apply a HCR engine, the
                system diverts to the new electrification sub-path of technologies that
                includes the P2HCR0, P2HCR1, and P2HCR2.
                 The P2HCR path introduced in the final rule analysis had similar
                constraints as the SHEVPS. Performance vehicles and vehicles with a
                high towing requirement were restricted from selection of the P2HCR
                technology. Restrictions that were applied used the same criteria
                described for the SHEVPS.
                (3) Plug-In Hybrids--PHEV20/30, PHEV50, PHEV20T, PHEV50T, PHEV20H,
                PHEV50H
                 The plug-in hybrid options in the NPRM included PHEV30 and PHEV50
                technologies. The plug-in technologies superseded the micro, mild, and
                strong hybrid electrification technologies and could only be replaced
                by full electric technologies. The path logic also allowed a PHEV30 to
                progress to a PHEV50.
                 In the NPRM, when a platform progressed to the plug-in hybrid
                technologies the powertrain was automatically modified. The engine
                technology was replaced by a high compression ratio engine (Eng26) and
                the transmission was replaced by the AT6 or DCT6 technology.
                 PHEV30 and PHEV50 were also constrained from application to
                vehicles with the potential for high towing demands.\1073\ This
                constraint was applied by restricting access to the pickup truck
                vehicle technology class. Additional specific vehicle platforms were
                restricted based on engineering judgment.
                ---------------------------------------------------------------------------
                 \1073\ Power split or Parallel-selecting the Right Hybrid
                Architecture: SAE 2017-01-1154. = Kapadia, J., Kok, D., Jennings,
                M., Kuang, M. et al., ``Powersplit or Parallel--Selecting the Right
                Hybrid Architecture,'' SAE Int. J. Alt. Power. 6(1):68-76, 2017,
                https://doi.org/10.4271/2017-0-1154.
                ---------------------------------------------------------------------------
                 Comments were received regarding the options for PHEV battery-
                electric technology. The comments are presented and discussed in
                Section VI.C.3.e) Electrification Technologies above, and resulted in
                the creation of additional technology options for plug-in hybrids, as
                well as a modification of available ranges. Comments were also received
                regarding the engine and transmission options used in the
                electrification technologies, these comments are also presented and
                discussed above in Section VI.C.3.e) Electrification Technologies.
                 For the final rule analysis, the plug-in hybrid options included
                PHEV20, PHEV50, PHEV20T, PHEV50T, PHEV20H, and PHEV50H. As with the
                NPRM, the plug-in technologies superseded the micro, mild, and strong
                hybrid technologies. For the final rule analysis, plug-in hybrid
                technologies were also mutually exclusive, and the PHEV20 technologies
                can progress to the PHEV50 technologies.
                 When a platform applied plug-in hybrid technologies in the final
                rule analysis, the engine and transmission technologies are superseded.
                For all plug-in technologies, an AT8L2 transmission is used. For the
                PHEV20/50 and PHEV20/50H, the engine is replaced by an Atkinson cycle
                based engine (Eng26). For the PHEV20/50T, the engine is replaced by the
                TURBO1 technology engine (Eng12).
                 The PHEV20/30 and PHEV20/50H path also had similar constraints as
                the SHEVPS in the final rule analysis. Performance vehicles and
                vehicles with a high towing requirement were restricted from selection
                of the PHEV20/30 and PHEV20/50H technologies. Restrictions that were
                applied used the same criteria described for the SHEVPS.
                (4) Battery Electric Vehicles
                 For the NPRM analysis, the BEV200 technology was applied as an end-
                of-path technology. The BEV200 technology was the only battery electric
                vehicle option. For the final rule analysis, the BEV300 was added as a
                technology option beyond the BEV200, as discussed in Section
                VI.C.3.a)(1)(f) Battery Electric Vehicles. BEV200 and BEV300 technology
                was applied in place of all engine and transmission technologies, and
                was an end of path technology.
                 For the final rule analysis, both the BEV 200 and BEV300 had phase-
                in cap limitations applied based on an analysis of the market
                availability and cost of batteries.\1074\ The BEV200 was limited to a
                greater extent than the BEV300, accounting for expected limits in
                market demand for the shorter-range BEV.\1075\ The phase-in capacity
                numbers were determined based on the results of the analysis of the
                National Energy Model System (NEMS) discussed in Section
                VI.D.1.b)(1)(b) Macroeconomic assumptions used to analyze economic
                consequences of the final rule.
                ---------------------------------------------------------------------------
                 \1074\ John Elkin, MIT finds that it might take a long time for
                EVs to be as affordable as you want, Digital Trends (November 23,
                2019), https://www.digitaltrends.com/cars/mit-study-finds-ev-market-will-stall-in-the-2020s/.
                 \1075\ MIT Energy Initiative. 2019. Insights into Future
                Mobility. Cambridge, MA: MIT Energy Initiative. http://energy.mit.edu/insightsintofuturemobility.
                ---------------------------------------------------------------------------
                (5) Fuel Cell Vehicle
                 For the NPRM analysis, FCV technology was also applied as an end of
                path technology. The FCV technology was also applied as end of path
                technology in the final rule analysis.
                 For the final rule analysis, a phase-in cap was assigned to FCV
                technology. The phase-in cap was assigned based on existing market
                share as well as an analysis of expected infrastructure availability
                during the time frame of regulation.1076 1092
                ---------------------------------------------------------------------------
                 \1076\ ``The 2018 EPA Automotive Trends Report,'' https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends. Last accessed Aug 23, 2019.
                ---------------------------------------------------------------------------
                d) Electrification Effectiveness Modeling and Resulting Effectiveness
                Values
                 For this analysis, the agencies considered a range of
                electrification technologies which, when modeled, resulted in varying
                levels of effectiveness at reducing fuel consumption. Each technology
                consists of many different complex sub-systems with unique component
                efficiencies and operational modes. As discussed further below, the
                systems that contribute to the effectiveness of an electrified
                powertrain in the analysis include the vehicle's battery, electric
                motors, power electronics, and accessory load. Procedures for modeling
                each of these sub-systems are discussed below, and also in Section
                VI.B.3 Technology Effectiveness Values and in the FRM Argonne Model
                Documentation.
                 The modeled electrification technologies included micro hybrids,
                mild hybrids, strong hybrids, plug-in hybrids, and full electric
                vehicles. This section discusses how Autonomie was used to model these
                technologies' effectiveness. The models for the micro hybrids included
                a SS12V system model; mild hybrid models included BISG system models
                and CISG system models; strong hybrid models included SHEVP2 system
                models and SHEVPS system models; and finally, electric vehicle models
                included BEV system models and FCV system models.
                [[Page 24479]]
                (1) Electric Motors, Power Electronics and Accessory Load
                 Each electrified powertrain type possesses a unique effectiveness
                for reducing fuel consumption. Autonomie determines the effectiveness
                of each electrified powertrain type by modeling the basic components,
                or building blocks, found in each powertrain, and then combining the
                components modularly to determine the overall efficiency of the entire
                powertrain. The basic building blocks that comprise an electrified
                powertrain in the analysis included the battery, electric motors, power
                electronics, and accessory loads. Autonomie identified which components
                comprise each electrified powertrain type, and how these components are
                interlinked within each unique electrified powertrain architecture.
                This creates a model for each electrified powertrain architecture that
                simulates how efficiently energy is transferred through each system.
                For example, Autonomie determines a BEV's overall efficiency by
                considering the efficiencies of the battery, the electric traction
                drive system (the electric machine and power electronics) and
                mechanical power transmission devices. Or, for a SHEVP2, Autonomie
                combines a very similar set of components to model the electric portion
                of the hybrid powertrain, and then also includes the combustion engine
                and related power transmission components.
                 For the NPRM and this final rule analysis, Autonomie employed a set
                of electric motor efficiency maps, which originated from two Oak Ridge
                National Laboratory (ORNL) studies: one for a traction motor and an
                inverter, the other for a motor/generator and
                inverter.1077 1078 Autonomie also used test data validations
                from technical publications to determine the efficiency of certain
                electric motors. The electric motor efficiency maps are visual
                measurements of percent efficiency of power as a function of torque and
                motor RPM, and were based on representative production vehicles,
                especially for base and maximum speeds as well as maximum torque curve.
                The maps were used to determine the efficiency characteristics of the
                motors, but were scaled such that their peak efficiency value
                corresponded to the latest state of the art technologies for different
                electrified powertrains. The maps also included some of the losses due
                to power transfer through the electric machine.\1079\ Table VI-86
                details the electric machine efficiency map sources for the different
                powertrain configurations used for the NPRM.
                ---------------------------------------------------------------------------
                 \1077\ See PRIA, at 374.
                 \1078\ Oak Ridge National Laboratory (2008). Evaluation of the
                2007 Toyota Camry Hybrid Synergy Drive System. Submitted to the U.S.
                Department of Energy; Oak Ridge National Laboratory (2011). Annual
                Progress Report for the Power Electronics and Electric Machinery
                Program.
                 \1079\ See Chapters 4.7 and 5.5 in the FRM ANL Model
                Documentation.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.207
                 For the final rule, the agencies used the same efficiency maps as
                the NPRM, except for BEVs. The agencies updated the BEV electric motor
                efficiency for the final rule analysis using data from a more recent
                technical publication.\1081\ The agencies also scaled the maps to have
                peak efficiencies ranging from 96-98 percent depending on the
                powertrain type.\1082\ Table VI-87 below shows powertrain types and the
                source of data used for the final rule.
                ---------------------------------------------------------------------------
                 \1080\ Burak Ozpineci, Oak Ridge National Laboratory Annual
                Progress Report for the Power Electronics and Electronic Motors
                Program, ORNL/SPR-2014/532, https://info.ornl.gov/sites/publications/Files/Pubs3253422.pdf, November 2014. (Nissan Leaf data
                was used for FCV powertrain type).
                 \1081\ Faizul Momen, Electric Motor Design of General Motors'
                Chevrolet Bolt Electric Vehicle, 2016-01-1228, SAE International,
                April 5, 2016.
                 \1082\ See. Chapter 5.5 in FRM ANL Model Documentation.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24480]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.208
                 Battery performance data (e.g., internal resistance, open circuit
                voltage) were measured using individual cell testing on a bench using
                standard test procedures, and BatPaC was used to design battery packs
                of different capacities and cell counts. The battery utilization (e.g.
                SOC range) were developed based on numerous vehicle test data.\1083\ In
                addition, as discussed further below, for the NPRM analysis, the
                agencies resized the battery pack only with the addition of incremental
                mass reduction technology levels. For this final rule, the agencies
                updated the modeling to consider battery resizing with the application
                of all road load reduction technologies. Accordingly, a more
                appropriately-sized battery pack could result in lower vehicle mass,
                resulting in potentially improved effectiveness.
                ---------------------------------------------------------------------------
                 \1083\ Kim, N., & Jeong, J. (2017). Control Analysis and Model
                Validation for BMW i3 Range Extender. SAE Technical Paper 2017-01-
                1152. doi:10.4271/2017-01-1152. Jeong, J. K. (2019). Analysis and
                Model Validation of the Toyota Prius Prime. SAE World Congress. SAE.
                Namdoo Kim, A. R. (2017). Vehicle Level Control Analysis for Voltec
                Powertrain. Presented at the 30th International Electric Vehicle
                Symposium and Exhibition (EVS30). Stuttgart, Germany. Hanho Son, N.
                K. (2015). Development of Performance Simulation for a HEV with CVT
                and Validation with Dynamometer Test Data. Presented at the 28th
                International Electric Vehicle Symposium (EVS28). Kintex, Korea.
                ---------------------------------------------------------------------------
                 Beyond the powertrain components, Autonomie also considered on-
                board accessory devices that consume energy and affect overall vehicle
                effectiveness. Some electrical power is consumed by electrical
                accessories such as headlights, radiator fans, wiper motors, engine
                control units (ECU), transmission control unit (TCU), cooling systems,
                and safety systems, in addition to driving the motor and the wheels. In
                real-world driving, the electrical accessory load on the powertrain
                varies depending on the how features are used and the condition the
                vehicle is operating in, such as for night driving or hot weather
                driving. However, for regulatory test cycles related to fuel economy,
                the electrical load is repeatable because the fuel economy and
                CO2 regulations control for these factors, as discussed in
                Section VI.B.3 Technology Effectiveness Values.\1084\ Accessory loads
                during test cycles do vary by powertrain type and vehicle technology
                class, since distinctly different powertrain components and vehicle
                masses will consume different amounts of energy.
                ---------------------------------------------------------------------------
                 \1084\ NHTSA Benchmarking, ``Laboratory Testing of a 2017 Ford
                F-150 3.5 V6 EcoBoost with a 10-speed transmission.'' DOT HS 812
                520.
                ---------------------------------------------------------------------------
                 The baseline fleet consists of hundreds of different vehicle types
                that vary in the amount of accessory electrical power that they
                consume. For example, vehicles with different motor and battery sizes
                will require different capacities of electric cooling pumps and fans to
                manage component temperatures. Autonomie has built-in models that can
                simulate these varying sub-system electrical loads. However, for the
                NPRM and this final rule analysis, the agencies used a fixed (by
                vehicle technology class and powertrain type), constant power draw to
                represent the effect of these accessory loads on the powertrain. The
                agencies intended and expected that fixed accessory load values would,
                on average, have similar impacts on effectiveness as found on actual
                manufacturers' systems. This process was in line with the past
                analyses, such as in the Draft TAR and the EPA Proposed
                Determination.\1085\ \1086\ For assumptions regarding accessory load
                modeling for the rulemaking timeframe, the agencies relied on research
                and development data from DOE's Vehicle Technologies Office and Argonne
                Advanced Mobility Technology Laboratory, as well as input from
                automotive manufacturers.\1087\ \1088\ \1089\
                ---------------------------------------------------------------------------
                 \1085\ Draft Technical Assessment Report (July 2016), Chapter 5.
                 \1086\ EPA Proposed Determination TSD (November 2016), at p.2-
                270.
                 \1087\ DOE VTO Power Electronics Research and Development.
                https://www.energy.gov/eere/vehicles/vehicle-technologies-office-
                electric-drive-systems. Last Accessed Jan 2, 2020.
                 \1088\ ANL Advanced Mobility Technology Laboratory (AMTL).
                https://www.anl.gov/es/advanced-mobility-technology-laboratory. Last
                Accessed Jan 2, 2020.
                 \1089\ DOE's lab years are ten years ahead of manufacturers
                potential production intent (i.e 2020 Lab Year is MY 2030).
                ---------------------------------------------------------------------------
                [[Page 24481]]
                 Table VI-88 below shows the NPRM assumptions for all the vehicle
                classes and powertrain types for accessory loads.\1090\ Data from AMTL
                D \3\ testing were used to designate electric loads for different types
                of powertrains.\1091\
                ---------------------------------------------------------------------------
                 \1090\ See NPRM ANL Assumptions Summary.
                 \1091\ ANL Energy Systems Division Downloadable Dynamometer
                Database: https://www.anl.gov/es/downloadable-dynamometer-database.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.209
                BILLING CODE 4910-59-C
                 For the final rule analysis, the agencies updated the electrical
                load assumptions for many of the powertrain types and classes,\1092\
                based on further consideration of comments from the Alliance on the
                2016 Draft TAR and EPA Proposed Determination.\1093\ \1094\ These
                assumptions are provided below, in Table VI-89.
                ---------------------------------------------------------------------------
                 \1092\ See ANL Assumptions Summary, ANL--All
                Assumptions_Summary_FRM_06172019_FINAL.
                 \1093\ Alliance of Automobile Manufacturers Comments on Draft
                TAR at p. 30. September 26, 2016.
                 \1094\ EPA Proposed Determination TSD (November 2016), at p.2-
                270.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.210
                [[Page 24482]]
                 CARB commented on NPRM non-battery component efficiency assumptions
                in two respects; first by claiming that the agencies relied on outdated
                data for electric machines and inverter efficiencies across all
                electrification applications,\1095\ and second by claiming that the
                agencies did not project any efficiency gains in those components over
                time.\1096\ CARB stated that the three vehicles benchmarked in the ORNL
                studies (MY 2007 Toyota Camry Hybrid, a MY 2011 Hyundai Sonata Hybrid,
                and MY 2012 Nissan Leaf) were inappropriate for the agencies to use to
                assess the costs and efficiencies for the same components in MY 2020-
                2030 vehicles, given the rapid development in the past ten years in
                automotive electrification. CARB cited the MY 2016 Chevrolet Volt and
                Bolt, and the MY 2016 Toyota Prius, as examples of vehicles that had
                undergone electric machine efficiency improvements from one generation
                to the next; those vehicles generally employed efficiency improvements
                including reduced electric motor volume and mass, reduced power
                inverter volume, increased electric motor peak power density, and
                reduced mechanical losses through friction reduction, among other
                improvements.
                ---------------------------------------------------------------------------
                 \1095\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 127.
                 \1096\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 128.
                ---------------------------------------------------------------------------
                 In support of their comments that the agencies did not project any
                efficiency gains in non-battery components over time, CARB faulted the
                agencies for not including data from the October 2015 ORNL progress
                report for electric drive technologies, stating that benchmarking data
                for a MY 2014 Honda Accord Hybrid inverter and traction motor
                components could have been used to compare against and update the data
                from the MY 2007 Toyota Camry Hybrid and MY 2011 Hyundai Sonata Hybrid
                efficiency maps benchmarked in the older ORNL report. CARB stated that
                the lack of consideration of this newer data was evidence that the
                agencies' data selection was biased to support weakening fuel economy
                standards.
                 CARB also cited 2017 research from Argonne's Autonomie group as a
                source of updated data that showed efficiency gains over time for
                electrification technologies not considered in the agencies' analysis,
                including increases in high voltage system peak efficiency, increases
                in high voltage specific power, and decreases in costs.\1097\ CARB
                stated that had the agencies included newer data in the analysis,
                including from the same data sources from which prior data came, the
                analysis would have not supported the agencies' proposal.
                ---------------------------------------------------------------------------
                 \1097\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 131. Note that comments on non-battery
                component costs are addressed in Section VI.C.3.e)(2) Non-Battery
                Electrification Component Costs.
                ---------------------------------------------------------------------------
                 The agencies agree that there have been improvements in non-battery
                component efficiency over the past few years, however CARB's
                characterization of the process used to employ the ORNL benchmarking
                data in the analysis was incorrect. Autonomie used high-level electric
                machine characteristics such as base and max motor speed from
                production vehicles along with generic efficiency map curves for each
                technology type, with peak efficiencies matching the current state of
                the art technologies discussed in ORNL reports. Although the source
                data for the electric machines were from older production vehicles, the
                peak electric motor and controller efficiencies were updated to reflect
                the latest available data. Specifically, the NPRM analysis modeled a 92
                percent peak efficiency for motors and controllers.\1098\
                ---------------------------------------------------------------------------
                 \1098\ See the Non_Vehicle_Attribute tab in the NPRM ANL
                Assumptions_Summary.
                ---------------------------------------------------------------------------
                 That said, the agencies also agreed that the analysis could use
                updated peak electric and controller efficiencies, and updated those
                for the final rule. For the final rule analysis, the agencies used 96
                percent efficiency for HEVs and PHEVS, and 98 percent peak efficiency
                for BEVS and FCEVs.\1099\ The agencies believe the final rule
                efficiencies are appropriate for the rulemaking timeframe.
                ---------------------------------------------------------------------------
                 \1099\ See the Non_Vehicle_Attribute tab in the FRM ANL
                Assumptions_Summary.
                ---------------------------------------------------------------------------
                 In addition, as discussed above, other changes for the final rule
                analysis include updating the electric motor sizing as a function of
                electric power to account for lower electric machine mass, updating the
                BEV electric machine map to use a newer efficiency map from the Chevy
                Bolt, updating baseline and reference vehicle mass assumptions to
                reflect latest machine weight technology development, and updating the
                electrical accessory loads for vehicle modeling to reflect data from
                vehicle benchmarking. Changes and updates to the Autonomie analysis are
                discussed throughout this electrification section and in the FRM
                Argonne Model Documentation. In addition, for this final rule analysis,
                the agencies used the latest Argonne BatPaC model to determine the
                battery pack mass and manufacturing costs for electric vehicle
                batteries. Updates to non-battery component efficiency were small in
                comparison to the impact of using updated battery modeling for the
                final rule analysis. Further discussion on battery modeling can be
                found in Section VI.C.3.e)(1) Battery Pack Modeling.
                (2) Modeling and Simulating Vehicles With Electrified Powertrains in
                Autonomie
                 Data from Argonne's AMTL was used to develop the electrified
                powertrain models in Autonomie. The modeled electrification components
                were sized based on performance neutrality needs, as discussed further
                below, and the control algorithms were based on Argonne -collected
                data.\1100\ Detailed discussion about the development of the HEV
                drivetrains can be found in the Autonomie modeling documentation.\1101\
                The modeled powertrains are not intended to represent any specific
                manufacturer's architecture, but are intended to act as surrogates
                predicting representative levels of effectiveness for each
                electrification technology.
                ---------------------------------------------------------------------------
                 \1100\ See FRM ANL Model Documentation.
                 \1101\ See NPRM ANL Model Documentation at p.92.
                ---------------------------------------------------------------------------
                 The agencies also broadly discussed in Section VI.B.3 Technology
                Effectiveness Values that certain technologies' effectiveness for
                reducing fuel consumption requires optimization through the appropriate
                sizing of the powertrain. This analysis iteratively minimizes the size
                of the powertrain components to maximize efficiency while at the same
                time enabling the vehicle to meet multiple performance criteria. The
                Autonomie simulations use a series of resizing algorithms which contain
                ``loops,'' such as an ``Acceleration Performance Loop (0-60 mph),''
                which automatically adjust the size of certain powertrain components
                until a criterion, for example 0-60 acceleration time, converges to a
                target value. As the algorithms examine different performance or
                operational criteria that must be met, no single criterion is allowed
                to degrade; once a resizing algorithm completes, all criteria will be
                met, and some may be exceeded as a necessary consequence of meeting
                others.
                 Autonomie applies different powertrain sizing algorithms depending
                on the type of vehicle considered because different types of vehicles
                not only contain different powertrain components to be optimized, but
                they must also operate in different driving modes. While the
                conventional powertrain sizing algorithm must consider only the power
                of the engine, the more complex algorithm for
                [[Page 24483]]
                electrified powertrains must simultaneously consider multiple factors,
                which could include the engine power, electric machine power, battery
                power and battery capacity. Also, while the resizing algorithm for all
                vehicles must satisfy the same performance criteria, the algorithm for
                some electric powertrains must also allow those electrified vehicles to
                operate in certain driving cycles without assistance of the combustion
                engine, and ensure the electric motor/generator and battery can handle
                the vehicle's regenerative braking power, all-electric mode operation
                and intended range of travel.
                 To establish the effectiveness of the technology packages,
                Autonomie simulated the vehicles performing compliance test cycles, as
                discussed in Section VI.B.3 Technology Effectiveness
                Values.1102 1103 1104 For vehicles with conventional
                powertrains and micro hybrids, Autonomie simulated the vehicles using
                the 2-cycle test procedures and guidelines.\1105\ For mild HEVs, strong
                HEVs, and FCVs, Autonomie simulated the 2-cycle test, with the addition
                of repeating the drive cycles until the final state of charge was
                approximately the same as the initial state of charge, a process
                described in SAE J1711. For PHEVs and BEVs, Autonomie simulated
                vehicles performing the test cycles per guidance provided in SAE
                J1711.\1106\ For BEVs, Autonomie simulated vehicles performing the test
                cycles per guidance provided in SAE J1634.\1107\
                ---------------------------------------------------------------------------
                 \1102\ EPA, ``How Vehicles are Tested.'' https://www.fueleconomy.gov/feg/how_tested.shtml. Last accessed Nov 14,
                2019.
                 \1103\ See FRM ANL Model Documentation at Chapter 6: Test
                Procedures and Energy Consumption Calculations.
                 \1104\ EPA Guidance Letter. ``EPA Test Procedures for Electric
                Vehicles and Plug-in Hybrids.'' Nov. 14, 2017. https://www.fueleconomy.gov/feg/pdfs/EPA%20test%20procedure%20for%20EVs-PHEVs-11-14-2017.pdf. Last accessed Nov. 7, 2019.
                 \1105\ 40 CFR part 600.
                 \1106\ PHEV testing is broken into several phases based on SAE
                J1711. Charge-Sustaining on the City cycle, Charge-Sustaining on the
                HWFET cycle, Charge-Depleting on the City and HWFET cycles.
                 \1107\ SAE J1634. ``Battery Electric Vehicle Energy Consumption
                and Range Test Procedure.'' July 12, 2017.
                ---------------------------------------------------------------------------
                 A survey of comments about the modeled effectiveness of
                electrification technologies showed most comments could be sorted in
                three major categories. The first, and largest category of comments,
                were concerned with effectiveness values used for the technologies.
                Specifically, commenters were concerned the values for the modeled
                effectiveness of the technologies were too low, particularly when
                compared to past analysis efforts. The second major category of
                comments were concerned with the size of the electrification components
                selected in the Autonomie tool, and used to simulate the system
                performance. Commenters were concerned because oversized components can
                lead to the system violating performance neutrality constraints and
                artificially increasing the cost of the technology. The third major
                category of comments were concerned not enough variety of technologies
                were represented in the electrification technology models.
                Specifically, commenters wanted additional engine technologies allowed
                to couple with electrification technologies.
                 Each of the comments from the first category will be referenced and
                addressed under the specific technology sections, below. However,
                broadly, two factors have led to the comments raised by stakeholders.
                First, as discussed throughout this document, the agencies avoided
                using performance values in the analysis that can be traced to specific
                implementation of a technology type. Thus, when comparing simulated
                performance to any specific real world vehicle, there will be a
                deviation. The modeled inputs are meant to represent the typical range
                of values for a technology--reasonable and realistic values--but are
                not likely to result in performance outputs that would equal any
                specific existing vehicle. Second, the modeling approach implemented in
                the NPRM and final rule analysis succeeds in isolating the effects of
                individual technologies to a higher degree then previous analysis. Due
                to the greater use of parametric modeling of full vehicle systems, the
                specific effects of technologies could be isolated to a higher degree
                from the amplifying or muting effects of other technologies. This
                isolation of effect often results in lower predicted effectiveness
                values for individual technologies than has been observed in previous
                analysis, where the isolation of effect was not as precise, and often
                attributed efficiency gains from a combination of technological changes
                to a single technology.
                 For the second major group of comments, the agencies mostly agreed
                with the stakeholder observations. The issues identified were
                investigated by the agencies and resulted in changes to the sizing
                algorithms used by the agencies for the final rule analysis. The
                agencies further investigated the constraints of performance neutrality
                and ensured those constraints were followed for sizing of
                electrification components. Further discussion of the changes made, as
                well as specific answers to comments under each technology section, can
                be found in the following technology subsections and in Performance
                Neutrality, Section VI.B.3.a)(6).
                 The third major group of comments from stakeholders were concerned
                with allowing more engine technologies to be incorporated in
                electrification systems. The agencies agreed with these comments and
                increased the number of technology combinations available. The new
                combinations are discussed in Section VI.C.3.a)(1) Electrification
                Technologies, as well as under each technology section below.
                (a) Micro and Mild Hybrid Vehicles
                 The micro and mild hybrid systems modeled in Autonomie represented
                SS12V and BISG technology (and CISG technology for the NPRM). SS12V and
                BISG were modeled using a similar approach because both systems have
                low peak power, low energy storage, and allow stop/start engine idle
                reduction. The effectiveness improvement from both technologies is
                attributable to the amount of fuel saved during engine idling period on
                the 2-cycle test. However, only the BISG system model allowed limited
                assist to propel the vehicle and limited regenerative braking. For
                further discussion of these system models, see the FRM Argonne Model
                Documentation.\1108\
                ---------------------------------------------------------------------------
                 \1108\ See FRM ANL Model Documentation at chapters 4.6, 4.7 and
                4.13.
                ---------------------------------------------------------------------------
                 Powertrain resizing was not employed for micro or mild hybrid
                system application, in either the NPRM or this final rule analysis.
                These systems have little to no impact on the vehicle performance
                metrics that would be adjusted by powertrain resizing, and in turn
                there would be limited or no benefit in attempting to resize upon
                application of these systems. For example, the micro hybrid SS12V
                system allows the engine to be turned off when the vehicle is fully
                stopped to reduce idle-stop fuel consumption, but the combustion engine
                size must be retained to maintain performance metrics such as
                acceleration. The main focus of mild hybrid vehicles is to provide
                idle-stop and capture some regenerative braking energy, and although
                they also can provide some assistance to the engine during the initial
                propelling of the vehicle, this is done to improve efficiency and does
                not significantly improve the acceleration performance of the vehicle.
                With BISG mild hybrids, the electric machine is linked to the engine
                through a belt, and thus the potential power assistance is usually
                limited. In the NPRM, the BISG system used an 806 Wh capacity battery
                [[Page 24484]]
                pack and a 10 kW motor/generator. For the final rule analysis, the 10
                kW motor/generator was paired with a 403 Wh battery pack to align with
                BISG systems emerging in the marketplace.
                 ICCT commented that the agencies unjustifiably reduced the
                CO2 and fuel consumption benefits of SS12V from the Draft
                TAR, including a reduction in the overall effectiveness benefit when
                the SS12V system was applied in combination with other
                technologies.\1109\ ICCT stated that the agencies should know the
                precise effectiveness improvement for SS12V technology based on EPA
                compliance data, and the agencies should report a full listing of all
                the baseline 2016 vehicle models with stop-start technology, with their
                test-cycle, and off-cycle improvement in g/mile and percent
                effectiveness. ICCT claimed that the agencies either intentionally
                ignored the full compliance benefits of SS12V technology or ``ignored
                the knowledge and expertise of the EPA engineering and compliance
                staff,'' and argued that not reporting the requested data would be
                ``hiding relevant data the agencies have readily available to more
                rigorously assess existing stop-start technologies and their impact for
                the rulemaking.'' ICCT also stated that the agencies did not
                appropriately include the full regulatory benefit (i.e., inclusion of
                the additional off-cycle ``credit'' under EPA's program or fuel
                consumption improvement value under NHTSA's program) of SS12V
                technologies due to their off-cycle improvements.\1126\
                ---------------------------------------------------------------------------
                 \1109\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-22.
                ---------------------------------------------------------------------------
                 HDS made a similar observation, noting that the SS12V benefit from
                the NRPM was similar to the 2012 TSD projection, but lower than the
                benefit quoted by stakeholders in the Draft TAR.\1110\ HDS cited the
                difference in fuel economy between two vehicles that were produced with
                and without a SS12V option (the 2015 Ford Fusion 1.5L TGDI and the 2015
                Mazda 3 i-ELOOP) which suggested effectiveness values for SS12V of
                about 3.3 percent for both vehicles. HDS also cited a Bosch
                presentation that claimed newer SS12V systems could provide
                effectiveness of up to 6 percent. HDS argued that this actual data and
                supplier data supported a benefit of at least 3.3 percent, which they
                stated was double the benefit in the NRPM analysis.
                ---------------------------------------------------------------------------
                 \1110\ H-D Systems, Attachment 1, Docket No. NHTSA-2018-0067-
                11985, at 44.
                ---------------------------------------------------------------------------
                 The agencies disagree with ICCT and HDS' comments regarding the
                effectiveness of the SS12V technology modeled in the NPRM analysis. The
                implementation of the full vehicle simulation approach used in the
                NPRM, and carried forward to the final rule analysis, clearly defines
                the contribution of individual technologies and separates those
                contributions from other technologies. The modeling approach also shows
                when technologies have amplifying or muting interactions. In some
                cases, this may appear as a reduction in performance compared to
                previous analysis. The agencies modeled the SS12V system in conjunction
                with all the IC engine and transmission combinations. The results of
                this parametric modeling accounted for each engine and transmission
                combination's unique fuel consumption rate at idle.\1111\ The range of
                effectiveness for the technology in the NPRM analysis is a result of
                these differences. This range of values will result in some modeled
                effectiveness values being close to real-world measured values, and
                some modeled values that will depart from measured values, depending on
                the level of similarity between the modeled hardware configuration and
                the real-world hardware configuration. This modeling approach comports
                with the National Academy of Science 2015 recommendation to use full
                vehicle modeling supported by application of lumped improvements at the
                sub-model level.\1112\ The approach allows the isolation of technology
                effects in the analysis supporting an accurate assessment.
                ---------------------------------------------------------------------------
                 \1111\ For example, when idling, a larger eight-cylinder engine
                has more friction and pumping losses than a smaller four-cylinder
                engine, and therefore will save more fuel when the engine is shut-
                off at rest.
                 \1112\ National Research Council. 2015. Cost, Effectiveness, and
                Deployment of Fuel Economy Technologies for Light-Duty Vehicles.
                Washington, DC--The National Academies Press. https://www.nap.edu/catalog/21744/cost-effectiveness-and-deployment-of-fuel-economy-technologies-for-light-duty-vehicles, at 292.
                ---------------------------------------------------------------------------
                 For both the NPRM and final rule analysis, the agencies assigned
                SS12V technology to vehicles in the analysis fleet using compliance
                data, and used compliance data to assign a vehicle's baseline fuel
                economy value. The market data file indicated the presence of SS12V on
                a vehicle, and accordingly, the vehicles reported to include SS12V
                technology in the analysis fleet were modeled with the technology. For
                more discussion on how technologies were assigned to the vehicle
                platforms in the analysis fleet, please see Section VI.B.1 Analysis
                Fleet. The agencies accounted for the contribution of the SS12V
                technology in the analysis fleet by using the reported compliance fuel
                economy values as the baseline fuel economy values for vehicles that
                included the technology. The analysis fleet fuel economy values were
                the reported final compliance values for the given vehicle platform and
                should include the benefits from all technologies on the vehicle
                platform.\1113\ The agencies also captured the off-cycle credits
                provided to a manufacturer for the existence of the technology in the
                manufacturer's fleet. For the NPRM and final rule analysis, the
                manufacturers' fleets are modeled with baseline year compliance-
                reported off-cycle credits. Further, for the final rule analysis, the
                agencies increased the application of off-cycle credits in the
                analysis, as discussed in Section VI.B.2.a) Off-cycle and A/C
                Efficiency Adjustments to CAFE and Average CO2 Levels.
                ---------------------------------------------------------------------------
                 \1113\ Sec. 32904. Calculation of average fuel economy, https://uscode.house.gov/browse/[email protected]/subtitle6/partC/chapter329&edition=prelim.
                ---------------------------------------------------------------------------
                 Commenters similarly disagreed with the BISG effectiveness
                presented in the NPRM analysis, suggesting the resulting effectiveness
                improvement should be at a range of 4 percent to 6 percent.\1114\ Such
                commenters claimed that it was unclear why effectiveness values were so
                much lower than previous effectiveness estimates. More specifically,
                comments centered on (1) arguing that the agencies' modeling of BISG
                and CISG systems in Autonomie likely underestimated the resulting
                effectiveness values; (2) suggesting that the values in prior documents
                like the Draft TAR and the 2015 NAS report were more accurate; and (3)
                comparing modeled effectiveness values to claimed values achieved by
                actual on-road vehicles and mild hybrid systems.
                ---------------------------------------------------------------------------
                 \1114\ ICCT, Attachment 3, Docket No. NHTSA-2018-0067-11741;
                California Air Resources Board, Attachment 2, Docket No. NHTSA-2018-
                0067-11873; Roush Industries, Attachment 1, NPRM Docket No. NHTSA-
                2018-0067-11984; H-D Systems, ``HDS final report,'' Docket No.
                NHTSA-2018-0067-11985; Union of Concerned Scientists, Attachment 2,
                Docket No. NHTSA-2018-0067-12039.
                ---------------------------------------------------------------------------
                 CARB claimed that the agencies failed to disclose the necessary
                details to conclude why mild hybrid systems were projected to have
                lower efficiency values than past estimates. CARB also concluded the
                lack of engine downsizing when adding a BISG/CISG system and the lack
                of adjusting transmission drive ratios and shift logic were reasons why
                BISG/CISG effectiveness was underpredicted.\1115\ CARB claimed not
                resizing the engines resulted in a ``less than optimized system that
                does not take full advantage
                [[Page 24485]]
                of the mild hybrid system.'' \1116\ CARB argued that the agencies'
                assumption that manufacturers ``would not optimize the engine and
                transmission when installing a CISG is not realistic and results in
                improper pairing of advanced gasoline engines and transmissions in the
                modeling and leads to underestimation of the efficiency benefits.'' As
                mentioned above, CARB stated that manufacturers ``often are required to
                make a[n] engine casting change to accommodate the system,'' and when
                doing so, ``no manufacturer would fail to pair the system with an
                optimally sized engine and configured transmission to take full
                advantage of the system's capabilities.'' \1117\
                ---------------------------------------------------------------------------
                 \1115\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 163.
                 \1116\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 185.
                 \1117\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 186.
                ---------------------------------------------------------------------------
                 CARB also inquired into whether the Argonne modeling ``took full
                advantage'' of the system, using Daimler's EQ Boost system, that
                provides temporary boosts for acceleration and enables engine shut-off
                during coasting events, as an example.\1118\ Similarly, CARB noted that
                CISG systems' ability to provide low end torque makes it an ``ideal
                technology to pair with an engine technology that may have poor low end
                torque but improved efficiency under other conditions; examples could
                include an HCR engine sized with minimal low end torque to maximize
                efficiency improvements in other operating conditions or a turbocharged
                downsized engine equipped with a larger turbine to reduce backpressure
                but provide improved efficiency over a larger portion of the engine
                map.'' \1119\ CARB stated that manufacturers are using such systems to
                boost engine torque at higher operating speeds so they can keep the
                engine operating in a more efficient region.
                ---------------------------------------------------------------------------
                 \1118\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 163.
                 \1119\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 163.
                ---------------------------------------------------------------------------
                 Commenters also cited data from suppliers that produce 48V BISG
                systems, including data from TULA that showed a 11 percent fuel economy
                benefit from a 48V system,\1120\ data from a Delphi 48V system
                prototype installed on a Honda Civic that showed a 10 percent reduction
                in CO2 emissions levels,\1121\ and data from Continental
                showing a 13 percent fuel savings improvement from its BISG
                system.\1122\ ICCT also cited its supplier and technology report on
                hybrids that estimated the benefit of mild hybrid technology at 12.5
                percent, which it characterized as ``remarkably similar'' to that
                achieved by the 2019 RAM pickup truck.\1123\ HDS noted that even if the
                effectiveness values from TULA are regarded as optimistic because they
                are the developers of the technology, EPA's previous modeling results
                of 8-9 percent effectiveness ``appear reasonable in light of what is
                observed from certification data.'' \1124\ ICCT ultimately recommended
                the agencies revise the effectiveness value for mild hybrid systems to
                include a CO2 effectiveness value of 12.5 percent.\1125\
                ---------------------------------------------------------------------------
                 \1120\ H-D Systems, Attachment 1, Docket No. NHTSA-2018-0067-
                11985, at 45.
                 \1121\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 160.
                 \1122\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 160.
                 \1123\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-24.
                 \1124\ H-D Systems, Attachment 1, Docket No. NHTSA-2018-0067-
                11985, at 45.
                 \1125\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-25.
                ---------------------------------------------------------------------------
                 Commenters also stated that the effectiveness estimates for CISG
                systems were significantly understated, \1126\ with UCS characterizing
                CISG systems as showing ``virtually no benefit whatsoever for CISG over
                BISG, and in many cases actually show[ing] an increase in fuel
                consumption.'' \1127\ UCS stated this was a dramatic departure from
                previous Autonomie results, and with ``no explanation whatsoever''
                given for the decrease in technology effectiveness.
                ---------------------------------------------------------------------------
                 \1126\ Union of Concerned Scientists, Attachment 2, Docket No.
                NHTSA-2018-0067-12039; Roush-Industries, Attachment 1, Docket No.
                NHTSA-2018-0067-11984; California Air Resources Board, Attachment 2,
                Docket No. NHTSA-2018-0067-11873.
                 \1127\ Union of Concerned Scientists, Attachment 2, Docket No.
                NHTSA-2018-0067-12039, at 3.
                ---------------------------------------------------------------------------
                 The agencies agree with commenters that the NPRM analysis of mild
                hybrid technologies could be more representative of production vehicles
                and vehicles likely to be produced during the rulemaking time period.
                The agencies further conclude that the NPRM analysis overestimated the
                costs of such technologies. Thus, for the final rule analysis, the
                agencies only considered one 48V BISG system in the mild hybrid
                technology category. The 48V mild hybrid BISG system used the same 10
                kW electric motor as the one used in the NPRM analysis, and the 48V
                BISG battery pack was also reduced in size to 403 W-hr from 806 W-hr to
                reflect more accurately the size of battery packs available in the
                market. In addition, the Autonomie model increased the usable battery
                capacity, increasing the duration of electric motor use by the vehicle
                before starting the engine. The specifications and assumptions for the
                48V BISG system are further discussed in the FRM Argonne Model
                Documentation and FRM Argonne Assumptions Summary.1128 1129
                The discontinued use of the CISG technology is discussed in Section
                VI.C.3.a)(1)(c) Electrification Technologies, Mild Hybrids.
                ---------------------------------------------------------------------------
                 \1128\ See FRM ANL Model Documentation, at 4.6, 4.13, and 5.7.
                 \1129\ FRM ANL Assumptions Summary (see Model Documentation
                tables in Section VI.A.7 Structure of Model Inputs and Outputs).
                ---------------------------------------------------------------------------
                 The agencies disagree with comments stating incremental
                effectiveness estimated by Autonomie modeling was incorrect because the
                effectiveness values deviated from past effectiveness values estimated
                in the agencies' rulemakings or from real-world values measured on
                specific vehicles. As discussed in previous sections, the
                implementation of the full vehicle simulation approach used in the NPRM
                analysis and carried forward to the final rule analysis clearly defines
                the contribution of individual technologies through the application of
                parametric modeling. This approach clearly separated the contributions
                of each technology. The modeling approach also showed the amplifying or
                muting interactions between technologies. In some cases, this may
                appear as reduced performance in comparison to previous analysis. The
                agencies also strongly disagree that they should use the performance
                values for any specific vehicle as representative of all mild hybrid
                systems.
                 CARB also commented that the agencies' decision to use a fixed
                final drive ratio and fixed shift logic based on the selected
                transmission did not allow for efficiency improvements when mated with
                electrified powertrains, with specific regards to mild hybrid BISG and
                CISG systems.\1130\ CARB stated that based on the information disclosed
                in the NPRM, ``it appears that Argonne did not utilize the system in
                these manners nor did they allow for changes in gear ratios, final
                drive ratio, or transmission shift logic to optimize for efficiency
                improvements when mated with different electrified powertrains.''
                \1131\ Roush Industries similarly stated that the analysis under-
                predicted the potential improvements of employing a BISG system because
                the engine could operate at a lower RPM with the help of the torque
                assist of the electric motor/generator, with a change to the final
                [[Page 24486]]
                drive ratio and transmission shift logic, but the analysis did not do
                so.\1132\
                ---------------------------------------------------------------------------
                 \1130\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 185.
                 \1131\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 185.
                 \1132\ Roush Industries, Attachment 1, Docket No. NHTSA-2018-
                0067-11984, at 16.
                ---------------------------------------------------------------------------
                 The agencies disagree with CARB and Roush Industries' claims about
                the gear ratio and shift logic used for the NPRM. As discussed in
                Section VI.C.2.d) Transmission Effectiveness Modeling and Resulting
                Effectiveness Values, manufacturers commonly maintain the same gear
                hardware across vehicle platforms and applications, relying on controls
                and shift strategy to achieve optimization. Autonomie maintained gear
                hardware but customized the shifting strategy for each unique vehicle
                system modeled \1133\ to reflect real-world manufacturing strategies
                more accurately.
                ---------------------------------------------------------------------------
                 \1133\ FRM ANL Model Documentation, at 4.4.5.
                ---------------------------------------------------------------------------
                 CARB also commented that the performance modeled by the Autonomie
                tool in the NPRM analysis failed to remain neutral for over 80 percent
                of the modeled systems with mild hybrids. CARB felt the over-
                performance was ``indicating some portion of the system capability was
                improperly modeled to improve performance rather than reduce
                CO2 emissions.'' \1134\
                ---------------------------------------------------------------------------
                 \1134\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 163.
                ---------------------------------------------------------------------------
                 The agencies agree with CARB's observations about the performance
                of mild hybrid combinations. The mild hybrid configuration exhibited
                higher performance in comparison to non-mild hybrid configurations in
                the NPRM analysis. For the final rule analysis, the agencies updated
                sizing and control of the mild hybrid systems to minimize performance
                changes and maintain neutrality. As discussed earlier in this chapter,
                updates include using smaller battery systems, updated algorithms, and
                updated component weights. For further discussion of performance
                neutrality for the final rule, see the Performance Neutrality Section
                VI.B.3.a)(6).
                 Finally, ICCT commented that the agencies should include off-cycle
                and ``game-changing'' pickup truck credits in the effectiveness
                estimates for hybrid pickup trucks, as ``[i]t is the responsibility of
                the agencies to include all applicable credits with their technology
                packages calculations and their projections, including any additional
                credits that will automatically accrue.'' \1135\
                ---------------------------------------------------------------------------
                 \1135\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-25.
                ---------------------------------------------------------------------------
                 While the agencies included many compliance flexibilities in the
                modeling for the final rule analysis, hybrid pickup truck credits were
                not modeled. The referenced pickup truck credit is set to expire for
                all pickup trucks after MY 2021, so in analyzing this comment the
                agencies considered what technologies manufacturers could apply to
                pickup trucks through that model year to meet the requirements
                specified in the regulation. To receive credit in a model year,
                manufacturers must produce a quantity of improved full size pickup
                trucks--improvement characterized by including either hybrid technology
                or improved emissions performance--such that the proportion of
                production of such vehicles, when compared to the manufacturer's total
                production of full size pickup trucks, is not less than an amount
                specified in that model year. The agencies determined that, based on
                manufacturers' MY 2017 pickup truck offerings characterized in the
                analysis fleet and with the technology considered in this rule, no
                pickup truck manufacturer could meet the criteria set by EPA to qualify
                for the mild credit before the credit is set to expire. For the strong
                HEV credit, the agencies considered that forcing the application of
                strong HEV pickups to meet the minimum threshold of 10 percent of the
                fleet in order to earn the incentive credits would significantly
                increase the cost for compliance and be less cost-effective than other
                technology pathways. As the analysis seeks the most cost-effective
                pathway for compliance, the agencies disagree the analysis should force
                the application of strong HEV technology to at least 10 percent of full
                size pickup trucks. However, the agencies did allow and simulated
                maximum off-cycle and A/C off-cycle FCIVs for all manufacturers in the
                CAFE model for both the CAFE and CO2 programs during the
                rulemaking time frame. So, while the agencies did not model pickup
                truck credits specifically, the final rule analysis allowed
                manufacturers to reach the maximum off-cycle credit cap during the
                rulemaking timeframe.
                (b) Strong Hybrid Vehicles
                 The power-split hybrid (SHEVPS) model in Autonomie included a
                power-split device, two electric machines and an engine, and allowed
                various interactions between these components. The SHEVP2 model in
                Autonomie is based on the pre-transmission (P2) configuration where the
                electric motor is placed between the engine and transmission for direct
                flow of power to the wheels. The vehicle can be propelled either by the
                combustion engine, electric motor, or both simultaneously, but the
                speed/efficiency region of operation for SHEVP2s under any engine/motor
                combination is ultimately dictated by the transmission gearing and
                speed. Detailed discussion of SHEVPS and SHEVP2 modeling and validation
                are provided in the Argonne Model Documentations.\1136\ Autonomie full
                vehicle models representing strong hybrids were based on vehicle test
                data from vehicle benchmarking.
                ---------------------------------------------------------------------------
                 \1136\ FRM ANL Model Documentation, at Chapters 4.13, 4.16 and
                6.0.
                ---------------------------------------------------------------------------
                 As discussed previously in this section, power-split hybrids
                utilize a combustion engine, two electric machines and a planetary gear
                set along with a battery pack to propel the vehicle. The smaller motor/
                generator (EM1) is used to control the engine speed and uses the engine
                to either charge the battery or to supply additional electric power to
                the second ``drive'' motor. The more powerful drive motor/generator
                (EM2) is permanently connected to the vehicle's final drive and always
                turns with the wheels. The SHEVPS resizing algorithm makes an initial
                estimate of the size of the engine, battery, and electric motors. The
                initial estimates for the combustion engine and EM2 sizes are based on
                the peak power required for acceleration performance and the continuous
                power required for gradeability performance. The initial estimates for
                the battery and EM1 powers are based the maximum regenerative braking
                power. With these initial size estimates, the algorithm computes the
                vehicle mass, and simulations are run to determine if 0-60 and 50-80
                mph acceleration performance is acceptable. If acceleration is not
                satisfactory (too fast or too slow), the algorithm iteratively adjusts
                the sizes of the engine, motors, and battery, and runs simulations
                until a minimum powertrain size is found that meets all requirements.
                With each iteration, the engine, battery, and motor characteristics
                were also updated for gradeability performance and regeneration, if
                necessary. Figure VI-32 below shows the general steps of the SHEVPS
                sizing algorithm. Detailed descriptions are available in section 8.3 of
                the FRM Argonne Model Documentation.
                BILLING CODE 4910-59-P
                [[Page 24487]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.211
                 A parallel hybrid (SHEVP2) uses a combustion engine and a multi-
                speed transmission-integrated electric motor (EM1), as discussed
                previously in this section. As is done with SHEVPS, the SHEVP2 resizing
                algorithm creates a starting point by making an initial estimate of the
                size of the engine, battery, and electric motor based on performance
                criteria or an estimated regenerative braking power, in turn
                calculating the associated vehicle mass. The algorithm then uses a
                simulation loop to find a more precise value of regenerative braking
                power generated in the UDDS ``city driving'' cycle, and adjusts the
                electric motor size and vehicle mass accordingly. Next, the algorithm
                uses simulation loops to optimize the engine, motor, and battery sizes
                in relation to acceleration performance criteria. In the event that the
                acceleration criteria requires downsizing the powertrain, the electric
                motor size is not reduced as this would not be suitable for the
                handling of regenerative braking power. If the acceleration criteria
                cause the electric motor to increase in size, the algorithm then
                returns to the regenerative braking loop and subsequently all other
                loops until all components are optimized. Figure VI-33 below shows a
                simplified sizing algorithm for SHEVP2s.
                BILLING CODE 4910-59-C
                 In the NPRM, the acceleration optimization loops in the SHEVP2
                algorithm did not resize the powertrain if the resulting acceleration
                time was less than the target. This strategy was intended to avoid
                reducing the engine size compared to the conventional vehicle,
                mimicking an observed marketplace trend in which parallel hybrid models
                tend to retain similar engine sizes as the non-hybrid models bearing
                the same nameplate. However, in some cases this resulted in overly
                aggressive SHEVP2 acceleration times; to further maintain performance
                neutrality, the final rule sizing algorithm for standard (non-
                performance) SHEVP2 vehicle powertrains was changed to allow engine
                downsizing such that acceleration performance could converge toward the
                target value. This algorithm update is also detailed in Section
                VI.B.3.a)(6), Performance Neutrality.
                 CARB, ICCT, Meszler and ACEEE commented that some combinations of
                advanced engines mated with strong hybrids were illogical and
                inefficient.\1137\ \1138\ \1139\ \1140\ The commenters specifically
                discussed combinations of SHEVP2 with TURBO2 and CEGR1 technologies
                that stated the incremental effectiveness resulted in near zero to
                negative value, but also clarified that not all combinations showed
                inappropriate effectiveness. CARB further expanded that ``[t]hese are
                not likely combinations utilized by manufacturers as they unnecessarily
                add both gasoline technology and hybrid technology that negates many of
                the benefits of the advanced gasoline technology. This error in the
                Agencies' modeling leads to inflated technology costs on vehicles that
                are converted into P2HEVs.'' \1141\
                ---------------------------------------------------------------------------
                 \1137\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 155.
                 \1138\ American Council for an Energy-Efficient Economy, ACEEE
                SAFE NPRM comments, Docket No. NHTSA-2018-0067-12122-22, at 8.
                 \1139\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-25.
                 \1140\ Comments from Meszler Engineering Services, Attachment 2,
                NPRM Docket No. NHTSA-2018-0067-11723, at 14.
                 \1141\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 186.
                ---------------------------------------------------------------------------
                 The agencies now conclude that the NPRM included certain engine and
                strong hybrid pairings that resulted in incremental effectiveness that
                exceeded a reasonable level of performance neutrality. The agencies
                also agree that Autonomie should model strong hybrid technology
                combinations with other engine technologies. In response to these
                comments, for the final rule analysis the agencies updated the CAFE
                model to allow the use of HCR engine technologies with strong hybrids,
                as discussed in Section VI.C.1.c)(4) Engine Maps, HEV Atkinson Cycle
                Engines, and improved full vehicle modeling of turbocharged engine
                combinations. These changes were discussed in Section VI.B.3.a)(1)
                Full-Vehicle Modeling, Simulation Inputs and Data Assumptions and
                Section VI.C.2.d)(1)(a) Shifting Controller.
                 In addition, the agencies limited adoption of advanced engine
                technologies with strong hybrids in cases where the electrification
                technology would have little effectiveness benefit beyond the benefit
                of the advanced engine system, but
                [[Page 24488]]
                would substantially increase costs. Specifically, the agencies did not
                model strong hybrid technologies with VCR engines (eng26a) and eBoost
                engines (eng23c). The agencies believe that manufacturers would not
                consider these combinations because the combination of electrification
                and advanced engine technologies are not as cost-effective as other
                technologies.
                c) Plug-In Hybrid Vehicles
                 The effectiveness of the PHEV systems in the analysis was dependent
                on both the vehicle's battery pack size and range, in addition to the
                other fuel economy-improving technologies on the vehicle (e.g.,
                aerodynamic and mass reduction technologies). For the NPRM analysis,
                the electrification components were sized to achieve the specified all-
                electric range (AER) on the combined cycle (UDDS + HWFET) on the basis
                of adjusted energy values. As mentioned above, the PHEV would provide
                propulsion energy for a limited range in addition to start-stop or
                idle-stop. The NPRM analysis classified PHEVs into two levels: (1)
                PHEV30 indicating a vehicle with an AER of 30 miles; and (2) PHEV50
                indicating a vehicle with AER of 50 miles.
                 The resizing algorithm for plug-in hybrid (PHEV) vehicles,
                similarly as for SHEVs, considered the power needed for acceleration
                performance and all-electric mode operation (compared to regenerative
                braking for SHEVs); the PHEV resizing algorithms used those metrics for
                an initial estimation of engine, motor(s) and battery powers, and
                battery capacity. The initial mass of the vehicle was then computed,
                including weight for a larger battery pack and charging
                components.\1142\ However, since PHEVs offer expanded electric driving
                capacity, their resizing algorithm must also yield a powertrain with
                the ability to achieve certain driving cycles and range in electric
                mode, in which the engine remains off all or the majority of the
                operation. The analysis sized the PHEV electric motors and battery
                powers to be capable of completing either the City Cycle (UDDS) or US06
                (aggressive, high speed) driving cycle in electric mode, and the
                battery energy storage capacity to achieve the specified all-electric
                range on the 2-cycle tests on the basis of adjusted energy
                values.1143 1144
                ---------------------------------------------------------------------------
                 \1142\ FRM ANL Model Documentation, at 8.3 Vehicle Powertrain
                Sizing Algorithms.
                 \1143\ Battery sizing and definition of combined 2-cycle tests
                all-electric range is discussed in detail in ANL Autonomie Model
                Documentation Chapter 6 Test Procedure and Energy Consumption
                Calculation.
                 \1144\ ANL has incorporated SAE J1711 standard into Autonomie
                Modeling. J1711: Society of Automotive Engineers Recommend Practice
                for Measuring Exhaust Emissions and Fuel Economy of Hybrid-Electric
                Vehicles, Including Plug-In Hybrid Vehicles.
                ---------------------------------------------------------------------------
                 The final rule analysis classified PHEVs into four technology
                levels, as discussed previously: (1) PHEV20 indicating a vehicle with
                an AER of 20 miles and powertrain system based on SHEVPS hybrid
                architecture; (2) PHEV50 indicating a vehicle with an AER of 50 miles
                and powertrain system based on SHEVPS hybrid architecture; (3) PHEV20T
                indicating a vehicle with an AER of 20 miles and powertrain system
                based on SHEVP2 hybrid architecture; and (4) PHEV50T indicating a
                vehicle with AER of 50 miles and powertrain system based on SHEVP2
                hybrid architecture.\1145\ The PHEV20, PHEV20T, PHEV50, and PHVE50T
                resizing algorithms were functionally equal, and differed only in the
                type of electric mode driving cycle simulated in each one (UDDS for
                PHEV20/20T, or US06 for PHEV50/50T). These algorithms simulated the
                driving cycles in an iterative loop to determine the size of the
                electric motors and the battery required to complete the cycles. In the
                case of PHEV20 and PHEV20T, the power of the electric motors and
                battery must be sized to propel the vehicle through the UDDS cycle in
                ``charge-depleting (CD) mode;'' in this mode, the electric machine
                alone propels the vehicle except during high power demands, at which
                point the engine may turn on and provide propulsion assistance. The
                PHEV50 and PHEV50T motor(s) and battery must be sized to power the
                vehicle through the US06 cycle in ``electric vehicle (EV) mode,'' where
                the engine is off at all times. Then, all PHEV algorithms adjusted the
                battery capacity, or vehicle range, by ensuring the battery energy
                content was sufficient to complete a simulated UDDS+HWFET combined
                driving cycle, based on EPA-adjusted energy consumption. Finally, the
                engine, electric motor(s), and battery powers were then sized
                accordingly to meet 0-60 and 50-80 mph acceleration targets. All loops
                were repeated until the acceleration targets were met without needing
                to resize the electric motors, at which point the resizing algorithm
                finished. Figure VI-34 below shows the general steps of the PHEV sizing
                algorithm. Detailed steps can be seen in section 8.3 of the FRM Argonne
                Model Documentation.
                ---------------------------------------------------------------------------
                 \1145\ As discussed previously, the NPRM analysis included
                PHEV30 instead of PHEV20. However, the related resizing algorithm is
                applicable to either.
                ---------------------------------------------------------------------------
                [[Page 24489]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.213
                 Meszler, CARB, and BorgWarner provided comments on the
                effectiveness of the PHEV models. The commenters were concerned with
                underperformance of the technology, sizing of the components, and the
                variety of PHEV technologies available.
                 Meszler commented that PHEVs in the 2016 analysis fleet were
                inappropriately constrained in their future fuel economy potential by
                the ratio of baseline electric-only fuel economy to baseline engine-on
                fuel economy; and those vehicles should be allowed to improve that
                ratio over time, identically to vehicles that adopt PHEV technology
                during the analysis period.\1146\
                ---------------------------------------------------------------------------
                 \1146\ Meszler Engineering Services, Attachment 2, NPRM Docket
                No. NHTSA-2018-0067-11723 at 32.
                ---------------------------------------------------------------------------
                 The agencies must use the SAE J1711 method for determining the fuel
                economy for the PHEV systems. The use of SAE J1711 and the underlying
                duel fuel vehicle fuel economy calculations are defined by
                statute.\1147\ However, it is important to note that PHEVs are not
                excluded from applying greater range technologies within the PHEV
                technology paths; that is, a PHEV with a lower AER can progress to
                become a PHEV with a longer AER.
                ---------------------------------------------------------------------------
                 \1147\ 49 U.S.C. 32901(b)(1).
                ---------------------------------------------------------------------------
                 CARB commented that several aspects of the agencies' PHEV modeling
                contributed to increased PHEV costs. CARB stated that the electric
                motors were oversized, that all-electric vehicle efficiencies were low,
                and that the lack of battery resizing for road load reductions other
                than mass reduction resulted in battery energy capacities much higher
                than production vehicles.\1148\ CARB stated the modeled battery
                capacity to achieve a given range (kWh/mi) was larger than what exists
                on several representative production vehicles.
                ---------------------------------------------------------------------------
                 \1148\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 149. Specific comments related to costs
                are discussed in Section VI.C.3.e) Overview of Electrification
                Costs, below.
                ---------------------------------------------------------------------------
                 The agencies agreed with CARB's comments that electric motors and
                batteries may be oversized. As a result, the agencies reviewed the
                sizing algorithms and methods used in the NPRM analysis and updated the
                model for the final rule analysis. The updates resulted in smaller
                motor sizes and battery pack sizes for electrified powertrains, as
                discussed above. In addition, the review also resulted in a change to
                the range categories used for the PHEVs in the final rule analysis; the
                final rule analysis classified PHEVs into two levels: (1) PHEV20
                indicating a vehicle with an AER of 20 miles; and (2) PHEV50 indicating
                a vehicle with AER of 50 miles. For more discussion on the change in
                classifications see Section VI.C.3.a)(1)(e) Electrification
                Technologies, Plug-in Hybrids.
                 BorgWarner commented that ``PHEVs and HEVs are complex systems and
                should be modeled in detail,'' and further provided, ``[t]herefore,
                modeling should be inclusive of all approaches of PHEV and HEV and not
                be limited only to Atkinson Cycle engines.'' \1149\ In response, the
                agencies created additional powertrain options for PHEV technologies
                for the final rule analysis. The additional PHEV technologies included
                a plug-in HEV using a turbocharged engine. The additional PHEV paths
                used in the final rule analysis are described in Section
                VI.C.3.a)(1)(e) Electrification Technologies, Plug-in Hybrids.
                ---------------------------------------------------------------------------
                 \1149\ BorgWarner, BorgWarner NPRM public comments 10-26-2018
                Final, Docket No. NHTSA-2018-0067-11895, at 10.
                ---------------------------------------------------------------------------
                d) Battery Electric Vehicles
                 Battery electric vehicles (BEVs) are vehicles with all-electric
                drive and with vehicle systems powered by energy-optimized batteries
                charged primarily from grid electricity. The effectiveness
                [[Page 24490]]
                of BEV powertrains is dependent on the efficiency of the components
                that transfer power from the battery to the driven wheels. These
                components include the battery, electric machine, power electronics,
                and mechanical gearing. For the analysis, electric machine efficiency
                was based on efficiency maps derived from actual electrified vehicles,
                and was scaled such that the peak efficiency value corresponded to the
                latest state-of-the-art technologies. The range of the battery electric
                vehicles depends on the vehicle's class and the battery pack size. For
                the NPRM analysis, manufacturers could apply BEV technology with an AER
                of 200 miles. As discussed previously, the final rule analysis added a
                BEV 300 to reflect vehicles in the market for the MY 2017 analysis
                fleet. For further detailed discussion of how BEV sub-models are
                simulated in Autonomie see the FRM Argonne model documentation.\1150\
                ---------------------------------------------------------------------------
                 \1150\ FRM ANL Model Documentation, at 4.6, 4.7, 4.13, 4.14, and
                5.8.
                ---------------------------------------------------------------------------
                 The resizing algorithm for BEVs is functionally the same as the
                PHEV algorithm; the difference is that BEVs do not use a combustion
                engine, and thus this component was not included in the BEV algorithm.
                To begin, initial estimates of motor and battery powers were calculated
                based on the criteria of acceleration performance, gradeability
                performance, and vehicle range. Then, the algorithm successively ran
                four simulation loops to fine tune the powertrain size to ensure that
                all performance and operational criteria were maintained. First, the
                BEV motor and battery were sized to power the vehicle through the US06
                cycle. Next, the battery capacity was adjusted to ensure the energy
                content is sufficient to complete a simulated UDDS+HWFET combined
                driving cycle, based on EPA adjustment factors to represent sticker
                values, and meet the vehicle range requirement. Finally, the electric
                motor and battery powers were sized accordingly to meet 0-60 and 50-80
                mph acceleration targets. If either acceleration simulation loop
                resulted in a change to the electric motor size, the algorithm repeated
                all simulation loops. Once the acceleration targets were met without
                any resizing of the electric motors, the algorithm finished. Figure VI-
                35 below shows a simplified sizing algorithm for BEVs.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.214
                 Meszler Engineering Services, commenting on behalf of NRDC, argued
                that the fuel economy for a vehicle adopting BEV technology was
                inappropriately dependent on the petroleum-based fuel economy of the
                transforming vehicle.\1151\ Meszler reiterated that the fuel economy of
                the internal combustion engine that BEV technology replaces does not
                have any impact on the efficiency of the resulting BEV, and the
                electric machine ``should not care'' whether it replaces a high or low
                efficiency engine, and should be modeled accordingly.
                ---------------------------------------------------------------------------
                 \1151\ Meszler Engineering Services, Attachment 2, NPRM Docket
                No. NHTSA-2018-0067-11723 at 33.
                ---------------------------------------------------------------------------
                 The agencies agree with Meszler that BEV effectiveness should be
                independent of the vehicle powertrain it will replace in production.
                This is, in fact, how the vehicle model and simulation was performed in
                Autonomie. Autonomie models the capabilities of each unique full
                vehicle system independently, including BEVs. As BEV technology is
                adopted by vehicles, the CAFE model uses the Autonomie databases to
                determine the added incremental efficiency that will
                [[Page 24491]]
                bring a specific vehicle up to the appropriate level. Since the CAFE
                model considers a variety of vehicle types with differing powertrain
                types, vehicle technology classes, performance criteria, and physical
                properties (curb weight, etc.), each with a different overall
                effectiveness, the observed efficiency increment needed to achieve BEV
                effectiveness will vary with each case. While these increments may
                differ, the final effectiveness of a BEV is independent of the
                powertrain it replaced. The effectiveness used in the CAFE model
                represents the difference between the performance of the full vehicle
                models--the full vehicle model representing the baseline vehicle and
                the full vehicle model representing the end-state with all additional
                fuel economy improving technology applied, as discussed in Section
                VI.B.3 Technology Effectiveness Values.
                 ICCT alleged that the agencies did not assess BEV efficiency
                improvements from road load reductions (i.e., from mass reduction, tire
                rolling resistance, or aerodynamic improvements) to reduce the battery
                and power electronic component sizing costs.\1152\ CARB similarly
                commented that battery packs were improperly sized, resulting in
                underestimation of electrified vehicle effectiveness. CARB stated that
                the NPRM constraints on battery sizing caused electrified vehicles to
                end up with oversized, less cost-effective battery packs. CARB further
                stated that battery designs are more scalable than engines and could
                thus be adjusted by manufacturers even at incremental technology
                steps.\1153\
                ---------------------------------------------------------------------------
                 \1152\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-82.
                 \1153\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 145.
                ---------------------------------------------------------------------------
                 For reference, battery resizing in the NPRM was constrained in the
                same manner as other powertrain components, such as the combustion
                engine. Resizing would typically be associated with a major vehicle or
                engine redesign, which in turn would justify the high costs of changing
                the powertrain. In the NPRM, the battery pack and other powertrain
                components were not resized for other improvements in incremental
                technologies such as AERO and ROLL. The agencies agree that battery
                packs, due to their modularity, should be capable of being resized at
                relatively lower cost and complexity, and thus should not be subject to
                the same resizing restrictions applied to other powertrain components
                such as conventional combustion engines. In consideration of CARB and
                ICCT's comments on battery pack resizing, for the final rule, the
                agencies allowed SHEV, PHEV, and BEV battery packs to be resized at all
                incremental technology steps, including for road load reduction
                technology improvements (aerodynamics, rolling resistance reduction,
                and low levels of mass reduction). This avoided the additional cost and
                range associated with oversized battery packs on BEVs and other
                electrified vehicles.
                 CARB commented that the NPRM analysis oversized battery packs that
                targeted 200-mile label range, resulting in exaggerated battery pack
                costs. CARB also stated that some MY 2016-2018 BEVs exist that have a
                higher efficiency than simulated for BEV200s in Autonomie. They further
                argued that although these vehicles were assigned BEV200s, their actual
                range was greater than 200 miles.\1154\
                ---------------------------------------------------------------------------
                 \1154\ California Air Resources Board, Attachment 2, Docket No.
                NHTSA-2018-0067-11873, at 147.
                ---------------------------------------------------------------------------
                 We agree with CARB that the NPRM modeled and simulated battery
                packs were oversized and that the AERs for BEVs did not match the
                current and expected future vehicle AERs. In response to these
                comments, for the final rule analysis, the agencies removed certain
                constraints from the Autonomie battery sizing algorithm, allowing
                batteries to be sized as function of all road load reduction
                technologies. As discussed earlier, this additional battery sizing is
                feasible due to the modularity of battery pack construction. This
                update allowed the battery pack cost and mass to better reflect the
                actual required energy capacity and power, and improved the efficiency
                of modeled BEVs. The agencies also updated the modeling of electric
                machines used in BEVs to reflect improvements in efficiency.
                Furthermore, the agencies added the BEV300 (with an AER of 300 miles)
                to the final rule analysis, providing a better representation of
                production BEVs with more than 200 miles of range. For more discussion
                on BEV300 and electrification efficiency improvements, see Sections
                VI.C.3.a)(1) Electrification technologies and VI.C.3.d)(1) Electric
                Motors, Power Electronics and Accessory Load.
                e) Fuel Cell Vehicles
                 The fuel-cell system in the analysis was modeled to represent
                hydrogen consumption as a function of the produced power, assuming
                normal-temperature operating conditions with a peak system efficiency
                of 60 percent, including the balance of plant.\1155\ The system's
                specific power is 650 W/kg. The hydrogen storage technology selected
                was a high-pressure tank with a specific weight of 0.04 kg H2/kg, sized
                to provide a 320-mile range on the 2-cycle tests on the basis of
                adjusted energy values.
                ---------------------------------------------------------------------------
                 \1155\ Power needed for supporting components and auxiliary
                systems. The balance of plant in a fuel cell system is the auxiliary
                equipment required to ensure the fuel cell operates as a reliable
                power source. This may include fuel reformers and pumps, for
                example.
                ---------------------------------------------------------------------------
                 The sizing algorithm for FCVs was similar to PHEVs and BEVs, but
                adapted to size the specific components of a FCV powertrain: the
                electric motor, fuel-cell, hydrogen (H2) fuel tank, and
                battery pack. The electric motor drives the wheels needed to propel the
                vehicle. During very low power operation, the battery pack alone powers
                the motor/wheels, depleting the battery charge. At moderate driving
                loads, the fuel-cell provides electrical power (generated by consuming
                stored H2) to the motor and also to charge the battery.
                Under heavy loads, both the fuel cell and battery deliver electric
                power to the motor. To begin, initial estimates of motor, fuel cell,
                and battery powers are calculated based on criteria for acceleration
                performance, gradeability performance, and vehicle range. Then, the
                algorithm successively runs four simulation loops to finetune
                powertrain size, ensuring that all performance and operational criteria
                are maintained. First, the FCV motor and battery are sized to power the
                vehicle through the US06 cycle. Next, the on-board mass of H2 fuel, as
                well as the fuel tank mass are adjusted to ensure the vehicle can
                complete a simulated 2-cycle test and meet the range requirement.
                Finally, the electric motor and fuel cell powers are sized accordingly
                to meet 0-60 and 50-80 mph acceleration targets. If either acceleration
                simulation loop results in a change to the electric motor size, the
                algorithm repeats all simulation loops. Once the acceleration targets
                can be met without any resizing of the electric motor, the algorithm
                completes. Figure VI-36 below shows a simplified sizing algorithm for
                FCVs.
                [[Page 24492]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.215
                 The agencies did not receive comments on FCV modeling in Autonomie.
                For the final rule analysis, the agencies used the same FCV model and
                simulations to estimated effectiveness values.
                e) Electrification Costs
                 The primary factors that influence the cost and effectiveness of
                hybrid or battery electric vehicles are the cost and efficiency of the
                energy storage components and electric machines. Energy storage
                components include battery cells, battery management systems, and
                thermal management systems. The electric machine components include
                electric motors, power electronics, controllers, and other devices that
                support thermal management.
                 Charging infrastructure is an essential component for PHEVs and
                BEVs, and may add to the total cost of ownership of the vehicle.
                However, most households are equipped with a 110-volt outlet for level
                1 charging, for which no additional cost is incurred. Installing a
                level 2 charging outlet (220-volt) will add cost to the total ownership
                of the vehicle but decreases charging time. The price of level 2
                residential charging equipment varies, but typically ranges from $500
                to $2,000 before installation and state or utility incentives.\1156\
                ---------------------------------------------------------------------------
                 \1156\ U.S. Department of Energy Office of Energy Efficiency and
                Renewable Energy, Charging at Home, https://www.energy.gov/eere/electricvehicles/charging-home (last visited March 20, 2020).
                ---------------------------------------------------------------------------
                 For this final rule analysis, the agencies used Argonne's BatPaC
                modeling tool to develop battery pack manufacturing costs as well as
                weight.\1157\ Battery packs were sized in terms of the vehicle's energy
                and power requirement and costs were estimate for each of the simulated
                technology combinations. The Argonne team used BatPaC to create a
                ``lookup table'' with battery pack size (energy and power) and cost as
                well as weight data for the full vehicle simulations to ``reference,''
                to avoid the need for conducting a full BatPaC simulation for each
                unique vehicle modeled in the analysis. The table included cost data
                for each technology key and vehicle technology classes. As discussed
                below, Autonomie runs linearly interpolate between points in the lookup
                tables when deriving final values from BatPaC, the differences between
                using BatPaC for each configuration and the interpolation using the
                lookup table was insignificant.
                ---------------------------------------------------------------------------
                 \1157\ The agencies used BatPaC version 3.0 (released in 2015)
                for the NPRM and BatPaC version 3.1 (June 2018) for the final rule.
                ---------------------------------------------------------------------------
                 The agencies used the cost of electric machines from U.S. DRIVE's
                October 2017 report, ``Electrical and Electronics Technical Team
                Roadmap.'' In industry, manufacturers use different types of electric
                machines resulting in a range of actual costs for the systems. To
                capture this range, the agencies considered a single type of high
                efficiency electric machine, representative of the range of technology
                available in the rulemaking timeframe, uniquely sized for each of the
                simulated combinations. For the final rule analysis, the cost of the
                electric machine was determined using a dollar-per-kilowatt metric. The
                agencies sized the electric machines using the method discussed in
                Section VI.C.3.d) Electric Effectiveness Modeling and Resulting
                Effectiveness Values.
                 The following sections discuss the method used for modeling battery
                and non-battery component costs, the learning curves applied to those
                costs, and the total costs for each type of electrification technology
                considered in this final rule analysis.
                (l) Battery Pack Modeling
                 BatPaC is a software designed for policymakers and researchers
                interested in estimating the manufacturing cost of lithium-ion
                batteries for electric drive
                [[Page 24493]]
                vehicles.\1158\ BatPaC is used to estimate the cost of manufacturing
                lithium-ion batteries and examine trade-offs that result from different
                battery performance specifications such as power and energy capacity.
                BatPaC includes a library of lithium ion electrode combinations and
                inputs for all the parameters associated with materials and
                manufacturing operations in a factory.
                ---------------------------------------------------------------------------
                 \1158\ BatPaC: Battery Manufacturing Cost Estimation, Argonne
                National Laboratory, https://www.anl.gov/tcp/batpac-battery-manufacturing-cost-estimation.
                ---------------------------------------------------------------------------
                 Specifically, BatPaC models stiff-pouch, laminated prismatic format
                cells, placed in double-seamed, rigid modules. The model supports
                liquid- and air-cooling, accounting for the resultant structure,
                volume, cost, and heat rejection capacity. The model considers cost of
                capital equipment, plant area and labor for each step in the
                manufacturing process. The model places relevant limits on electrode
                coating thickness, and considers limits applicable to current and near-
                term manufacturing processes. The model also considers annual pack
                production volumes and economies of scale for high-volume production.
                 BatPaC calculations are based on a generic pack designs that
                reasonably represents the weight and manufacturing cost of batteries
                deployed commercially. The advantage of using this approach is the
                ability to model wide range of commercial design specifications for the
                various classes of vehicles. This modeling approach is particularly
                advantageous because the data from commercially available battery packs
                is limited and varies widely with respect to the underlying
                specifications (power and energy) and constraints (mass, volume,
                dimensions, durability) set by the manufacturer.
                 BatPaC is a Microsoft Office Excel spreadsheets-based model. The
                data needed to design and build a battery pack, such as dimensions of
                the cell, estimate of materials, and manufacturing cost, are provided
                in the model, with the manufacturing costs for the designed battery
                based on a ``baseline plant'' designed for a battery of intermediate
                size and production scale so as to establish a center-point for other
                designs. BatPaC can be configured with alternative chemistries,
                charging constraints, battery configurations, production volumes, and
                cost factors for other battery designs by customizing these parameters
                in the modeling tool.
                 For this analysis, running individual BatPaC simulations for each
                full vehicle simulation requiring an electrified powertrain would have
                been computationally intensive and impractical, given that
                approximately 750,000 simulated vehicles out of the 1.2 million total
                simulated vehicles had an electrified powertrain. Accordingly, staff at
                Argonne built ``lookup tables'' with BatPaC to provide battery pack
                manufacturing costs, battery pack weights, and battery pack cell
                capacities for vehicles modeled in the large-scale simulation runs.
                 To build the lookup tables, Argonne staff selected a range of
                minimum and maximum values for battery pack power (kW) and battery pack
                energy (kWh) for each vehicle powertrain based on a combination of
                market analysis and analysis of the Autonomie simulations that were run
                for the NPRM and final rule. The performance requirements (vehicle
                acceleration times, EV range, etc.) were defined from set assumptions
                and validated from existing vehicles.\1159\ The range, as well as the
                number of power and energy points considered to generate each lookup
                table, varies across powertrains. The minimum and maximum power and
                energy values have been selected to encompass current designs. For
                example, one end of the spectrum is representative of the MY 2016-2017
                Tesla Model S 100D (100 kWh total battery energy, 335-mile range),
                while the other end of the spectrum is representative of the 2017
                Mitsubishi iMiEV (16 kWh total battery energy, 62-mile range). The
                components were then sized in Autonomie across all vehicle classes to
                define the minimum and maximum values to be considered, as shown in
                Table VI-90.
                ---------------------------------------------------------------------------
                 \1159\ See Final Rule Argonne Model Documentation Section 5.9,
                Battery Performance and Cost Model (BatPaC).
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.216
                 Figure VI-37 illustrates the inputs generated in Autonomie to
                create the BatPaC-based lookup tables, and the outputs characterized in
                the BatPaC-based lookup tables that are used to provide estimates
                referenced in the agencies' analysis. A linear interpolation was then
                performed in MATLAB to determine the associated values for battery pack
                manufacturing cost, weight, and cell capacity.
                [[Page 24494]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.217
                 Figure VI-38 shows the linear relationship between cost, power, and
                weight used to generate the compact passenger car BEV200 technology
                class lookup table presented in Figure VI-39. As seen from the figures
                below, the energy values produced by BatPaC consist of a fairly linear
                relationship with respect to power and energy for a vehicle class.
                Since Autonomie runs would linearly interpolate between the points in
                the lookup tables when deriving the final values from BatPaC, the
                differences between using BatPaC for each configuration and the
                interpolation using the lookup table were insignificant.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.218
                BILLING CODE 4910-59-C
                 Figure VI-39 details the estimates of $ per kWh at the pack level
                generated from the lookup table for BEV200 compact cars used in the
                final rule analysis. As discussed further below, the specific battery
                costs for each simulated vehicle were presented for the NPRM (and now
                for the final rule) in the docketed Argonne assumptions files and in
                the vehicle simulation database included in the CAFE model.
                [[Page 24495]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.219
                 During the Autonomie large-scale simulation runs, calling the
                BatPaC model for each individual simulation would have been
                computationally intensive. Using the MATLAB lookup tables reduced the
                time to run the approximately 750,000 simulations significantly, which
                in turn reduced the total simulation run time for all of the technology
                combinations by several days with insignificant impact on the
                analytical results.
                (a) BatPaC Inputs and Assumptions
                 The Argonne documentation describing the analysis performed for the
                NPRM, ``A Detailed Vehicle Simulation Process To Support CAFE
                Standards,'' detailed the specific assumptions that Argonne's experts
                used to simulate batteries and their associated costs for the full
                vehicle simulation modeling.\1160\ In addition, detail on the NPRM
                electrification analysis was presented in the PRIA.\1161\ While the
                Argonne Summary of Main Component Assumptions Excel file correctly
                identified the chemistry used in the NPRM analysis as NMC333,\1162\ the
                PRIA inadvertently described that NMC441 was used. The agencies
                presented selected lookup table battery cost values in the Argonne
                Summary of Main Component Assumptions Excel file,\1163\ as shown above,
                and the specific battery costs for each simulated vehicle were
                presented for the NPRM and final rule in the vehicle simulation
                database included in the CAFE model.
                ---------------------------------------------------------------------------
                 \1160\ Islam S. Ehsan. Moawad, Ayman. Kim, Namdoo. Rousseau,
                Aymeric. ``A Detailed Vehicle Simulation Process to Support CAFE
                Standards.'' ANL/ESD-18/6. Energy Systems Division, Argonne National
                Laboratory (2018).
                 \1161\ PRIA at 362-384.
                 \1162\ ANL--All Assumptions Summary, NHTSA-2018-0067-0005.
                 \1163\ ANL--Summary of Main Component Performance Assumptions
                NPRM, NHTSA-2018-0067-0003.
                ---------------------------------------------------------------------------
                 Several commenters claimed that costs for electrification
                technologies were too high, especially regarding battery costs (note
                that comments on non-battery component costs are addressed separately
                in Section VI.C.3.e)(2) Non-battery Electrification Component Costs,
                below).\1164\ Several commenters pointed to text in interagency review
                documents that stated the NPRM battery modeling costs were higher than
                what EPA recommended,\1165\ and higher than what EPA had obtained from
                the most recent version of the BatPaC model.\1166\
                ---------------------------------------------------------------------------
                 \1164\ Meszler Engineering Services, NHTSA-2018-0067-11723
                Attachment 2; National Coalition for Advanced Transportation, NHTSA-
                2018-0067-11969; Workhorse Group Inc., NHTSA-2018-0067-12215;
                International Council on Clean Transportation, NHTSA-2018-0067-
                11741; California Air Resources Board, NHTSA-2018-0067-11873.
                 \1165\ California Air Resources Board, NHTSA-2018-0067-11873.
                 \1166\ Boulder County Public Health et al., NHTSA-2018-0067-
                11975; International Council on Clean Transportation, NHTSA-2018-
                0067-11741.
                ---------------------------------------------------------------------------
                 CARB commented that the agencies incorrectly identified and
                assessed existing technologies, improperly oversized components and
                batteries for the modeled vehicle classes, and underestimated
                technology efficiency through improper modeling.\1167\ CARB also
                submitted supplemental comments (discussed further, below) stating that
                the PRIA and the underlying modeling were inconsistent regarding which
                exact battery chemistries were modeled for every electrified model in
                the fleet, which CARB argued was crucial for understanding the battery
                compositions and thus their production costs.\1168\
                ---------------------------------------------------------------------------
                 \1167\ California Air Resources Board, NHTSA-2018-0067-11873.
                 \1168\ California Air Resources Board, NHTSA-2018-0067-4166.
                ---------------------------------------------------------------------------
                 ICCT stated that the agencies misrepresented the leading research
                on both battery and electric vehicle costs, with the result being that
                electric vehicles were so costly that they were modeled to remain at
                approximately the same penetration in 2025 with the Augural 2025 fuel
                economy and adopted 2025 CO2 standards, as they were in mid-
                2018 (i.e., between 1.5 percent and 2 percent of new vehicle
                sales).\1169\ ICCT stated that the agencies' inputs failed to reflect
                the latest industry data on future potential electric vehicle cost
                parity with combustion vehicles. ICCT commented that through a
                combination of incorrectly high electric vehicle prices (which, they
                argue, do not reflect Argonne or other leading battery research groups'
                work), and modeling restrictions on electric vehicles, the agencies
                unduly inflated technology costs of electric vehicles to comply with
                the standards. ICCT argued that although the agencies purported to use
                state-of-the-art tools like the BatPaC model for battery costs, the
                cost calculations erroneously pushed up electric vehicles' incremental
                costs above $10,000 per vehicle. ICCT claimed that the agencies
                introduced errors that artificially pushed up the battery costs higher
                than indicated by BatPaC and other experts in the field.
                ---------------------------------------------------------------------------
                 \1169\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741.
                ---------------------------------------------------------------------------
                 NCAT noted that the PRIA described some ways in which the modeling
                increased battery costs, namely, that the battery pack costs were
                adjusted
                [[Page 24496]]
                upwards, the cost of the battery management system increased, and a
                cost for a battery automatic and manual disconnect unit was
                added.\1170\ Regardless, NCAT stated that the agencies analysis was not
                sufficiently transparent, and argued that the battery costs were
                significantly overestimated in the modeling supporting the NPRM.
                Boulder County Public Health and other Colorado municipal organizations
                claimed that overstated battery costs had the effect of
                mischaracterizing and downplaying the benefits of increased numbers of
                electric vehicles as part of the vehicle fleet.\1171\ Commenters also
                argued that discrepancies existed between battery costs used in the
                rulemaking documents and battery costs found in the Argonne database,
                referring specifically to BISG and CISG costs (discussed further
                below).\1172\
                ---------------------------------------------------------------------------
                 \1170\ National Coalition for Advanced Transportation, NHTSA-
                2018-0067-11969, citing PRIA at 366-67.
                 \1171\ Boulder County Public Health et al., NHTSA-2018-0067-
                11975.
                 \1172\ Meszler Engineering Services, NHTSA-2018-0067-11723
                Attachment 2; International Council on Clean Transportation, NHTSA-
                2018-0067-11741.
                ---------------------------------------------------------------------------
                 In addition to comments claiming that the agencies' battery cost
                projections were incorrect or difficult to interpret, many commenters
                submitted general information about the state of battery technology and
                cost advances now and as projected into the future. For example, NCAT
                stated that battery technology has improved and battery costs have
                fallen dramatically, due in part to reduced material costs,
                manufacturing improvements, and higher manufacturing volumes.\1173\ In
                compliment, NCAT asserted that the demand for EVs is growing
                ``dramatically.''
                ---------------------------------------------------------------------------
                 \1173\ National Coalition for Advanced Transportation, NHTSA-
                2018-0067-11969. NCAT also stated that the increase in mass
                manufacturing of lithium-ion storage is expected to continue to
                reduce battery prices.
                ---------------------------------------------------------------------------
                 ICCT stated that the agencies' analysis of electric vehicle costs
                and the resulting extremely low penetration levels was not in line with
                automakers' announcements, which included statements that they would
                produce far greater numbers of electric vehicles to comply with
                standards around the world.
                 ICCT summarized projections of electric vehicle battery costs for
                2020-2030, and stated that the agencies did not analyze the studies and
                automaker announcements they cited to understand the potential for
                cost-effective electric drive technology.\1174\ ICCT stated the data
                they reviewed included a variety of different technologies, production
                volumes, and cost elements, and although there were differences in
                methods for each, ``they generally include in some variation of
                material, process, overhead, depreciation, warranty, and profit
                costs.'' ICCT summarized the results of their review, projecting that
                battery pack costs will decline to $150/kWh by 2020-2023 and then to
                about $120-$135/kWh by 2025, with the exception of Tesla, which reports
                costs of $150 kWh in 2018 and projected costs of $100/kWh by 2022. ICCT
                stated that the results of this review were corroborated in the
                aforementioned EPA interagency comments on battery costs used in the
                proposal.
                ---------------------------------------------------------------------------
                 \1174\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741.
                ---------------------------------------------------------------------------
                 NCAT stated that the average price of a battery pack dropped from
                $1,000/kWh in 2010 to $209/kWh in 2017, demonstrating a decrease of 79
                percent in seven years.\1175\ NCAT stated Tesla is on track to achieve
                $100/kWh by the end of 2018, and Audi has been buying batteries at
                $114/kWh, according to trade press reports.\1176\ NCAT also cited BNEF
                analyses showing that battery costs are projected to continue to
                decline substantially,\1177\ specifically projecting a decrease in
                battery cost of 77 percent between 2016 and 2030. Accordingly, NCAT
                stated that EVs will be less expensive to buy than conventional
                gasoline vehicles by 2025 in the United States.\1178\ Workhorse
                similarly echoed the assertion that EV costs will reach parity with
                conventional vehicle costs before 2025.\1179\
                ---------------------------------------------------------------------------
                 \1175\ National Coalition for Advanced Transportation, NHTSA-
                2018-0067-11969, citing Bloomberg New Energy Finance, ``Electric
                Vehicle Outlook: 2018,'' https://bnef.turtl.co/story/evo2018?teaser=true.
                 \1176\ National Coalition for Advanced Transportation, NHTSA-
                2018-0067-11969, citing Fred Lambert, ``Tesla to achieve leading
                $100/kWh battery cell cost this year, says investor after
                Gigafactory 1 tour'' (Sept. 11, 2018), https://electrek.co/2018/09/11/tesla-100-kwh-battery-cost-investor-gigafactory-1-tour/.
                 \1177\ National Coalition for Advanced Transportation, NHTSA-
                2018-0067-11969, citing Bloomberg New Energy Finance, ``Electric
                Vehicle Outlook: 2018,'' https://bnef.turtl.co/story/evo2018?teaser=true.
                 \1178\ National Coalition for Advanced Transportation, NHTSA-
                2018-0067-11969, citing Jess Shankleman, ``Pretty Soon Electric Cars
                Will Cost Less Than Gasoline'' (May 26, 2017), https://www.bloomberg.com/news/articles/2017-05-26/electric-cars-seen-cheaper-than-gasoline-models-within-a-decade; Jess Shankleman, ``The
                Electric Car Revolution Is Accelerating'' (July 6, 2017), https://www.bloomberg.com/news/articles/2017-07-06/the-electric-car-revolution-is-accelerating. NCAT also noted that the up-front cost
                parity does not take into consideration the fuel savings and
                maintenance savings over the lifetime of EV use as compared to
                gasoline vehicle use.
                 \1179\ Workhorse Group Inc., NHTSA-2018-0067-12215.
                ---------------------------------------------------------------------------
                 NCAT also cited the ICCT Efficiency Technology and Cost Assessment,
                which concluded that, primarily because of rapid developments in
                battery pack technologies, EV costs will be reduced by $4,300-$5,300
                per vehicle by 2025 compared to EPA's prior estimates in support of the
                MY 2017-2025 standards.\1180\ In that report, ICCT concluded that
                battery costs of $140/kWh is a realistic estimated value by 2025, as
                compared with EPA estimates in the 2016 Mid-Term Evaluation (MTE)
                analysis of $180-200/kWh.\1181\
                ---------------------------------------------------------------------------
                 \1180\ National Coalition for Advanced Transportation, NHTSA-
                2018-0067-11969, citing ICCT, ``Efficiency Technology and Cost
                Assessment for U.S. 2025-2030 Light-duty Vehicles'' (Mar. 2017) at
                11, 15, available at http://www.theicct.org/US-2030-technology-cost-assessment.
                 \1181\ Id.
                ---------------------------------------------------------------------------
                 NCAT also cited improvements in manufacturing techniques,
                specifically by Tesla, as an example of how batteries are being
                manufactured in large volumes with high quality at low cost.\1182\ NCAT
                stated that in mid-2018, Tesla was producing batteries at its
                Gigafactory 1 facility at an annualized rate of roughly 20 GWh, making
                it the highest-volume battery plant in the world.\1183\ NCAT and other
                commenters also cited Bloomberg's New Energy Finance research stating
                that the average energy density of EV batteries is improving at around
                5-7 percent per year.
                ---------------------------------------------------------------------------
                 \1182\ National Coalition for Advanced Transportation, NHTSA-
                2018-0067-11969, citing Tesla, Inc., S.E.C. Form 10-K (Feb. 22,
                2018) at 3-4, available at https://www.sec.gov/Archives/edgar/data/1318605/000156459018002956/tsla-10k-20171231.htm.
                 \1183\ National Coalition for Advanced Transportation, NHTSA-
                2018-0067-11969, citing Tesla, ``Tesla Gigafactory,'' https://www.tesla.com/gigafactory (last visited Oct. 25, 2018).
                ---------------------------------------------------------------------------
                 Finally, Workhorse commented that they have more than ten years of
                experience in the field of designing and assembling battery packs, and
                their business plans are predicated on battery costs much lower than
                assumed by the agencies.\1184\
                ---------------------------------------------------------------------------
                 \1184\ Workhorse Group Inc., NHTSA-2018-0067-12215.
                ---------------------------------------------------------------------------
                 As explained above, the agencies consulted with and relied on
                Argonne battery experts to develop inputs to the BatPaC model and
                generate the battery cost lookup tables used as references for the
                Autonomie full-vehicle simulations, as detailed in Argonne's
                documentation supporting the NPRM analysis.\1185\ As explained further
                below, the agencies also directed CARB to information about the NPRM
                battery cost analysis available
                [[Page 24497]]
                in the public docket in response to their FOIA request.
                ---------------------------------------------------------------------------
                 \1185\ Islam S. Ehsan. Moawad, Ayman. Kim, Namdoo. Rousseau,
                Aymeric. ``A Detailed Vehicle Simulation Process to Support CAFE
                Standards.'' ANL/ESD-18/6. Energy Systems Division, Argonne National
                Laboratory (2018).
                ---------------------------------------------------------------------------
                 Commenters are correct that the EPA Draft TAR and Proposed
                Determination estimates for battery sizing and cost were different than
                the NPRM analysis. For the Draft TAR and in the Proposed Determination,
                a separate battery and motor sizing spreadsheet was built to determine
                the energy and power requirements for PHEVs and BEVs at different curb
                weights, and then BatPaC was used to determine specific energy (kWh/kg)
                and the battery pack cost estimate.\1186\ For this NPRM and final rule,
                the energy requirements for PHEVs and BEVs were determined using
                Autonomie simulations with the integrated BatPaC lookup table to select
                the appropriate battery pack size, cost, and weight. As discussed in
                Sections VI.B.3.a)(4) How Autonomie Sizes Powertrains for Full Vehicle
                Simulation and VI.B.3.a)(6) Performance Neutrality, the Autonomie full-
                vehicle simulation modeling assessed metrics to ensure performance
                requirements were met for every modeled vehicle. Appropriately
                accounting for vehicle metrics and individual vehicle power and weight
                requirements resulted in some of the differences observed between the
                Draft TAR and Proposed Determination estimates and the estimates
                presented in the NPRM and this final rule.
                ---------------------------------------------------------------------------
                 \1186\ Draft TAR at 5-315.
                ---------------------------------------------------------------------------
                 For the final rule, the agencies considered these public comments,
                market observations, literature, industry reports, and additional
                research. In addition, as described further below and in the Argonne
                documentation accompanying this final rule, Argonne consulted the
                A2Mac1 database for additional data points on batteries that were used
                to inform the final rule battery cost modeling.
                 As discussed above, BatPaC version 3.0 was used for the NPRM
                analysis because that was the most up-to-date version of BatPaC
                available at the time the NPRM analysis was being conducted. BatPaC
                version 3.1, released after the NPRM analysis was completed, was used
                for this final rule because that was the most up-to-date version of
                BatPaC available at the time the final rule analysis was being
                conducted.
                 The agencies note that BatPaC version 4.0 has been released since
                the analysis was completed for this final rule. Specifically, that
                version was released on January 14, 2020, after the rule had been
                submitted for interagency review. The default battery chemistry in
                BatPaC version 4.0 continues to be NMC622, which as discussed further
                in Section (i) below, reflects the reasonable assumption this chemistry
                will likely continue to be used in the rulemaking timeframe based on
                its commercial application and market trends towards higher-nickel,
                lower-cobalt content chemistries.\1187\ As explained in this section,
                and further in Section (c) below, the agencies' modeled costs for
                battery packs aligns with current industry estimates and closely tracks
                future projections of battery pack costs from the Department of
                Energy's Vehicle Technology Office (DOE VTO) lab
                targets.1188 1189
                ---------------------------------------------------------------------------
                 \1187\ The agencies note that BatPaC version 4.0 provides a new
                option to build battery packs with NMC811.
                 \1188\ Freyermuth, Vincent. Rousseau, Aymeric. ``Impact of
                Vehicle Technologies Office Targets on Battery Requirements.'' ANL/
                ESD-16/22. Energy Systems Division, Argonne National Laboratory
                (2016).
                 \1189\ Hummel et al., UBS Evidence Lab Electric Car Teardown--
                Disruption Ahead?, UBS (May 18, 2017), https://neo.ubs.com/shared/d1ZTxnvF2k/.
                ---------------------------------------------------------------------------
                 In addition to using BatPaC version 3.1 for this final rule, BatPaC
                assumptions were updated to reflect what the Argonne battery experts
                and the agencies believed would be representative and attainable of
                battery manufacturing trends in the rulemaking timeframe. Section (ii)
                provides additional information on BatPaC inputs and assumptions that
                were updated for the final rule based on public comments and the
                agencies own market observations and additional research. In addition,
                as discussed further below, for the final rule, the calculated battery
                pack weight and manufacturing cost was compared with the battery pack
                cost and weight data obtained through various benchmarking studies. The
                agencies believe that the Argonne methodology for producing the
                hundreds of thousands of battery pack cost estimates required for the
                full-vehicle modeling and simulation resulted in reasonable estimates
                of battery pack costs. The following sections provide additional
                context and response to comments on specific BatPaC inputs and
                assumptions used in the NPRM and final rule.
                (i) Chemistry
                 The choice of chemistry for battery cells depends on the
                application and consideration of cost, energy density, and safety,
                among other factors. The PRIA described the battery pack cell chemistry
                used for different powertrain types modeled in the NPRM analysis.\1190\
                For Micro HEVs, BISG HEVs, CISG HEVs, and Full HEVs, the agencies used
                LFP-G, rather than LMO-G, because the latter has a limited lifespan
                which is expected to degrade functionality over a vehicle's lifetime,
                and has greater limitations on available ranges of battery charge and
                discharge rates. As described above, for PHEVs and BEVs, the Argonne
                ``Summary of Main Component Performance Assumptions'' file correctly
                stated that NMC333 was used, however the PRIA misstated that NMC441 was
                used.
                ---------------------------------------------------------------------------
                 \1190\ PRIA at 373.
                ---------------------------------------------------------------------------
                 Both UCS and CARB commented on the agencies' choice of battery
                chemistry, with UCS noting that this choice can have a large impact on
                performance and materials costs, and therefore on the modeled cost of
                drivetrain electrification.
                 First, both commenters stated that the NPRM documentation was
                inconsistent and unclear. UCS noted the discrepancy between the PRIA
                and Argonne model documentation, and also that the rulemaking documents
                stated the most recent version of Argonne's BatPaC model was used to
                estimate battery costs, but the default lithium ion chemistry in the
                current BatPaC model is NMC622. UCS stated the choice of NMC variant
                effects battery costs, as NMC622 replaces more expensive cobalt with
                nickel. UCS further stated it was not possible to determine the
                magnitude of the cost error in the PHEV and BEV battery pack costs,
                only that the costs were likely higher than current battery cost data
                supported.
                 CARB stated that the agencies' selected battery chemistries
                represented a step backward from previous analysis done for the Draft
                TAR. CARB claimed that the biggest lithium-ion production companies
                have indicated that they will use NMC811 for BEVs, and therefore NMC441
                or NMC333 would not represent current technology going into BEVs or
                near-future BEV battery technology. CARB stated that NMC811 technology
                was expected to come to market in 2019, which is far sooner than
                anticipated, even in the agencies' prior analyses.
                 Commenters also noted that the chemistry chosen for mild and strong
                hybrids differed from what is used in current and announced HEVs. UCS
                stated that all non-plug-in hybrids in the proposed rule analysis used
                lithium iron phosphate (LFP) chemistry, but in practice, most hybrids
                on the road did not use this chemistry. UCS referenced the Toyota Prius
                and the new RAM 1500 pickup as examples of vehicles that do not use LFP
                chemistry. CARB similarly stated that the NPRM battery chemistry
                selection for PHEV and strong hybrid batteries does not represent many
                of the batteries that are being deployed in the market, nor have been,
                for several years now, but did not provide an alternative chemistry
                they believed to be better
                [[Page 24498]]
                represented in the market. CARB stated that this resulted in a
                ``misappropriation of higher costs for electrification technologies in
                the Agencies' analysis, and further highlights the Agencies' sudden
                lack of knowledge about electrification, despite the far more
                directionally correct projections in previous analysis for the 2016
                Draft TAR and EPA's Proposed Determination.''
                 Similarly, UCS pointed to a discrepancy in strong hybrid battery
                costs between the proposed rule estimates (greater than $1,200, even
                for the small car classes) and an estimate from Argonne in 2017 ($614),
                to argue that the lack of detailed information made it impossible to
                determine if the choice of battery chemistry was responsible for the
                discrepancy.
                 The agencies carefully considered these comments. As stated above,
                the agencies disagree that the discrepancy in the Argonne Summary of
                Main Component Performance Assumptions file and the PRIA over the use
                of NMC333 for the NPRM analysis limited commenters ability to comment
                on battery chemistry, as both UCS and CARB communicated a belief that
                the agencies choice of battery chemistry contributed to the overstated
                battery costs in the NPRM. The agencies understand how the choice of
                chemistry impacts battery costs, and many of the commenters' concerns
                intertwined the NPRM choice of battery chemistry with the NPRM battery
                costs. Here, the agencies respond to comments on the choice of
                chemistries. The agencies will also discuss costs below.
                 As stated earlier, although manufacturers use different battery
                chemistries in various HEV, PHEV, and BEV applications, the choice of
                chemistry for a given application depends on several factors including
                safety, stability, and functional requirements (high power or high
                energy requirements for performance) of the battery pack. In
                determining whether to select one battery chemistry over another, the
                agencies concluded that using commercially proven technologies that
                represented the current cost of production was more reasonable than
                assuming additional technologies would come to fruition during the
                rulemaking timeframe, and attempting to project the cost and
                effectiveness of such technologies. While there is ongoing research and
                development in battery chemistry and in other battery related
                technologies that have the potential to reduce costs and increase
                battery capacity, these technologies have yet to be proven viable for
                commercial use.\1191\
                ---------------------------------------------------------------------------
                 \1191\ Recent Advances in Energy Chemical Engineering of Next-
                Generation Lithium Batteries, Engineering, Volume 4, Issue 6
                (December 2018), at 831-847. Available at https://www.sciencedirect.com/science/article/pii/S2095809918312177. Some
                examples include lithium-sulfur battery cell chemistry and solid-
                state electrolyte battery cells.
                ---------------------------------------------------------------------------
                 In addition, as discussed throughout this document, the agencies
                considered technologies that manufacturers could use to comply with
                standards in the rulemaking timeframe that reasonably represented the
                state of technology across the industry. While the battery chemistries
                used in commercial vehicles are largely confidential business
                information, proprietary teardown reports are one source of information
                used to learn more about the chemistries actually employed in the
                market. For both the NPRM and final rule, the agencies consulted
                Argonne's battery experts to determine the chemistries that should be
                modeled in the BatPaC analysis. Argonne consulted A2Mac1 battery pack
                teardown reports, which confirmed that indeed, manufacturers use a
                range of chemistries across the electrified vehicle types. Selecting
                battery chemistries that can reasonably represent the range employed in
                the market ensured that the analysis better captured the average of
                costs across the industry.
                 For example, in addition to the reasons listed in the NPRM, LFP has
                been proven in commercial use, as identified in literature and battery
                teardown reports.\1192\ This presented a basis for using LFP, as the
                chemistry was reasonably representative of chemistries used in mild and
                strong hybrids at the time of the analysis. The agencies also
                considered that LFP's lower cost compared to other potential HEV
                battery chemistries (contrary to commenters' statements) made it more
                attractive for vehicles with tight cost constraints, even with the
                associated lower energy density.
                ---------------------------------------------------------------------------
                 \1192\ Details of cell chemistry and battery cooling system are
                described in Nelson, Paul A., Gallagher, Kevin G., Bloom, Ira D.,
                and Dees, Dennis W. Modeling the Performance and Cost of Lithium-Ion
                Batteries for Electric-Drive Vehicles--SECOND EDITION (2012),
                available at https://publications.anl.gov/anlpubs/2015/05/75574.pdf.
                ---------------------------------------------------------------------------
                 Similarly, although EPA selected NMC622 as the modeled battery
                chemistry for the Draft TAR, manufacturers were also using other NMC
                chemistries in hybrid and BEV applications in that timeframe depending
                on the required application. The chemistry selected for the NPRM,
                NMC333, was selected based on proprietary teardown reports that
                demonstrated the chemistry's commercial use: a survey of twelve MY 2013
                to MY 2018 HEVs, PHEV, and BEVs showed that NMC333 was used in eleven
                of those vehicles, and NMC622 was only used in one.\1193\
                ---------------------------------------------------------------------------
                 \1193\ A Detailed Vehicle Simulation Process To Support CAFE and
                CO2 Standards for the MY 2021--2025 Final Rule Analysis,
                Section 5.9 Battery Performance and Cost Model (BatPaC), referencing
                A2Mac1 Automotive Benchmarking, https://a2mac1.com.
                ---------------------------------------------------------------------------
                 Accordingly, the agencies believe that assuming LFP-G as the
                modeled cell chemistry for HEVs and NMC333 as the modeled PHEV and BEV
                chemistry for the NPRM analysis of battery costs was not unreasonable,
                based on their demonstrated commercial use in a range of electric
                vehicle applications. However, employing BatPaC version 3.1 for the
                final rule analysis also presented the opportunity to update the
                modeled battery chemistry used to assess battery costs.
                 The agencies similarly consulted Argonne battery experts on battery
                chemistry and trends to inform the final rule analysis. Argonne staff
                used the A2Mac1 database to determine real-world battery chemistry and
                configurations in different electric vehicle applications. As shown in
                the Argonne Full Vehicle Modeling documentation for the final rule, the
                A2Mac1 battery pack teardown analysis provided an array of data points
                on battery chemistries for different electric vehicle applications,
                among other relevant battery pack data, that informed the final rule
                battery analysis.\1194\
                ---------------------------------------------------------------------------
                 \1194\ Id.
                ---------------------------------------------------------------------------
                 In determining which of these chemistries would best represent the
                range of chemistries demonstrated in the market, the agencies
                considered several issues. Due to the increasing manufacturing volume
                of battery packs with NMC, it is expected that NMC battery cells will
                continue to be used in battery packs across different electric vehicle
                applications in the future. The agencies considered concerns about NMC
                formulations with varying cobalt content, and issues including the
                current and future cost of cobalt,\1195\
                [[Page 24499]]
                and the cobalt supply chain.\1196\ These concerns, among others, have
                led to the market shift towards cathode active materials with a higher
                fraction of nickel and less cobalt.\1197\ Manufacturers have
                demonstrated the use of NMC622, which contains more nickel and less
                cobalt than NMC333, in different electric vehicle applications. In
                addition, as CARB noted and has been reported in the news for some
                time, the expected next step in battery chemistries using even less
                cobalt is NMC811. However, the shift to higher-nickel-content
                chemistries is not without challenges; increasing nickel content
                results in lower thermal stability, leading to safety concerns.\1198\
                ---------------------------------------------------------------------------
                 \1195\ See, e.g., MIT Energy Initiative. 2019. Insights into
                Future Mobility, at 78. Cambridge, MA: MIT Energy Initiative (``. .
                . significant uncertainty remains about the steady-state price of
                cobalt in the future as demand and supply continues to increase
                [internal citation omitted]. Under our base case scenario, global
                demand for cobalt in 2030 from new EV sales (even if all EVs use
                batteries with the high nickel content of NMC811) would reach
                approximately 80% of the world's total cobalt output in 2016.
                Considering that only 15% of the worldwide demand for cobalt in 2017
                was used in EV batteries (Jackson 2019), an increase in demand of
                this magnitude might result in higher prices for cobalt. Thus,
                automakers may need to move to different battery chemistries that
                are less reliant on cobalt to avoid raw materials shortages and
                price volatility.'').
                 \1196\ See, e.g., Todd C. Frankel, The Cobalt Pipeline: Tracing
                the path from deadly hand-dug mines in Congo to consumers' phones
                and laptops, Washington Post (Sept. 30, 2016), https://www.washingtonpost.com/graphics/business/batteries/congo-cobalt-mining-for-lithium-ion-battery/?itid=lk_inline_manual_9&tid=lk_inline_manual_9; Peter Whoriskey and
                Todd C. Frankel, Tech giants pledge to keep children out of cobalt
                mines that supply smartphone and electric-car batteries, Washington
                Post (Dec. 20, 2016), https://www.washingtonpost.com/news/the-switch/wp/2016/12/20/tech-giants-pledge-to-keep-children-out-of-cobalt-mines-that-supply-smartphone-and-electric-car-batteries/.
                 \1197\ See, e.g., Gohlke, David, and Zhou, Yan. Assessment of
                Light-Duty Plug-In Electric Vehicles in the United States, 2010-
                2018. United States: N. p., 2019. Web. doi:10.2172/1506474 (citing
                Berman, Kimberly, Jared Dziuba, Colin Hamilton, Richard Carlson,
                Joel Jackson, and Peter Sklar, 2018. ``The Lithium Ion Battery and
                the EV Market: The Science Behind What You Can't See.'' BMO Capital
                Markets, February 2018. https://bmo.bluematrix.com/docs/pdf/079c275e-3540-4826-b143-84741aa3ebf9.pdf); MIT Energy Initiative.
                2019. Insights into Future Mobility, at 77. Cambridge, MA: MIT
                Energy Initiative. http://energy.mit.edu/insightsintofuturemobility.
                 \1198\ Schipper, Florian, Evan M. Erickson, Christoph Erk, Ji-
                Yong Shin, Frederick Francois Chesneau, and Doron Aurbach. 2017.
                ``Review--Recent Advances and Remaining Challenges for Lithium Ion
                Battery Cathodes I. Nickel-Rich, LiNixCoyMnzO2.'' Journal of the
                Electrochemical Society 164, no. 1 (1): A6220-A6228. https://doi.org/10.1149/2.0351701jes.
                ---------------------------------------------------------------------------
                 For the final rule analysis, based on these considerations, the
                agencies in consult with Argonne determined that it was reasonable to
                model HEV, PHEV, and BEV batteries using NMC622 as the cathode active
                material, as shown in Table VI-91 below.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.220
                 The agencies recognize that there will be advancements in battery
                chemistries during the rulemaking timeframe. As discussed further in
                Section (3), below, the analysis accounts for the potential that
                battery costs will decrease, but in a technology-agnostic manner. The
                agencies used BatPaC to model battery costs for the analysis by
                modeling battery prices in a specific year--in this case, MY 2020--and
                then used learning curves to reduce the cost of batteries over time.
                The learning curves act as a proxy for potential future improvements in
                battery chemistry and other battery-related advancements that would
                reduce costs. Using the learning curves in this way makes it
                unnecessary to make inherently uncertain projections of potential
                future improvements in battery chemistry over time.
                 BatPaC version 4.0, which contains NMC811 as a chemistry option,
                was released after the analysis for this rule was completed. However,
                the cost estimates generated in BatPaC version 3.1 using NMC622, with
                discussed learning curves applied resulted in estimated $/kWh battery
                pack costs, during the rule making time frame within a reasonable range
                of other estimated projections that considered NMC811 as the
                predominant battery chemistry. As discussed further in Section (3), a
                significant shift in battery chemistry alone is only one factor
                required to significantly lower battery costs; other developments like
                increases in battery pack production quantities and cell yield (plant
                efficiencies) would be required to reach the commonly-cited $100/kWh
                target.
                 The agencies recognize that the specific chemistries manufacturers
                may choose for future model years may or may not be the same as the
                chemistries selected by the agencies for the analysis. However, this
                approach mirrors the approach taken to modeling technology
                effectiveness and cost used across the analysis; the modeled technology
                effectiveness and cost represents a level of performance representative
                of the typical range of performance across industry. If the agencies
                modeled pre-production battery chemistries unlikely to be widely
                adopted by the industry for several years, the analysis would likely
                under-predict the actual cost and effectiveness of electrification
                technology application. Accordingly, the agencies determined that using
                LFP-G as the modeled chemistry of choice for mild hybrids and NMC622 as
                the modeled chemistry of choice for strong HEVs, PHEVs, and BEVs was
                reasonable.
                 The agencies also refined other inputs and assumptions used for
                modeling battery costs in BatPaC, based on a review of public comments
                and subsequent review of market research, technical publications, and
                other information.
                 Argonne continuously studies the battery pack designs of existing
                electrified vehicles in the market, using, among other information,
                detailed battery pack teardown analysis reports spanning a range of
                electrified vehicle types and vehicle classes produced over a range of
                MYs. For the final rule, Argonne utilized detailed battery pack
                teardown analysis reports for 10 MY
                [[Page 24500]]
                2013 to MY 2018 vehicles from A2mac1,\1199\ as shown in the Table VI-92
                below.
                ---------------------------------------------------------------------------
                 \1199\ Argonne Vehicle Modeling for Safer Affordable Fuel
                Efficient (SAFE) Vehicles Final Rulemaking, Section 5.9 Battery
                Performance and Cost Model (BatPaC), referencing A2Mac1 Automotive
                Benchmarking, https://a2mac1.com.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.221
                 The teardown analysis reports were used to evaluate different
                battery pack design criteria, including battery pack power, battery
                pack energy, battery pack configuration, total number of cells per
                module, number of modules per pack, battery pack mass, energy density
                (cell/pack), cell voltage, battery pack voltage, cathode chemistry,
                cell capacity, and pack capacity. The metrics data collected from
                teardown analysis were used to estimate the battery pack manufacturing
                cost and mass (energy density-Wh/kg) in BatPaC for these exemplar
                vehicles from the A2Mac1 database. The data collected was also used to
                validate the battery pack design assumptions in BatPaC for the final
                rule. The four metrics that BatPaC provides are: Battery pack
                manufacturing cost, battery pack weight (energy density-Wh/kg), battery
                pack capacity (Ah) and nominal battery pack voltage. Since the A2mac1
                teardown reports do not avail the manufacturing costs of these battery
                packs, the analyses and comparisons were limited to the scope of the
                other three criteria.
                 For the NPRM, Argonne used the U.S. Department of Energy VTO
                targets for battery energy density (Wh/kg) for high energy and power
                density-(W/kg) for high powered batteries.\1200\ As a result of the
                analysis discussed above Argonne updated the method of estimating
                battery pack weight for each battery pack design in the final rule
                analysis. The analysis revealed greater influences on battery pack
                design by usable energy density characteristics then was initially
                assumed for the NPRM. For the final rule analysis BatPaC was used for
                battery pack weight estimates along with manufacturing cost estimates.
                ---------------------------------------------------------------------------
                 \1200\ Modeling the Performance and Cost of Lithium-Ion
                Batteries for Electric-Drive Vehicles, ANL/CSE-19/2.
                ---------------------------------------------------------------------------
                 As discussed further in Section VI.C.3.e)(1)(c) Battery Pack Costs,
                the number of cells per pack influenced total battery pack costs for
                the final rule. As result of the analysis discussed above Argonne
                updated the number of cells in each battery. For the final rule
                analysis battery cell counts increased or decreased for some battery
                pack designs, while battery counts for some designs remained the same.
                Argonne's process for evaluating different design criteria for
                electrified vehicles is detailed further in the Argonne model
                documentation.\1201\
                ---------------------------------------------------------------------------
                 \1201\ A Detailed Vehicle Simulation Process To Support CAFE and
                CO2 Standards for the MY 2021-2026 Final Rule Analysis, Section 5.9
                Battery Performance and Cost Model (BatPaC).
                ---------------------------------------------------------------------------
                 The agencies also updated other BatPaC inputs and assumptions based
                on additional market information or research. For the NPRM, the
                agencies modeled battery packs in BatPaC using the default values
                associated with the baseline manufacturing plant, including an annual
                production rate of 100,000 batteries.\1202\
                ---------------------------------------------------------------------------
                 \1202\ See Nelson, Paul A., Gallagher, Kevin G., Bloom, Ira D.,
                and Dees, Dennis W. Modeling the Performance and Cost of Lithium-Ion
                Batteries for Electric-Drive Vehicles--SECOND EDITION (2012), at 62.
                Available at https://publications.anl.gov/anlpubs/2015/05/75574.pdf.
                ---------------------------------------------------------------------------
                 The estimate for battery pack costs incorporates an assumption of
                the battery pack production volume. Both BatPaC version 3.0, used in
                the NPRM, and BatPaC version 3.1, used in the final rule, include a
                default value assumption of 100,000 battery pack units manufactured per
                year per manufacturing plant as well as the plant efficiency (cell
                yield) of 95 percent. For the final rule, the agencies adjusted the
                production volume assumption used in BatPaC version 3.1 to 25,000
                battery pack units, based on the analysis presented below.
                 As described in the BatPaC model documentation, the BatPaC models
                the differences in pack designs and how they affect the costs of one or
                more steps in the battery production process and the physical plant
                layout.\1203\ For example, increasing the power of the battery packs
                without increasing the number of cells, or cell capacity, results in
                the model increasing the area of the cells and decreasing the electrode
                coating thickness. This results in an increased cost of the coating
                equipment, the floor area occupied by the equipment, and the direct
                labor for the process.1204 1205 The agencies are aware that
                each manufacturer (not brand) has a unique battery pack design that
                differs from other manufacturers. Accordingly, it is likely that each
                manufacturer's BEV models had distinct characteristics, such as unique
                battery packaging space, energy requirements, thermal control systems,
                and safety systems, which cause battery pack designs to vary between
                each manufacturer.
                ---------------------------------------------------------------------------
                 \1203\ Nelson, Paul A., Ahmed, Shabbir, Gallagher, Kevin G., and
                Dees, Dennis W. Modeling the Performance and Cost of Lithium-Ion
                Batteries for Electric-Drive Vehicles, Third Edition (2019), at 100.
                Available at https://publications.anl.gov/anlpubs/2019/03/150624.pdf.
                 \1204\ Kupper et al, The Future of Battery Production for
                Electric Vehicles, Boston Consulting Group, (Sept. 11, 2018),
                https://www.bcg.com/publications/2018/future-battery-production-electric-vehicles.aspx.
                 \1205\ Id.
                ---------------------------------------------------------------------------
                 Thus, the agencies determined that even though one battery
                manufacturer
                [[Page 24501]]
                might manufacture batteries for multiple vehicle manufacturers, the
                default BatPaC assumption of 100,000 battery pack units manufactured
                per plant likely did not account for all of the cost differences in
                pack designs between manufacturers. Therefore, the agencies assumed the
                production volume of each battery pack type was reasonably represented
                by the BEV production volume for each manufacturer. The agencies also
                assumed that battery pack manufacturing plants operated at reasonable
                capacity during that timeframe, which would produce the lowest cost
                assumption.
                 The agencies analyzed BEV sales for MYs 2016-2019, referencing data
                collected by the Department of Energy.\1206\ Table VI-93 shows that
                individual manufacturer U.S. BEV sales are substantially below 100,000
                units per year except for Tesla, beginning in MY 2018 Tesla is a
                vertically integrated battery and BEV manufacturer, which is not the
                model the remainder of the industry has implemented, or intends to,
                based on the agencies current understanding. More specifically, Tesla
                sold more BEVs than all manufacturers combined in MYs 2016, 2018, and
                2019. 2017 was the only year in which all other manufacturers combined
                sold more BEVs than Tesla. Ultimately, in selecting a battery pack
                volume estimates for an industry-wide assessment, the agencies sought
                to accurately account for both the representative production volumes
                and representative practices applicable to the industry. As such, the
                agencies evaluated the average per manufacturer volumes, less the
                outlying and vertically integrated volumes of Tesla (shown in Table VI-
                94). As depicted in Table VI-93 and Table VI-94, the data show that the
                average annual sales of BEVs for individual manufacturers, excluding
                Tesla, is just 5% of the default battery pack production volume in
                BatPaC.
                ---------------------------------------------------------------------------
                 \1206\ Light Duty Electric Drive Vehicles Monthly Sales Updates,
                Argonne National Laboratory Energy Systems Division, https://www.anl.gov/es/light-duty-electric-drive-vehicles-monthly-sales-updates (last visited March 2, 2020); Maps and Data, Alternative
                Fuels Data Center, https://afdc.energy.gov/data/ (last visited March
                2, 2020).
                [GRAPHIC] [TIFF OMITTED] TR30AP20.222
                [[Page 24502]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.223
                 In consideration of this data, when estimating the production
                volume in the final rule analysis, the agencies selected a value of
                25,000 units per year per manufacturer as a reasonable estimate for the
                average industry for MY 2020, which is the base model year for
                estimated battery pack costs using BatPaC version 3.1. As discussed in
                Section VI.C.3.e)(3) Electrification Learning Curves, other model year
                battery pack costs are estimated using cost learning. Using the default
                production volume of 100,000 units per year per manufacturer, the
                agencies would have underestimated the actual cost of battery pack
                production for MY 2020, as the model assumes that production costs
                decrease as production volumes increase. By selecting the value of
                25,000 units per year per manufacturing plant, the battery cost
                estimate from the BatPaC model better aligned with the cost estimate
                published in industry-recognized reports such as the UBS MY 2016 Chevy
                teardown report.1208 1209 1210
                ---------------------------------------------------------------------------
                 \1207\ Note, for the assessment, Nissan and Mitsubishi are
                considered a single manufacturer.
                 \1208\ Proposed Determination TSD at 2-127.
                 \1209\ Based on the battery cell to battery pack ratio of 1.3 to
                1.5, the 2015-2019 cell-level figure of $145 per kWh used in the MY
                2016 Chevy Bolt would translate to approximately $190 to $220 per
                kWh on a pack level.
                 \1210\ Hummel et al., UBS Evidence Lab Electric Car Teardown--
                Disruption Ahead?, UBS (May 18, 2017), https://neo.ubs.com/shared/d1ZTxnvF2k/.
                ---------------------------------------------------------------------------
                 The agencies performed a sensitivity study for production volume
                using BatPaC version 3.1. The cost of the battery pack dropped by 15
                percent on average when the production volume was changed from 25,000
                to 100,000 units per year. The sensitivity analysis showed that
                manufacturing plant volume has a significant impact on battery pack
                costs and therefore it is important to use realistic production volume
                estimates for the battery pack cost analysis.
                 Manufacturing plant efficiency is another parameter important to
                estimate battery pack costs. BatPaC version 3.1 defines manufacturing
                plant efficiency in terms of cell yield, or the number of cells that
                are usable out of the total number of cells that the plant
                produced.\1211\ Since battery pack technology and battery pack
                manufacturing processes are proprietary, the data on plant efficiencies
                are not widely reported. While BatPaC uses a default cell yield (plant
                efficiency) value of 95 percent, Argonne battery experts have used an
                85 percent cell yield value to represent the current production yield
                for internal DOE studies.\1212\ By selecting an 85 percent cell yield
                value for the final rule analysis, the agencies aligned the cell yield
                value assumption with internal DOE studies.
                ---------------------------------------------------------------------------
                 \1211\ Cells might not be usable because of, for example,
                manufacturing defects, among other reasons.
                 \1212\ Argonne National Laboratory, BatPaC Model Software,
                https://www.anl.gov/cse/batpac-model-software (last visited March
                19, 2020). Argonne used an 85% cell yield assumption in its
                Estimated Cost of EV Batteries 2018-19 analysis.
                ---------------------------------------------------------------------------
                 In addition, as discussed in detail above, the final rule analysis
                was performed using BatPaC version 3.1, with NMC622 assumed as the
                battery chemistry for HEVs, PHEVs, and BEVs. Separate from the inputs
                and assumptions discussed here, the Argonne battery experts made a
                number of changes to BatPaC version 3.1, and these are extensively
                documented in the BatPaC manual,\1213\ as well as in Argonne model
                documentation for final rule.
                ---------------------------------------------------------------------------
                 \1213\ Nelson, Paul A., Ahmed, Shabbir, Gallagher, Kevin G., and
                Dees, Dennis W. Modeling the Performance and Cost of Lithium-Ion
                Batteries for Electric-Drive Vehicles, Third Edition (2019),
                available at https://publications.anl.gov/anlpubs/2019/03/150624.pdf.
                ---------------------------------------------------------------------------
                (b) Comments on Information Availability
                 In addition to comments that the agencies' battery pack costs were
                too high, the agencies received comments that the analysis for battery
                pack costs was unclear and not well documented. ICCT stated that the
                agencies largely obscured the BEV cost sources and calculations, which
                made it ``nearly impossible for even very interested researchers to
                understand how all the BatPaC costs translate into BEV costs that can
                be compared with other full-BEV costs in the literature.'' \1214\ ICCT
                stated that to enable meaningful public comments, the sources and cost
                calculations must be made explicit and the agencies must provide an
                additional public comment opportunity.\1215\
                ---------------------------------------------------------------------------
                 \1214\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741.
                 \1215\ Id.
                ---------------------------------------------------------------------------
                 CARB claimed that it could not comment meaningfully on the battery
                modeling for the NPRM analysis without extensive additional
                information.\1216\ As such, CARB submitted a letter to the agencies'
                NPRM docket posing, under FOIA, a number of questions pertaining to
                battery assumptions used for the modeling. This requested information
                concerned what version of BatPaC was used in the NPRM analysis, inputs
                incorporated into the BatPaC model; and information about how battery
                costs were generated for the analysis.
                ---------------------------------------------------------------------------
                 \1216\ California Air Resources Board, NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 Specifically, CARB's initial comments alleged that the agencies had
                not disclosed the exact version of BatPaC used, and had simply claimed
                to use the ``most up-to-date'' version of BatPaC,
                [[Page 24503]]
                and further that the agencies had not disclosed ``the BatPaC modeling
                files that were used, clear statements about what version of the model
                was used, or thorough descriptions of the inputs to those modeling
                runs.'' CARB claimed that without that information, ``there is no way
                to know what assumptions were made for raw material pricing, battery
                cell yields, pack electrical connection topology, battery production
                volume assumptions, or if any additional parameters were modeled, like
                rapid charging capability.'' CARB argued that these pieces were
                critical to understanding whether the BatPaC model was estimating
                proper battery pack cost values.
                 In a subsequent docketed comment submitted as an administrative
                appeal to NHTSA's FOIA response, CARB reasserted that, in fact, the
                ``most recent version'' of BatPaC had not been used, because the FOIA
                response stated clearly that version 3.0 had been used and Argonne had
                updated to version 3.1 in October 2017, which was the last version
                released before the NPRM was published. CARB further argued that NHTSA
                was ``choosing to withhold information about battery pack
                configurations,'' and that the agencies had not posted the BatPaC model
                version and files used for the NPRM to the agencies' dockets,
                inhibiting meaningful comment.
                 The majority of information sought by CARB's comment was already
                published in supporting documents and materials posted to the agencies'
                dockets and online websites for the NPRM. Nevertheless, in an effort to
                answer CARB's specific questions, NHTSA also processed the initial
                comment as a FOIA request and provided a written response directly to
                CARB within the comment period. This response both pointed CARB to the
                locations where the sought material could be located among the
                published NPRM materials, and expressly answered several of CARB's
                questions for clarification, such as identifying the specific version
                of BatPaC utilized in the NPRM analysis. For example, although the
                Argonne model documentation describing the battery modeling for the
                NPRM was included in the docket, the agencies' response directed CARB
                to the precise location in the docket where it could be found.
                 The agencies believe that the NPRM docket contained enough
                information for stakeholders to comment meaningfully. This is apparent
                from the voluminous comments the agencies received regarding the NPRM's
                electrification analysis--including from CARB. For example, as
                discussed above, CARB submitted extensive comments on each element of
                the battery cost modeling that CARB claimed the agencies did not
                adequately explain. As discussed above, CARB stated that the agencies'
                selected battery chemistries represented a step backward from previous
                analysis done for the Draft TAR. CARB noted that regardless of whether
                NMC441 or NMC333 was chosen for PHEVs and BEVs in the NPRM analysis,
                the biggest lithium-ion production companies have indicated that they
                will use NMC811 for BEVs, and therefore neither NMC441 nor NMC333 would
                represent current technology going into BEVs or near-future BEV battery
                technology. CARB stated that NMC811 technology is expected to come to
                market in 2019, which, the agencies note, is far sooner than
                anticipated, even in the agencies' prior analyses. CARB was accordingly
                able to communicate its opinion that NMC881 should have been used to
                model battery chemistries for the NPRM analysis, and that NMC441 or
                NMC333 should not be used.
                 As these comments demonstrate, in addition to the extensive
                comments listed above, the expansive information, data, and
                documentation concerning the Argonne BatPaC modeling analysis for the
                NPRM sufficiently enabled commenters to submit voluminous technical
                analysis regarding the electrification analysis. Moreover, while the
                docketed and published NPRM materials themselves afforded sufficient
                notice on these topics, the agencies even undertook the additional step
                of directly responding to CARB in writing in an attempt to address
                specific questions raised by CARB. This written correspondence both
                directed CARB to specific locations on the rulemaking dockets and
                agencies' websites where information CARB was seeking could be
                accessed, and even directly answered several of CARB's questions
                through narrative responses. Both CARB and other commenters submitted
                subsequent comments, which referenced the material described in this
                written response. Accordingly, the agencies consider the information
                provided with the NPRM sufficient to enable meaningful comment, which
                is underscored by the voluminous technical comments received on the
                electrification issues.
                 For this final rule, the BatPaC model version 3.1 (June 2018) model
                documentation has been included in the docket for this
                rulemaking.\1217\ Furthermore, Argonne's detailed documentation
                describing the modeling process used to support this final rule
                provides information and specific assumptions that Argonne's experts
                used to simulate batteries and their associated costs for the full
                vehicle simulation modeling.\1218\ These resources, in addition to the
                detailed description of the battery cost modeling process provided here
                and in the FRIA provide interested stakeholders the necessary tools to
                understand the battery cost modeling analysis.
                ---------------------------------------------------------------------------
                 \1217\ Nelson, Paul A., Ahmed, Shabbir, Gallagher, Kevin G., and
                Dees, Dennis W. Modeling the Performance and Cost of Lithium-Ion
                Batteries for Electric-Drive Vehicles, Third Edition (ANL/CSE-19/2),
                available at https://publications.anl.gov/anlpubs/2019/03/150624.pdf.
                 \1218\ A Detailed Vehicle Simulation Process To Support CAFE and
                CO2 Standards for the MY 2021-2026 Final Rule Analysis.
                ---------------------------------------------------------------------------
                c) Final Rule Battery Pack Costs
                 As discussed above, based on comments and additional research, the
                agencies updated the battery cost analysis for the final rule by
                relying on BatPaC version 3.1.\1219\ In addition, as outlined above and
                explained in more detail in the Argonne Model Documentation for this
                final rule, several inputs and assumptions were updated based on public
                comments, market research, and additional literature review. The
                agencies computed the average battery pack cost across all road load
                combinations for electrification technologies that could be reasonably
                compared between the NPRM and final rule.\1220\
                ---------------------------------------------------------------------------
                 \1219\ Modeling the Performance and Cost of Lithium-Ion
                Batteries for Electric-Drive Vehicles, Third Edition (ANL/CSE-19/2)
                provides a complete list of changes and assumptions incorporated in
                BatPaC version 3.1.
                 \1220\ Costs data is from the CAFE Model core file
                Battery_Costs.csv.
                ---------------------------------------------------------------------------
                 Table VI-95 to Table VI-99 show the differences between battery
                pack costs presented in the NPRM and final rule.\1221\ The tables show
                absolute cost differences between battery packs, which can vary for
                battery packs with different energy and power combinations. For
                example, as shown in Table VI-96, the cost difference between the NPRM
                and final rule for a Mild HEV battery pack with a 1kWh energy and 10kW
                power rating is -28 percent. Similarly, the cost difference in an HEV
                battery pack with a 1kWh battery energy and 40kW power rating is 5
                percent. In summary, the percentage increase or decrease in the table
                represents the
                [[Page 24504]]
                absolute cost differences between the battery packs used in NPRM and in
                final rule.
                ---------------------------------------------------------------------------
                 \1221\ The absolute cost differences shown here is by comparing
                the cost of battery pack with similar number of cells in the NPRM to
                the final rule cost lookup tables for compact and medium car. The
                cost differences between the NPRM and the final rule cost lookup
                tables for small SUV, medium SUV and Pickup trucks will be different
                from the table shown here.
                ---------------------------------------------------------------------------
                 Figure VI-40 to Figure VI-42 shows the average battery pack costs
                across all road load combinations for each applicable vehicle
                technology class for SHEVPS, PHEV50, and BEV200s between the NPRM and
                final rule.\1222\ Since the battery pack size varies for different road
                load combinations, the battery pack cost across different road load
                combinations varies as well. For example, there are 105 combinations of
                different mass reduction, aerodynamic improvements and rolling
                resistance improvements. The battery pack size for an initial road load
                condition that includes MR0, AERO0 and ROLL0 is larger, and therefore,
                the cost of the battery pack is higher as well. The battery pack size
                is smaller for the highest level of road load reduction such as in MR6,
                AERO20 and ROLL20, and the cost of battery pack is less as well.
                ---------------------------------------------------------------------------
                 \1222\ The agencies did not simulate SHEVPS and BEV200
                powertrain architectures on pickup trucks in the NPRM, so those are
                not included in the comparison.
                ---------------------------------------------------------------------------
                 Table VI-95 shows the cost difference in Micro HEV battery packs.
                The cost reduction is from the reduced number of cells in the battery
                pack.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.224
                 Table VI-96 shows percentage cost differences for mild hybrid
                (BISG) battery packs. The cost difference is due, in part, to
                accounting for BISG-related hardware costs, such as the battery
                management system, as part of the electric machine costs in this final
                rule.\1223\
                ---------------------------------------------------------------------------
                 \1223\ In the NPRM, additional hardware component costs were
                included as part of the battery pack cost.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.225
                 Table VI-97 shows the percentage cost differences for HEV battery
                packs. Even as the battery chemistry changed to NMC622, the cost
                increase is from the different battery pack production volume and plant
                efficiency assumptions used in the final rule.
                [[Page 24505]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.226
                 Figure VI-40 shows the difference in battery pack costs for SHEVPS
                applications between the NPRM and final rule. Power-split hybrids could
                not be used in pickup trucks due to their unique power and towing
                requirements, so those technology classes are not shown. In general,
                the cost of the battery pack in the final rule analysis increased due
                to the updated battery pack production volume and plant efficiency
                assumptions.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.227
                 Table VI-98 shows the percentage cost differences between the NPRM
                and final rule for PHEV50 battery packs. The cost increase in the
                PHEV50 battery pack shown here is mainly due to the increase in number
                of cells per pack as well as the other updated BatPaC assumptions.
                [[Page 24506]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.228
                 Table VI-94 shows the difference in average PHEV50 battery pack
                costs between the NPRM and final rule for all technology combinations.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.229
                 Table VI-99 shows the percentage cost differences for BEV battery
                packs. In the example shown in Table VI-99, the agencies compared the
                cost lookup table from the NPRM with 300 cells to the cost lookup table
                in the final rule analysis with 320 cells. The cost increase in the
                higher energy packs is due to the different battery pack production
                volume and plant efficiency value assumptions, along with the different
                battery chemistry assumption.
                [[Page 24507]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.230
                BILLING CODE 4910-59-C
                 Figure VI-42 shows the average cost of BEV200 battery packs across
                all technology combinations for technology classes that could be
                compared between the NPRM and final rule. As shown, for the final rule
                analysis, the average cost of a BEV200 battery pack is lower than the
                average cost of the NPRM BEV200 battery pack. For the final rule
                analysis, the agencies updated the motor efficiency map for BEVs (as
                explained in Section VI.C.3.d) Electrification Technology
                Effectiveness) and updated the glider share of the vehicles from 50
                percent of the curb weight to 71 percent of the vehicle curb weight (as
                explained in Section VI.C.4 Mass Reduction). In addition, the updated
                motor weight resulted in further reduced vehicle weights. This
                combination of improved vehicle assumptions resulted in reduced energy
                and power requirements in BEVs.
                 The agencies also observed that even as the number of cells in the
                battery pack increased from 300 to 320, and changes in production
                volume and plant efficiency values resulted in marginal cost increases
                for higher energy packs, the overall battery capacity requirement went
                down due to overall reduction in power and energy demand from electric
                vehicles.\1224\ A reduction in battery capacity leads to reduced cell
                size in a pack with number of cells and voltage. A reduction in cell
                size leads to cost reductions at the cell level and at the pack level.
                In general, a higher capacity battery pack is more expensive than a
                lower capacity battery pack due to the increase in cell size for a
                given number of cells and voltage.1225 1226
                ---------------------------------------------------------------------------
                 \1224\ As explained above, the energy density values in the NPRM
                were kept constant. For the final rule analysis, the power density
                varied to meet different power and energy requirements, as was
                observed through market research.
                 \1225\ Nelson, Paul A., Ahmed, Shabbir, Gallagher, Kevin G., and
                Dees, Dennis W. Modeling the Performance and Cost of Lithium-Ion
                Batteries for Electric-Drive Vehicles, Third Edition (ANL/CSE-19/2),
                at 15 (battery design worksheet). Available at https://publications.anl.gov/anlpubs/2019/03/150624.pdf.
                 \1226\ The amount of electrode materials and electrode area of
                the cells are determining cost factors in the battery. Higher
                capacity battery packs require additional manufacturing steps to
                increase the energy density of the pack.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24508]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.231
                 The graphs demonstrate the range of cost changes observed, with the
                other electrification technologies falling somewhere in between the
                extremes. In summary, the agencies observed that the BEV200 technology
                showed a cost reduction in battery packs across all vehicle platforms
                with the largest reductions occurring for the largest battery packs. In
                contrast the PHEV50 technology showed a cost increase in battery packs
                across all vehicle platforms with the smallest increase for the largest
                battery packs and the largest increase for the smallest battery packs.
                It is worth noting the cost decreases seen across the technologies are
                generally larger than the cost increases.
                 For the final rule, when possible, the calculated battery pack
                weight and manufacturing cost was also compared with the battery pack
                cost and weight data obtained through various benchmarking studies. For
                example, UBS reported a battery pack manufacturing cost of $12,500 from
                its 2017 Chevrolet Bolt teardown analysis.\1227\ Using a production
                volume of 25,000 packs per year per plant and similar battery pack
                design, BatPaC estimated a manufacturing cost of $10,680.\1228\ These
                comparisons were used to verify the different assumptions used in
                BatPaC and helps represent the battery packs for electrified vehicles
                used in representative market volume. Table VI-100 shows a comparison
                of specifications estimates for 60 kWh and 160 kW battery packs from
                the 2016 DOE VTO report 1229 1230 and BatPaC version 3.1
                (June 2018), and the Chevrolet Bolt. The comparison shows modeled and
                actual battery packs are in close agreement.
                ---------------------------------------------------------------------------
                 \1227\ Hummel et al., UBS Evidence Lab Electric Car Teardown--
                Disruption Ahead?, UBS (May 18, 2017), https://neo.ubs.com/shared/d1ZTxnvF2k/.
                 \1228\ $178/kWh x 60kWh = $10,680.
                 \1229\ Peter Faguy, Overview of the DOE Advanced Battery R&D
                Program (June 2015), https://www.energy.gov/sites/prod/files/2015/06/f23/es000_faguy_2015_o.pdf.
                 \1230\ Freyermuth, Vincent. Rousseau, Aymeric. ``Impact of
                Vehicle Technologies Office Targets on Battery Requirements.'' ANL/
                ESD-16/22. Energy Systems Division, Argonne National Laboratory
                (2016).
                ---------------------------------------------------------------------------
                [[Page 24509]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.232
                 In addition, the agencies compared the battery pack cost estimates
                generated using BatPaC to other current studies or studies cited by
                commenters. Table VI-101 summarizes battery pack estimates from
                selected studies in MYs for which that information was available.
                ---------------------------------------------------------------------------
                 \1231\ Not each study distinguished a DMC source year, so these
                values vary slightly based on inflation.
                 \1232\ Sources generally provided estimates for 2018 or 2020.
                 \1233\ Hummel et al., UBS Evidence Lab Electric Car Teardown--
                Disruption Ahead?, UBS (May 18, 2017), https://neo.ubs.com/shared/d1ZTxnvF2k/.
                 \1234\ Mosquet et al., The Electric Car Tipping Point, BCG (Jan.
                11, 2018), https://www.bcg.com/publications/2018/electric-car-tipping-point.aspx. This study provided cell cost estimates that the
                agencies converted to pack cost estimates using a multiplier of 1.3,
                as outlined in the Draft TAR at 5-124.
                 \1235\ Nic Lutsey and Michael Nicholas, Update on electric
                vehicle costs in the United States through 2030, ICCT (April 2,
                2019), available at https://theicct.org/publications/update-US-2030-electric-vehicle-cost. The presented values are $/kWh pack costs for
                mid-range electric cars/crossovers and SUVs.
                 \1236\ McKerracher et al., Electric Vehicle Outlook 2019--Free
                Interactive Report, Bloomberg New Energy Finance (May 2019), https://about.bnef.com/electric-vehicle-outlook/.
                 \1237\ Logan Goldie-Scot, A Behind the Scenes Take on Lithium-
                ion Battery Prices, Bloomberg New Energy Finance (March 5, 2019),
                https://about.bnef.com/blog/behind-scenes-take-lithium-ion-battery-prices/. BNEF projected the pack costs in 2018$ for 2018 as $176,
                and used the same value in the Electric Vehicle Outlook 2019 to
                describe pack cost levels ``today.''
                 \1238\ MIT Energy Initiative. 2019. Insights into Future
                Mobility. Cambridge, MA: MIT Energy Initiative. Available at http://energy.mit.edu/insightsintofuturemobility.
                 \1239\ Islam, E., Kim, N., Moawad, A., Rousseau, A., ``A Large-
                Scale Vehicle Simulation Study To Quantify Benefits & Analysis of
                U.S. Department of Energy VTO & FCTO R&D Goals.'' Report to U.S.
                Department of Energy. Contract ANL/ESD-19/10 (forthcoming).
                ---------------------------------------------------------------------------
                [[Page 24510]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.233
                 As shown in the table above, there are a range of cost estimates
                for battery packs. Each individual cost estimate is derived based on
                certain set of assumptions to arrive at a rate of cost reduction. Among
                all the different cost estimates, Bloomberg New Energy Finance (BNEF)
                has the most aggressive year-over-year cost reductions, based on the
                historical learning rate of 18% and their battery demand
                forecast.\1240\ Similar to other sources of cost estimates BNEF assumes
                improved battery chemistry and battery density increasing greater than
                200Wh/kg by 2030. In order for the battery manufacturer to achieve
                economies of scale, BNEF assumes a global battery manufacturing
                facility capable of producing battery packs for both stationary energy
                storage and vehicle applications.
                ---------------------------------------------------------------------------
                 \1240\ Logan Goldie-Scot, A Behind the Scenes Take on Lithium-
                ion Battery Prices, Bloomberg New Energy Finance (March 5, 2019),
                https://about.bnef.com/blog/behind-scenes-take-lithium-ion-battery-prices/.
                ---------------------------------------------------------------------------
                 A recent report from the Massachusetts Institute of Technology
                (MIT), the MIT Energy Initiative's Insights into Future Mobility, has
                the most conservative estimate among all the cost sources listed the
                Table VI-101. The authors use a more rigorous two-stage method of
                estimating composite battery learning curves independently for (a)
                battery material synthesis and minerals costs, and (b) battery pack
                production processes. The learning rates are defined as the cost
                reduction that results from cumulative volume doubling, and produce
                separate cost learning rates for the two stages of 3.5 percent and 16.5
                percent, respectively. The study argues that there are greater
                opportunities for cost learning in the production stage than the
                chemical synthesis stage, which is more mature. These cost estimates
                produce global EV fleet penetration rates that may not be as aggressive
                as other estimates, reaching only 33 percent by 2050. This study also
                assumes NMC811 will be available by 2030.
                 The cost estimates from other sources referenced above also include
                assumptions about higher levels of battery pack production and higher
                density battery cells. Most cost estimates assume improved battery
                chemistry, such as NMC811. As discussed above, the agencies determined
                that modeling assuming NMC622 was reasonable, based on current
                production vehicles, the relative uncertainty surrounding large-scale
                NMC811 deployment in the rulemaking timeframe, and the ability to
                account for lower battery pack costs over time with cost learning. The
                agencies also believe that, based on the market analysis and from the
                teardown analysis, improvements in battery chemistry may be slow to be
                applied in a widespread manner, and therefore the economies of scale
                required to achieve considerable cost reductions solely from
                improvements in chemistry may remain effusive during the rulemaking
                timeframe.
                 For these reasons, the agencies believe that the BatPaC-generated
                battery cost estimates using the updated inputs and assumptions are
                reasonable.
                2) Non-Battery Electrification Component Costs
                 Battery components are the biggest driver of the cost of
                electrification, however, non-battery electrification components also
                add to the total cost required to electrify a vehicle. In this
                analysis, the agencies accounted for the following non-battery
                component costs: Electric motor(s), inverter, and other power
                electronics including a bi-directional DC/DC converter, a voltage step
                down DC/DC converter, and an on-board charger. Collectively, these
                components (except for the on-board charger) are referred to as the
                electric traction drive systems (ETDS), or the electric machine. Non-
                plug-in hybrid electric vehicles include all of the listed components
                except for an on-board charger; PHEVs include all of the listed
                components; and BEVs include all of the listed components except, in
                some cases, a second motor.
                 For the NPRM, the agencies accounted for battery pack costs and
                ETDS costs independently.\1241\ The Alliance commented broadly in
                support of separating electrification hardware costs and battery costs,
                and stated that it was a positive change to the modeling.\1242\ The
                Alliance correctly noted that the separation allowed for separate
                learning rates and cost differentiation between the two distinct pieces
                of electrification technologies.
                ---------------------------------------------------------------------------
                 \1241\ PRIA at 362.
                 \1242\ Alliance of Automobile Manufacturers, NHTSA-2018-0067-
                12073, at 140.
                ---------------------------------------------------------------------------
                [[Page 24511]]
                 As stated in the PRIA,\1243\ the agencies derived the cost values
                for the EDTS using Argonne National Laboratory's ``Assessment of
                Vehicle Sizing, Energy Consumption, and Cost through Large-Scale
                Simulation of Advanced Vehicle Technologies'' report.\1244\ Generally,
                the agencies referred to this report in the PRIA as the DOE VTO report,
                as it was a report that reviewed results of the DOE VTO. Some
                commenters seemed confused by this alternative reference--even
                questioning why the agencies didn't rely on recent Argonne National
                Laboratory reports.\1245\ To clarify, this report was written by
                Argonne National Laboratory, and to avoid further confusion it is
                referred to using the full title throughout this rule.
                ---------------------------------------------------------------------------
                 \1243\ 83 FR 43047; PRIA at 362.
                 \1244\ Moawad, Ayman, Kim, Namdoo, Shidore, Neeraj, and
                Rousseau, Aymeric. Assessment of Vehicle Sizing, Energy Consumption
                and Cost Through Large Scale Simulation of Advanced Vehicle
                Technologies (ANL/ESD-15/28). United States (2016), available at
                https://www.autonomie.net/pdfs/Report%20ANL%20ESD-1528%20-%20Assessment%20of%20Vehicle%20Sizing,%20Energy%20Consumption%20and%20Cost%20through%20Large%20Scale%20Simulation%20of%20Advanced%20Vehicle%20Technologies%20-%201603.pdf.
                 \1245\ California Air Resources Board, NHTSA-2018-0067-11973, at
                130-31.
                ---------------------------------------------------------------------------
                 CARB expressed concerns with non-battery component effectiveness
                values, arguing that the agencies inappropriately relied on outdated
                data for electric machines and inverter efficiencies across all
                electrification applications, and further claiming that the agencies
                did not project any efficiency gains in those components over
                time.\1246\ Broadly, as these comments on effectiveness related to the
                NPRM non-battery component cost estimates, CARB claimed that the
                agencies failed to consider new data, including the 2015 ORNL Annual
                Progress Report for the Power Electronics and Electric Motors Program,
                and two Argonne studies, which rendered the analysis unrepresentative
                of actual technology costs.
                ---------------------------------------------------------------------------
                 \1246\ California Air Resources Board, NHTSA-2018-0067-11973, at
                130.
                ---------------------------------------------------------------------------
                 CARB also commented that the agencies did not provide any
                substantive discussion or documentation of how non-battery component
                costs were developed for the NPRM analysis. CARB claimed that
                dissonance existed between the PRIA description of voltage systems and
                associated costs needed for different performance classes, the
                Autonomie files, and the technologies input file, and that this served
                as an example of how the agencies failed to include information
                regarding how costs and cost differences were derived, or any component
                changes from previous analyses.
                 CARB also commented that the lack of disclosure of non-battery cost
                development information was an issue for other electrification
                technologies. CARB cited the increase in parallel (P2) and power-split
                (PS) hybrid systems costs relative to costs used in past agency
                analyses, noting that there was no discussion on what changed from the
                past analyses. CARB referenced a 2010 FEV teardown (Light Duty
                Technology Cost Analysis, Power-Split and P2 HEV Case Studies, EPA-420-
                R-11-015) study that the agencies had previously relied on for
                component costs, noting that not only did the agencies ignore that
                study in the NPRM, but that ICCT had commented 2010 FEV report
                overstated strong hybrid costs at the time of the study, making it
                likely that costs are likely to be lower now and even more so in the
                future. CARB claimed that the agencies provided no justification or
                rationale for the increases in strong hybrid modeled costs for the
                proposal, and that there was no meaningful way to comment on the exact
                components or cost changes that the agencies relied upon. Similarly,
                CARB cited EPA's 2016 Proposed Determination and associated public
                comments from Ford and Tesla on the Draft TAR for the proposition that
                non-battery costs, which were lower in the Draft TAR than the NPRM,
                were conservative and not overly optimistic.
                 Finally, in addition to the ORNL and Autonomie group studies that
                CARB referenced as examples of sources that provided updated data on
                non-battery component effectiveness and costs, CARB claimed that newer
                data existed from a UBS Global Research report that examined the
                component costs of a MY 2016 Chevrolet Bolt, and the agencies did not
                discuss why the newer data was not used in the NPRM analysis. CARB
                stated the significant upward adjustment in non-battery costs from
                previous analyses was not supported by industry input, analysis
                conducted by other outside sources, or by the agencies' previous
                analyses.
                 As explained above, for the NPRM the agencies relied on Argonne's
                ``Assessment of Vehicle Sizing, Energy Consumption, and Cost through
                Large-Scale Simulation of Advanced Vehicle Technologies'' for EDTS
                costs. In turn, the Assessment of Vehicle Sizing, Energy Consumption,
                and Cost through Large-Scale Simulation of Advanced Vehicle
                Technologies report referenced electric machine data provided by OEMs,
                suppliers, and Oak Ridge National Laboratory.\1247\ Regarding CARB's
                assertion that the agencies did not refer to the UBS Global Research
                report on the MY 2016 Chevy Bolt teardown for the NPRM, the agencies
                agree. The UBS Global Research report was not available at the time the
                CAFE model inputs were finalized for the NPRM analysis. That study,
                among others, was considered for the final rule.
                ---------------------------------------------------------------------------
                 \1247\ Moawad, Ayman, Kim, Namdoo, Shidore, Neeraj, and
                Rousseau, Aymeric. Assessment of Vehicle Sizing, Energy Consumption
                and Cost Through Large Scale Simulation of Advanced Vehicle
                Technologies (ANL/ESD-15/28), at 32.
                ---------------------------------------------------------------------------
                 For the final rule analysis, the agencies carefully considered
                comments and the referenced studies, as well as other studies. The
                agencies determined the cost and component efficiency estimates from
                U.S. DRIVE's October 2017 report, Electrical and Electronics Technical
                Team (EETT) Roadmap,\1248\ provided reasonable estimates to use in the
                final rule. The EETT Roadmap report reflected considerable work by the
                DOE VTO collaboratively with U.S. DRIVE, a government-industry
                partnership. The EETT Roadmap report estimated the 2017 manufacturing
                cost of a commercial on-road 100kW ETDS consisting of a single electric
                traction motor and inverter. The reported costs were approximately
                $1,800, with the cost of the electric motor accounting for $800, and
                approximately $1,000 for the inverter, equaling $18/kW for the ETDS.
                ---------------------------------------------------------------------------
                 \1248\ U.S. DRIVE, Electrical and Electronics Technical Team
                Roadmap (Oct. 2017), available at https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
                ---------------------------------------------------------------------------
                 The agencies also referenced the UBS MY 2016 Chevy Bolt teardown
                report to compare the cost of the ETDS.\1249\ To compare the costs, the
                agencies applied the $18/kW metric for ETDS as determined by EETT
                Roadmap report to the 150kW ETDS used in the MY 2016 Chevy Bolt ($18kW
                x 150kW = $2700). As shown in Table VI-102, the cost estimate from the
                above computation aligned with UBS MY 2016 Chevy Bolt teardown cost
                estimate. As a result, the agencies determined that it was appropriate
                to use $18/kW to estimate the cost of the ETDS for all hybrid and
                electric vehicle architectures for the final rule.
                ---------------------------------------------------------------------------
                 \1249\ Hummel et al., UBS Evidence Lab Electric Car Teardown--
                Disruption Ahead?, UBS (May 18, 2017), https://neo.ubs.com/shared/d1ZTxnvF2k/.
                ---------------------------------------------------------------------------
                 The EETT Roadmap report did not explicitly estimate the cost of
                other electrical equipment present in PHEVs and BEVs, such as on-board
                chargers, DC to DC converters, and charging cables, but recommended
                cost targets for the years 2020 and 2025. As a consequence, the
                agencies relied on the
                [[Page 24512]]
                UBS MY 2016 Chevy Bolt teardown report to estimate the cost of on-board
                chargers, DC to DC converters, and charging cables. Table VI-102 shows
                the cost estimate for the ETDS from the EETT Roadmap report and from
                the UBS MY 2016 Chevy Bolt teardown report, and the cost estimate for
                other electrical equipment from the same UBS report.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.234
                [GRAPHIC] [TIFF OMITTED] TR30AP20.235
                 While the EETT Roadmap report estimated the cost of the ETDS at the
                system level, the report did not itemize the cost of individual
                components in electric motor and inverter in 2017. However, the EETT
                Roadmap report provided target cost estimates for the motor and
                inverter system for the year 2025. As shown in Table VI-104, the EETT
                Roadmap report estimated a cost reduction of 73 percent for the
                inverter and 59 percent for the motor relative to 2017. Using the
                percentage cost reductions from 2025 to the on-road status as defined
                in the EETT Roadmap report, the agencies developed an estimated motor
                and inverter component cost for 2017. The resulting cost estimate for
                2017 using the scaling factor matches the $18/kW for motor and inverter
                ($10/kW for Inverter + $8/kW for motor). Since the motor and inverter
                component costs are developed based on a $/kW basis, the agencies
                applied the same $/kW metric for all hybrid and electric vehicle
                applications for the final rule analysis.
                ---------------------------------------------------------------------------
                 \1250\ U.S. DRIVE, Electrical and Electronics Technical Team
                Roadmap, at 12 (Oct. 2017), available at https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
                 \1251\ U.S. DRIVE, Electrical and Electronics Technical Team
                Roadmap, at 12 (Oct. 2017), available at https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
                 \1252\ T U.S. DRIVE, Electrical and Electronics Technical Team
                Roadmap, at 12 (Oct. 2017), available at https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
                 \1253\ U.S. DRIVE, Electrical and Electronics Technical Team
                Roadmap, at 18 (Oct. 2017), available at https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
                ---------------------------------------------------------------------------
                [[Page 24513]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.236
                [GRAPHIC] [TIFF OMITTED] TR30AP20.237
                [GRAPHIC] [TIFF OMITTED] TR30AP20.238
                BILLING CODE 4910-59-C
                 In addition, the EETT Roadmap report provided notably newer data
                than the 2010 FEV teardown study referenced by commenters. Based on
                these considerations, the agencies determined that the EETT Roadmap
                report provided reasonable costs to estimate the cost of EDTS
                components in the rulemaking timeframe.
                ---------------------------------------------------------------------------
                 \1254\ U.S. DRIVE, Electrical and Electronics Technical Team
                Roadmap, at 23 (Oct. 2017), available at https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
                ---------------------------------------------------------------------------
                (3) Electrification Learning Curves
                 The total incremental costs of electrification powertrain
                technologies are comprised of the DMC as modified by the learning
                curves for each individual powertrain component, which include
                batteries, non-battery components, and IC engines and transmissions
                (for hybrids and PHEVs). The PRIA showed the learning curves for
                battery and non-battery electrification technologies,\1255\ and listed
                the sources used to develop those curves, including the 2015 NAS
                report,
                [[Page 24514]]
                Wright-based learning curves,\1256\ and Argonne's 2016 Assessment of
                Vehicle Sizing, Energy Consumption, and Cost through Large-Scale
                Simulation of Advanced Vehicle Technologies.\1257\ Learning rates for
                batteries were also derived using Argonne's BatPaC model.
                ---------------------------------------------------------------------------
                 \1255\ PRIA at 380.
                 \1256\ Wright, T. P. (1936). Factors Affecting the Cost of
                Airplanes. Journal of Aeronautical Sciences, vol. 3 124-125. http://www.uvm.edu/pdodds/research/papers/others/1936/wright1936a.pdf.
                 \1257\ Moawad, Ayman, Kim, Namdoo, Shidore, Neeraj, and
                Rousseau, Aymeric. Assessment of Vehicle Sizing, Energy Consumption
                and Cost Through Large Scale Simulation of Advanced Vehicle
                Technologies (ANL/ESD-15/28). United States (2016). Available at
                https://www.autonomie.net/pdfs/Report%20ANL%20ESD-1528%20-%20Assessment%20of%20Vehicle%20Sizing,%20Energy%20Consumption%20and%20Cost%20through%20Large%20Scale%20Simulation%20of%20Advanced%20Vehicle%20Technologies%20-%201603.pdf.
                ---------------------------------------------------------------------------
                 For the NPRM, to develop the learning curves for non-battery
                components, the agencies consulted Argonne's 2016 Assessment of Vehicle
                Sizing, Energy Consumption, and Cost through Large-Scale Simulation of
                Advanced Vehicle Technologies report. The report provided estimated
                cost projections from the 2010 lab year to the 2045 lab year for
                individual vehicle components.1258 1259 The agencies
                considered the component costs used in electrified vehicles, and
                determined the learning curve by evaluating the year over year cost
                change for those components.
                ---------------------------------------------------------------------------
                 \1258\ ANL/ESD-15/28 at 116.
                 \1259\ DOE's lab year equates to five years after a model year,
                e.g., DOE's 2010 lab year equates to MY 2015.
                ---------------------------------------------------------------------------
                 The agencies used BatPaC version 3.0 to develop the NPRM learning
                curves for batteries. As discussed above, BatPaC calculations are based
                on generic pack design for a given set of inputs that could reasonably
                represent potential current and future designs. Because BatPaC does not
                simulate battery costs as a function of time, the agencies modified the
                battery volume inputs for MY 2015, MY 2020, MY 2025 to show costs in
                each of those MYs. Like the non-battery component analysis, a learning
                curve was developed from the year over year cost change, and this rate
                was used to develop the learning curves used in the NPRM.
                 CARB stated that publicly available data supported lower costs in
                the near term than what the applied learning curve rates would do to
                the battery costs developed by the agencies, and the agencies failed to
                consider new information or data to adjust battery costs.\1260\ CARB
                stated that considering the substantial volume of publicly available
                information and public input to the agencies' previous analysis,
                projected battery costs should have been adjusted even further downward
                for the NPRM. CARB stated that instead, the agencies moved costs upward
                without sufficient justification, and in contrast, the analysis for the
                Proposed Determination and 2016 Draft TAR provided far more
                justification for those modeled battery costs.
                ---------------------------------------------------------------------------
                 \1260\ California Air Resources Board, NHTSA-2018-0067-11873, at
                142-43.
                ---------------------------------------------------------------------------
                 As discussed in Section VI.B.4.d) Cost Learning, above, ICCT
                commented broadly on the change in approach to learning curves since
                the Draft TAR, stating that this change in approach led to lower
                decreases in costs over time in the NPRM than the Draft TAR analysis.
                ICCT compared EPA's Draft TAR learning curves and NPRM learning curves
                for batteries in MYs 2016-2025, concluding that there was a 29%
                reduction in learning for batteries from EPA's Draft TAR analysis to
                the NPRM analysis.
                 The agencies considered an array of both present and future cost
                estimates from various public and private sector organizations to
                validate the rate at which battery pack costs declined over time. These
                estimates, in addition to estimates submitted by commenters as
                discussed in BatPaC Inputs and Assumptions and Final Rule Battery Pack
                Costs are shown in Table VI-101. In addition, the agencies had to
                consider how to project learning rates out through 2050, as discussed
                in Section VI.B.4.d) Cost Learning and Section VI.C.3.e)(3)
                Electrification Learning Curves.
                 The agencies also assessed and reviewed literature evaluating more
                recent battery technology development.1261 1262 The NPRM
                analysis used a three percent learning rate per year from MY 2033 to MY
                2050. Learning rate forecasts from MY 2033 to MY 2050 for this final
                rule analysis were scaled down in steps from the previous analysis
                based on literature, market research, and Wright's learning curve
                assumptions.
                ---------------------------------------------------------------------------
                 \1261\ MIT Energy Initiative. 2019. Insights into Future
                Mobility. Cambridge, MA: MIT Energy Initiative. Available at http://energy.mit.edu/insightsintofuturemobility.
                 \1262\ Islam, E., Kim, N., Moawad, A., Rousseau, A., ``A Large-
                Scale Vehicle Simulation Study To Quantify Benefits & Analysis of
                U.S. Department of Energy VTO & FCTO R&D Goals.'' Report to U.S.
                Department of Energy. Contract ANL/ESD-19/10. (forthcoming).
                ---------------------------------------------------------------------------
                 It is difficult to predict which battery chemistry and production
                processes will be prevalent for electrified vehicles in MY 2030, let
                alone for MY 2050. The agencies reviewed potential battery chemistries
                that could come into readiness for adoption at different timeframes,
                such as MY 2030s to MY 2039, and MY 2040 to MY 2050.\1263\ It is
                possible that costs based on other lithium-ion based chemistries will
                learn at the same rate as lithium-ion NMC development. However, the
                same learning effect in battery production may not be additive across
                different chemistries, especially in learning effects related to
                battery production. Accordingly, the learning rates applied between MY
                2030 to MY 2039 considered development and increased volume for the
                same or similar battery chemistries as an NMC battery platform.\1264\
                Learning curves beyond MY 2040 were flattened further to ensure that
                the cost of batteries did not lower beyond the projected price of the
                raw materials. Further, new chemistries introduced in later years may
                learn at different rates than the curve identified for NMC-based
                chemistries. The battery pack cost learning rate that resulted from
                this exercise produced the schedule that appears in Table VI-96, which
                shows this final rule analysis battery pack cost reduction as function
                of time. By MY 2040, the pack cost has reduced by 54 percent.
                Accordingly, the estimated battery pack cost between MY 2040 and MY
                2050 as shown in Figure VI-43 below shows flatter curve.
                ---------------------------------------------------------------------------
                 \1263\ MIT Energy Initiative. 2019. Insights into Future
                Mobility. Cambridge, MA: MIT Energy Initiative, at p. 79. Available
                at http://energy.mit.edu/insightsintofuturemobility.
                 \1264\ For example, an NMC lithium-ion-based platform could move
                from a cathode composition of NMC622 to NMC811.
                ---------------------------------------------------------------------------
                [[Page 24515]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.239
                 The reference cost is defined for MY 2020 vehicles, and vehicles
                produced in subsequent years (as well as earlier years) use a per kWh
                cost that is a percentage of the 2020 cost. As the figure shows, the
                cost reduction is rapid through MY 2030, after which cost reductions
                slow considerably. As discussed above, the cost projections assumed
                different battery chemistries and different rates of cost learning.
                 The agencies expect there will be incremental improvements in
                battery chemistry, energy density, plant efficiency, and production
                volume over the timeframe modeled in the analysis. While each of these
                factors may have an impact on the rate at which battery costs decline
                over time, the agencies determined that using the same cost learning
                projection method from the NPRM to project learning rates out through
                2050 provided a reasonable method for accounting for something that is
                inherently uncertain. Accordingly, the learning curve used in the NPRM
                and in the final rule represent a composite learning curve irrespective
                of the type of battery chemistry, the production volume necessary to
                achieve economies of scale, or energy density of the battery pack. For
                the final rule, the agencies have performed sensitivity analyses
                varying the battery pack learning rate, and these analyses are
                presented in FRIA Chapter VII.E Sensitivity cases.
                (4) Electrified Powertrain Costs
                 For the NPRM analysis and carried forward for the final rule
                analysis, the total electrified powertrain costs were developed by
                summing individual component costs. The costs associated with the IC
                engine, transmissions, electric machines, and battery packs were
                combined to create a full-system cost, per Section VI.C.3.e)(2) Non-
                battery Electrification Component Costs, Section VI.C.3.e)(1) Battery
                Pack Modeling, Section VI.C.1.g) Engine Costs, and VI.C.2.e)
                Transmissions Costs. This approach assured all technologies
                appropriately contributed to the total system cost.
                 The Alliance commented in support of the agencies' accounting
                separately for the subsystems' costs and benefits for CISG, BISG, P2
                hybrid, power split hybrid (PS), and PHEV technologies.\1265\ The
                Alliance noted that these distinctions are important to capture the
                differences between various technologies, which can have separate
                packaging requirements, efficiency potentials, and vehicle
                applications. Ford echoed the Alliance comments on the modeling of
                electric vehicles in the NPRM, stating they supported the use of
                separate cost and benefits modeling for P2 and power split strong
                hybrid technologies.\1266\ Additionally, Ford commented that the
                modeling ``better reflects market realities by recognizing that
                manufacturers cannot simply pass on the entire incremental costs of
                hybrid, plug-in hybrid, and battery electric vehicles to the
                customers.''
                ---------------------------------------------------------------------------
                 \1265\ Alliance of Automobile Manufacturers, NHTSA-2018-0067-
                12073, at 140.
                 \1266\ Ford Motor Company, NHTSA-2018-0067-11928, at 10.
                ---------------------------------------------------------------------------
                 Comments from other stakeholders generally stated that the NPRM
                powertrain sizing approach resulted in costs for complete powertrains
                that were too high compared to other studies or market observations. In
                addition, as discussed in Section VI.C.1.g) Engine Costs, CARB also
                commented that the costs associated with IC engines were not excluded
                from the final costs of BEV vehicles.\1267\ CARB continued, stating
                that ``the final costs of BEV vehicles are higher due to the inclusion
                of the base absolute costs, to which the assigned BEV incremental cost
                would be added.'' The agencies agreed with CARB that inclusion of IC
                engine costs in the BEV cost was an error in the analysis.
                ---------------------------------------------------------------------------
                 \1267\ NHTSA-2018-0067-11873 at p.122.
                ---------------------------------------------------------------------------
                 In response to this comment, the agencies developed absolute costs
                for baseline engines for the CAFE Model so the absolute costs for IC
                engines could be removed from BEVs. In the final rule analysis, when a
                vehicle adopted BEV technology, the costs associated with IC powertrain
                systems were removed. As the vehicle walks through the technology tree,
                becoming a battery electric vehicle, the motor and inverter (ETDS)
                costs replaced the internal IC engine costs. Since the cost of the ETDS
                accounted for significant portion of the
                [[Page 24516]]
                total cost of electrification, it was important to accurately
                characterize the motor size (motor rating). To do this, the agencies
                used the MY 2017 market data file to compute the average engine power
                for each technology class.
                 For SHEVPS and SHEVP2 vehicles, as explained further in Section
                VI.C.3.e)(4)(c) Strong Hybrid Costs, the agencies computed the average
                rating for traction and generator motors across all road load
                combinations using Autonomie simulation runs. Since motor sizing varies
                based on road load levels, the average motor sizes acted as a mid-range
                representation for motor ratings across all road load combinations. The
                full range of motor sizes are driven by road load limits; the motor
                size for initial road load levels (MR0, AERO0 and ROLL0) would be
                larger compared to the motor size for highest level of road load
                reduction (MR6, AERO20 and ROLL20). After calculating the average motor
                size, the agencies applied the $18/kW metric (derived from the EETT
                Roadmap report) for both traction motors and generator motors. As
                discussed earlier, the agencies also used the cost of the CVTL2 as
                proxy to represent the cost of the eCVT used in power-split hybrid
                vehicle systems, and used the cost of the AT8L2 as proxy for the cost
                of the planetary gear set used in the P2 parallel hybrid system. The
                total cost of electrification for power-split hybrid vehicles includes
                the cost of the eCVT transmission, and the total cost of
                electrification for the P2 parallel hybrid vehicles includes the cost
                of the planetary gear set transmission.
                 CARB also submitted supplemental comments attempting a cost walk
                for electrified powertrain technologies, stating that inconsistencies
                in the model files and PRIA and lack of documentation about how the
                costs were derived ``[left] the public without the ability to
                understand why the costs are what they are and what should be
                applied.'' \1268\ Accordingly, a cost walk for a vehicle adopting an
                electrified powertrain is shown below. Additional comments on
                electrified powertrain costs are discussed in each individual
                technology section below, along with a discussion of changes made for
                the final rule in response to these comments.
                ---------------------------------------------------------------------------
                 \1268\ California Air Resources Board, NHTSA-2018-0067-12428, at
                25.
                ---------------------------------------------------------------------------
                 For the final rule analysis, the agencies have updated several
                electrification inputs and assumptions in response to these comments,
                as discussed in the previous sections. An example of how the costs are
                applied to a simulated vehicle platform's technology cost is discussed
                here, to assist CARB and other stakeholders in assessing
                electrification technology costs for the final rule analysis. The
                example shows the costs for a vehicle with conventional engine and
                transmission technology as it adds electrification technology.
                 The application of the electrification costs to an existing
                platform follows the same basic process for each technology on the
                electrification path. All technology costs used are for the model year
                of the electrification technology application. The first step is the
                process is the removal of the costs associated with the conventional
                drivetrain technologies. The next step is the application of the costs
                associated with the electrification technology. The costs include the
                cost of the engine, if applicable, transmission, non-battery
                components, and the battery pack. After the electrification costs are
                applied, other technology costs, such as aerodynamic or rolling
                resistance technologies are applied.
                 The specific example is the Toyota Rav4 LE AWD/XLE AWD simulated
                platform. The platform data were used from the reference run CAFE model
                standard setting vehicle_report.csv result file, augural standards
                results. The change in technology for the simulated platform was
                between MY 2023 and MY 2024. Table VI-107 shows the costing change
                between the MYs.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.240
                 Table VI-108 shows the costs, and where to find them, for the
                drivetrain components subtracted from the MY 2023 version of the
                platform. The costs for current engine and transmission were
                subtracted. To properly cost the engine it is important to note the
                engine was designated as a 4C1B engine, or, 4 cylinder 1 bank engine
                type. For more information about engine geometry designation in the
                technology input file please see Section VI.A.7 Structure of Model
                Inputs and Outputs.
                [[Page 24517]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.241
                 The costs for the new electrification technology were then applied.
                For the specific example the simulated vehicle platform is being
                converted to a PHEV20 powertrain. For all the technologies in the
                electrification path two major component groups were always added, the
                battery pack and the non-battery components. Hybrid electric
                technologies will also include the cost for an engine. Table VI-109
                shows the costing data for the non-battery pack electrification
                technology components, and where the cost data can be found.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.242
                 The battery pack is cost is determined by multiplying the baseline
                battery pack cost by the learn curve factor. Table VI-110 shows the
                calculation of the battery pack costs. The baseline battery costs are
                determined per discussions in Section VI.C.3.e)(1) Battery Pack
                Modeling.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.243
                 Table VI-111 shows a summary of the total cost application for the
                technology transition of the Rav4 example platform. The added costs of
                the addition of the LDB technology, improvement from AERO15 to AERO20,
                improvement from MR0 to MR1 are summarized. However, the costing data
                for these technologies can be found in the Technology Input file on the
                `SmallSUV' tab under each technology's respective rows.
                [[Page 24518]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.244
                 The following sections discuss specific electrification component
                cost comments on the NPRM, responses, and any relevant assumptions for
                the final rule analysis.
                a) Micro Hybrid Cost
                 As stated in PRIA, the cost of SS12V in NPRM included the cost of
                the battery, learning rate and retail price equivalent.\1269\ The
                assumed direct manufacturing cost (DMC) was the same as was used for
                the Draft TAR and the Proposed Determination,\1270\ but adjusted for
                learning and updated from 2013 to 2016 dollars. Cost learning made the
                cost of SS12V presented in the NPRM slightly lower than the Proposed
                Determination.
                ---------------------------------------------------------------------------
                 \1269\ Footnote n. 364 in PRIA; Table 6-32 and Table 6-33.
                 \1270\ Draft TAR Table 5.210.
                ---------------------------------------------------------------------------
                 ICCT compared the agencies' NPRM cost effectiveness estimate for
                SS12V with EPA's Proposed and Final Determination analyses, and
                concluded that the latter analyses found SS12V cost nearly $100 less
                than the agencies found in the NPRM, with a higher effectiveness
                benefit.\1271\ ICCT noted its difficulty in evaluating whether SS12V
                technology was actually cost-effective, since the NPRM CAFE model added
                the incremental cost of BISG over SS12V. ICCT stated that because SS12V
                is not as cost effective as other technologies in the electrification
                technology pathway, such as BISG, the analysis' estimate of SS12V costs
                was exaggerated and resulted in an unrealistic increase in compliance
                costs.
                ---------------------------------------------------------------------------
                 \1271\ International Council on Clean Transportation,
                ``Attachment 3_ICCT 15page summary and full comments appendix,''
                NHTSA-2018-0067-11741, at I-63.
                ---------------------------------------------------------------------------
                 While BISG is more expensive than the SS12V, BISG provides
                additional benefits such as smoother start-stop (reduced vibration
                during each start-stop event), launch assist and/or torque assist
                (during certain sudden acceleration while passing or load at low speed
                for short burst of time). Therefore, the effectiveness of SS12V should
                not be compared to BISG. The agencies have always considered BISG as a
                separate technology. Also, the effectiveness of SS12V in the Proposed
                Determination was determined using ALPHA modeling. A peer reviewer
                noted that ``[a]ccording to the documentation review, ALPHA's stop/
                start modeling appears to be very simplistic.'' \1272\ As discussed in
                Section VI.B.3 Autonomie model, the Autonomie tool simulates the
                technology as part of the full vehicle system, accounting for
                interactions with other technologies, and therefore the agencies
                believe the full-vehicle simulations provide more realistic
                effectiveness estimates than the value from the Proposed Determination.
                For these reasons, the agencies disagree with ICCT's assertions. For
                SS12V, the agencies continued to use the costs from the NPRM, which are
                consistent with the Draft TAR and Proposed Determination. The ETDS
                costs presented in the final rule do not include the cost of the
                battery.
                ---------------------------------------------------------------------------
                 \1272\ Peer Review of ALPHA Full Vehicle Simulation Model, at C-
                4, available at https://nepis.epa.gov/Exe/ZyPdf.cgi?Dockey=P100PUKT.pdf.
                ---------------------------------------------------------------------------
                b) Mild Hybrid Cost
                 The belt integrated starter generator (BISG) and crank integrated
                starter generator (CISG), sometimes referred to as mild hybrid systems,
                provide idle-stop capability and use a higher voltage battery with
                increased energy capacity over typical automotive batteries. The higher
                voltage allows the use of a smaller, more powerful and efficient
                electric motor/generator which replaces the standard alternator. For
                the NPRM the agencies developed the costs for the mild hybrid systems
                assuming the use of a 115V system. The battery, motor, and supporting
                components were sized and costed based on this voltage level.
                 Many commenters asserted that the costs presented in the NPRM
                analysis for BISG and CISG systems were inflated or incorrect.\1273\
                ICCT noted that because mild hybrid systems were
                [[Page 24519]]
                widely adopted by the fleet under the augural standards, the high cost
                of those systems had a significant impact on the costs of the
                standards.\1274\
                ---------------------------------------------------------------------------
                 \1273\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741; Union of Concerned Scientists, NHTSA-2018-0067-
                12039; Fiat Chrysler Automobiles, NHTSA-2018-0067-11943; Alliance of
                Automobile Manufacturers, NHTSA-2018-0067-12073; California Air
                Resources Board, NHTSA-2018-0067-11873.
                 \1274\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741, at I-24.
                ---------------------------------------------------------------------------
                 Meszler Engineering Services noted that the NPRM documentation
                presented BISG/CISG battery costs that were ``not unreasonable,'' and
                that the CAFE model database of battery costs used for NPRM analysis
                included estimates for those electrification technologies that were
                $259 higher than those presented in the NPRM documentation.\1275\
                Meszler surmised that it initially appeared as if the model may have
                been applying a redundant RPE factor to BISG/CISG costs, but noted that
                the determination that the costs differed from those documented by a
                constant absolute offset made that assumption an unlikely possibility.
                ---------------------------------------------------------------------------
                 \1275\ Meszler Engineering Services, NHTSA-2018-0067-11723
                Attachment 2.
                ---------------------------------------------------------------------------
                 ICCT and UCS both noted the discrepancy between the reported
                battery costs in the PRIA and costs reported in the NPRM Autonomie
                simulation databases.\1276\ ICCT disagreed with the agencies' approach
                to modeling batteries in the NPRM analysis, stating that ``[n]ot only
                is [the Argonne] database exceedingly difficult to access to modify
                battery costs (as battery costs should be a user input), but it makes
                it much harder to see how battery costs affect mild hybrid costs over
                time.'' \1277\ Claimed difficulties aside, ICCT concluded that the
                battery costs were outdated and grossly overstated, based on the tables
                in section 6.3.9.12 of the PRIA and the outputs of the low battery cost
                sensitivity case, which ICCT stated were more closely aligned with EPA
                and other research on battery costs. ICCT presented its own best
                estimate of NPRM BISG costs, stating that they were not able to make
                the PRIA and datafile costs match up.
                ---------------------------------------------------------------------------
                 \1276\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741; Union of Concerned Scientists, NHTSA-2018-0067-
                12039.
                 \1277\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741.
                ---------------------------------------------------------------------------
                 Several commenters noted that the costs of BISG/CISG systems were
                higher for Small Cars/SUVs and Medium Cars than for Medium SUVs and
                Pickup trucks, which the Alliance and FCA described as ``implausible''
                and ``misaligned with industry understanding,'' and which ICCT
                described as ``contrary to basic engineering logic, which holds that a
                system which would be smaller and have lower energy and power
                requirements would be less expensive, not more.'' \1278\ Both ICCT and
                UCS stated that regardless of alleged errors in costs between
                technology classes, even the lower of the values presented in the PRIA
                overestimated the cost of mild hybrid batteries.\1279\
                ---------------------------------------------------------------------------
                 \1278\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741.
                 \1279\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741; Union of Concerned Scientists, NHTSA-2018-0067-
                12039.
                ---------------------------------------------------------------------------
                 The Alliance and FCA urged the agencies to update the CAFE model to
                address this issue so that the cost of compliance was properly
                reflected in the results. To estimate the impact of the error, the
                Alliance and FCA modified the technology input file so that the Medium
                SUV and Pickup truck electrification costs were changed to be identical
                to the Small Car/SUV and Medium Car costs for SS12V, BISG, and CISG,
                and re-ran the CAFE model to show an estimated $13 billion increase in
                compliance costs under the augural standards with the error
                corrected.\1280\
                ---------------------------------------------------------------------------
                 \1280\ Fiat Chrysler Automobiles, NHTSA-2018-0067-11943;
                Alliance of Automobile Manufacturers, NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 Conversely, CARB modified the fuel consumption improvement
                estimates for BISG systems to match those predicted by Argonne in a
                recent report after calculating the smallest modified improvement from
                MYs 2015-2025 for five vehicle classes, resulting in efficiency
                improvements of 8.5-11 percent.\1281\ CARB also reduced the non-battery
                costs for Small Car/SUVs to match the non-battery costs for Medium SUV
                and Pickup trucks, which CARB stated still reflected higher costs than
                those previously used by EPA in the Proposed Determination. CARB did
                not modify the battery costs, but did comment that they were overstated
                by approximately 50 percent ``due to the erroneous oversizing of the
                battery.'' CARB's modified run decreased average vehicle technology
                costs by a range of $300-$500 per year, ``reflecting an approximate 25
                percent drop in 2029 model year incremental technology costs to meet
                the existing standards relative to the rollback standards.''
                ---------------------------------------------------------------------------
                 \1281\ California Air Resources Board, NHTSA-2018-0067-11873
                (``Specifically, the fuel consumption improvements modeled by ANL in
                the most recent report for DOE were utilized in place of the
                assumptions used for the Agencies' analysis. As noted above, ANL,
                via Autonomie modeling, identified efficiencies between 8.5 percent
                to 12.7 percent for mild hybrids, relative to both gasoline spark
                ignited and relative to turbocharged gasoline spark ignited across
                five different vehicle classes. Using approximately the smallest
                modeled improvement across the 2015 to 2025 model years for each of
                the five classes, improvements of 8.5 percent-11 percent were
                utilized for a modified CAFE Model run.'').
                ---------------------------------------------------------------------------
                 Commenters also pointed to prior agency analyses, studies, and
                applications of BISG systems to provide examples of what they believed
                BISG system costs should be, with ICCT arguing that the agencies' cost
                values for BISG/CISG systems were contrary to the research and
                evidence.\1282\ HDS noted that the 2018 PRIA estimate was approximately
                double the estimate from the 2016 Draft TAR, that the difference in
                battery costs between those two analyses did not explain the
                difference, and that there was no discussion in the PRIA that did
                so.\1283\
                ---------------------------------------------------------------------------
                 \1282\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741.
                 \1283\ H-D Systems, NHTSA-2018-0067-11985.
                ---------------------------------------------------------------------------
                 UCS stated that BISG system costs have already reached that which
                was predicted in EPA's first Final Determination, published in 2017,
                for 2025, and would decline further because of continued volume-based
                learning.\1284\ UCS also cited a 2018 Argonne report that estimated the
                battery component cost for a mild hybrid system to be $159.35, and a
                Chevrolet Malibu eAssist teardown study that estimated total battery
                subsystem direct costs at $166, and battery modules, power
                distribution, and covers at $120 in direct manufacturing costs.\1285\
                UCS summarized that the aforementioned costs are less than half the
                costs listed in the PRIA and approximately one quarter of the
                ``BatPaCCost'' value given in the Argonne input files. UCS also cited
                cost estimates from the 2015 NAS report and two EPA reports, and
                concluded that the agencies did not sufficiently explain why the NPRM
                cost data differed so substantially from this other available
                information.
                ---------------------------------------------------------------------------
                 \1284\ Union of Concerned Scientists, NHTSA-2018-0067-12039.
                 \1285\ Id. (citing [Component Cost, ANL 2017k]).
                ---------------------------------------------------------------------------
                 ICCT cited its own 2016 study of supplier costs with estimates for
                48V mild hybrid systems, estimating the system cost at $600-$1,000
                (with costs on the lower side for cars and the higher side for light
                trucks) in the 2025 timeframe.\1286\ ICCT pointed to the RAM 1500
                pickup truck as an example of a vehicle with a BISG system that ``has
                already validated the ICCT figures in 2019.'' ICCT noted that the BISG
                system, branded as eTorque, was first offered as a ``free standing''
                option on the RAM 1500 truck for $800, and that price was recently
                raised to $1,450. ICCT stated that even with the higher price, applying
                the agencies' RPE of 1.5 means
                [[Page 24520]]
                that the direct manufacturing cost is less than $1,000, which is less
                than the $1,616 direct manufacturing cost estimate in the NPRM for 2016
                pickup trucks.\1287\ Similarly, UCS cited the $500 premium that General
                Motors charged for the technology on its Chevrolet Silverado pickup
                trucks with eAssist.\1288\
                ---------------------------------------------------------------------------
                 \1286\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741.
                 \1287\ ICCT also stated that the eTorque system offered improved
                performance and driveability and contributes to higher payload and
                towing ratings for 2019 compared with 2018, and noted that the
                agencies ``have completely failed to account for the consumer value
                of the utility benefits'' from the system. The agencies' approach to
                simulating performance neutrality and the consumer benefit of
                increased performance are discussed in Section VI.B.3.a)(6)
                Performance Neutrality.
                 \1288\ Union of Concerned Scientists, NHTSA-2018-0067-12039.
                ---------------------------------------------------------------------------
                 The agencies reviewed all of the comments and information provided.
                It appears there may have been confusion about what costs were used for
                the Draft TAR and NPRM. For the Draft TAR, non-battery BISG costs,
                including learning and RPE, were $1,701 compared to $1,186 for the NPRM
                (both costs in 2018 dollars). Therefore, the costs for the NPRM were
                lower than for the Draft TAR when cost accounting is on an equivalent
                basis.
                ---------------------------------------------------------------------------
                 \1289\ Table 5.131 in Draft TAR ($1,045 x 1.5 = $1567.5 in
                2013$. (Absolute cost, without batteries. This includes learning and
                Retail Price Equivalent).
                 \1290\ Table 6-32 in PRIA (Absolute Electrification Cost without
                batteries. This includes learning and Retail Price Equivalent).
                 \1291\ See Table I 19--Cost and Mass Estimate of BISG
                components.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.245
                 The agencies also determined the cost presented by EPA in Draft TAR
                (see Table 5.131 in Draft TAR) was the direct manufacturing cost of the
                BISG system, and not the retail price equivalent. The Draft TAR cost
                estimate in Table VI-112 includes the RPE and costs updated from 2013
                to 2018 dollars. The agencies agree with the commenters about the
                discrepancy in the cost of the battery pack for the BISG system
                presented in PRIA and in CAFE model. To avoid any confusion, Table VI-
                112 shows the non-battery costs of the BISG system.
                 After considering the comments and reviewing the approach used in
                the NPRM, the agencies agreed updating the cost of the BISG system was
                appropriate for the final rule analysis. Adjustments were based on
                using a 48V BISG system instead of the 115V system used for the NPRM.
                For the final rule, the agencies considered several cost sources,
                including the EPA-sponsored FEV report titled: Light-Duty Vehicle
                Technology Cost Analysis on 2013 Chevrolet Malibu ECO with eAssist BAS
                Technology Study.\1292\ Based on the teardown study, EPA estimated the
                direct manufacturing cost of the BISG system (without batteries) to be
                $1,045 in 2013 dollars. This included a cost adjustment for reduced
                voltage insulation. The agencies also considered the 2019 Dodge Ram
                eTorque system retail price. A cost of $1,195 for water-cooled system
                and $1,450 for air-cooled system in 2018 dollars was deduced from the
                retail price of eTorque assist (BISG) system. The 2015 NAS report
                estimated the cost range of BISG technology at $888 to $1,164 in 2010
                dollars in 2025.\1293\ This is equivalent to a range of $1,020 to
                $1,337.27 in 2018 dollars in 2025. The agencies also reviewed
                confidential business information on BISG cost and mass estimates
                provided by manufacturers.
                ---------------------------------------------------------------------------
                 \1292\ Light Duty Vehicle Technology Cost Analysis 2013
                Chevrolet Malibu ECO with eAssist BAS Technology Study, FEV P311264
                (Contract no. EP-C-12-014, WA 1-9).
                 \1293\ Cost, Effectiveness and Deployment of Fuel Economy
                Technologies for Light-Duty Vehicles, National Academy of Sciences,
                2015.
                ---------------------------------------------------------------------------
                 For the final rule analysis, the agencies used the A2Mac1 database
                to develop a bill of materials for BISG systems. The agencies sourced
                cost estimates for the motor, inverter and DC-DC converter from the
                2017 EETT roadmap report.\1294\ The agencies used BatPaC model version
                3.1 to perform a standalone analysis determining the cost of a battery
                pack for the 48V system.1295 1296 Table VI-113 shows the
                cost and mass estimates for BISG components used in the final rule.
                ---------------------------------------------------------------------------
                 \1294\ U.S. DRIVE, Electrical and Electronics Technical Team
                Roadmap (October 2017), https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
                 \1295\ A Detailed Vehicle Simulation Process To Support CAFE and
                CO2 Standards for the MY 2021--2026 Final Rule Analysis, at Table
                50.
                 \1296\ BatPac 10032018 BISG Version 3.1--28June2018_FINAL.
                ---------------------------------------------------------------------------
                [[Page 24521]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.246
                 The agencies compared the cost estimates in the 2017 EETT roadmap
                report and found they aligned well with cost estimates from sources
                cited by commenters. For reference, Table VI-113 above showed the cost
                estimate for BISG system (without the battery) used in Draft TAR, NPRM
                and in Final Rule. Furthermore, the agencies considered the Alliance
                and FCA analysis, provided in their respective comments, recommending
                the use of the same BISG system cost for both cars and
                trucks.1297 1298 This analysis, supplemented with CBI data,
                demonstrated that the costs for implementing BISG systems on different
                vehicle classes was not appreciably different. The agencies agree with
                this assessment. For the final rule analysis, the cost of the BISG
                system is the same for cars, SUVs, and pickups.
                ---------------------------------------------------------------------------
                 \1297\ Fiat Chrysler Automobiles, NHTSA-2018-0067-11943, at 85.
                 \1298\ Alliance of Automobile Manufacturers, NHTSA-2018-0067-
                12073, at 140-42.
                ---------------------------------------------------------------------------
                (c) Strong Hybrid Cost
                 In the NPRM and this final rule analysis, the total cost for strong
                hybrids (SHEVP2 and SHEVPS) included the electric machine, battery
                pack, IC engine, and transmission. Discussed earlier in Section
                VI.C.3.d) Electrification Effectiveness Modeling, each strong hybrid
                powertrain is optimized for the given vehicle class by appropriate
                sizing of the electric machine, IC engine and battery pack.
                Accordingly, the costs represent the optimized system. For the NPRM,
                the agencies referred to the ``Assessment of vehicle sizing, energy
                consumption, and cost through large-scale simulation of advanced engine
                technologies'' report to estimate the cost and effectiveness for
                different hybrid systems for the NPRM.\1299\ For the final rule, as
                discussed in Section 2) and further below, the agencies sourced cost
                estimates from the October 2017 U.S. DRIVE report, ``Electrical and
                Electronics Technical Team Roadmap.'' \1300\
                ---------------------------------------------------------------------------
                 \1299\ Moawad, Ayman, Kim, Namdoo, Shidore, Neeraj, and
                Rousseau, Aymeric. Assessment of Vehicle Sizing, Energy Consumption
                and Cost Through Large Scale Simulation of Advanced Vehicle
                Technologies (ANL/ESD-15/28). United States (2016), available at
                https://www.autonomie.net/pdfs/Report%20ANL%20ESD-1528%20-%20Assessment%20of%20Vehicle%20Sizing,%20Energy%20Consumption%20and%20Cost%20through%20Large%20Scale%20Simulation%20of%20Advanced%20Vehicle%20Technologies%20-%201603.pdf.
                 \1300\ U.S. DRIVE, Electrical and Electronics Technical Team
                Roadmap (October 2017), https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
                ---------------------------------------------------------------------------
                 SHEVP2 and SHEVPS have different characteristics and in turn have
                different costs, as reflected in both the NPRM and this final rule
                analysis. The cost for engines and transmissions for SHEVP2s are based
                on estimates discussed further in Sections VI.C.1 Engine Path and
                VI.C.2 Transmission Path, respectively. The cost for SHEVP2 electric
                machines and battery packs were dependent on their sizes, which were
                optimized by the Autonomie sizing algorithm. SHEVPS total powertrain
                costs includes the optimized battery pack, electric machine, an
                Atkinson engine, and the CVT.
                 Many commenters generally stated that the costs of hybrid
                technology were overestimated in comparison to prior agency estimates
                and other publicly available sources, and that the agencies'
                documentation of hybrid system costs was unclear.
                 Meszler Engineering Services commented that the net costs of
                vehicles that apply SHEVP2 technology were in error, resulting from the
                way that the CAFE model applied HCR, CEGR and TURBO technology in
                combination with the SHEVP2 strong hybrid system.\1301\
                ---------------------------------------------------------------------------
                 \1301\ Meszler Engineering Services, NHTSA-2018-0067-11723.
                ---------------------------------------------------------------------------
                 HDS claimed that cost estimates for both SHEVP2 and SHEVPS were
                significantly higher than the Draft TAR estimates, differing by a
                factor of about 2 for SHEVP2 and by a factor of 2.5 for SHEVPS, with no
                justification given for the increase in costs.\1302\ HDS noted that the
                SHEVPS cost estimates were particularly surprising since the costs have
                been investigated extensively since that technology was introduced to
                the market over a decade ago. HDS stated that the 2016 TAR estimates
                were in line with other analyses like the NAS
                [[Page 24522]]
                estimate, and consistent with actual retail price increments observed
                in the market.
                ---------------------------------------------------------------------------
                 \1302\ H-D Systems, NHTSA-2018-0067-11985.
                ---------------------------------------------------------------------------
                 HDS also pointed to cost estimates based on teardown studies
                sponsored by EPA and the European Union,\1303\ public cost data
                disclosed by suppliers of hybrid systems, and the retail prices of
                available hybrid vehicles as estimates that contradict the agencies'
                NPRM cost estimates. HDS compared the European Vehicle Market Phase 1
                FEV cost analysis to the costs published by EPA in the TAR, concluding
                that the EU costs ``even at [levels adjusted for the strength of the
                Euro] are quite similar to EPA estimates of $2,650 to $3,300 (depending
                on vehicle size) published in the TAR for the P2 hybrid, and also shows
                that the PS hybrid is just 7 percent more expensive than the P2
                hybrid.'' HDS stated that battery costs have also certainly decreased
                since 2012 when the report was written, so current costs are estimated
                to be approximately $400 less than the values cited above.
                ---------------------------------------------------------------------------
                 \1303\ Id., citing FEV, Light-Duty Vehicle Technology Cost
                Analysis-European Vehicle Market (Phase 1), (2012, updated 2013),
                available at https://www.theicct.org/.
                ---------------------------------------------------------------------------
                 HDS also cited a methodology to estimate costs from retail price
                increments in the market,\1304\ stating that a typical cost-to-retail
                price ratio is 1.5. Applying this methodology, the cost of the SHEVPS
                hybrid as used by Ford and Toyota would be in the $2,500 to $3,000
                range, the cost of a SHEVP2 as used by Hyundai Kia would be $2,250, and
                the cost of a low volume and/or luxury model system would be estimated
                at $3,300 for a SHEVP2.
                ---------------------------------------------------------------------------
                 \1304\ Id. (citing Vincentric Hybrid Analysis, executive
                summary, www.vincentric.com/Home/IndustryReports/HybridAnalysis
                October2014.aspx.).
                ---------------------------------------------------------------------------
                 Similarly, ICCT stated that the agencies failed to analyze properly
                the dozens of hybrid vehicles in the marketplace, their costs which
                were lower than the agencies assumed, and their rapid improvements from
                automakers and suppliers competitively developing lower cost components
                for those vehicles.\1305\ ICCT observed an incremental price increase
                in the analysis for hybrid vehicles under the augural standards of
                approximately $6,600 per hybrid vehicle in 2017 and $4,800 in 2025, and
                concluded that this was not a plausible result considering hybrid
                component costs and full-vehicle prices in the marketplace in 2016 as
                well as the technology improvement that continues to enter the fleet.
                ICCT stated that the agencies must set a maximum cost premium for full
                hybrids of $2,500 in 2017, declining linearly to $1,400 by 2025 for
                mid-size cars and crossovers, with cost components likely scaling by
                vehicle power requirements (up for pickups, down for smaller cars),
                which it stated the agencies must also account for in the modeling.
                ---------------------------------------------------------------------------
                 \1305\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741.
                ---------------------------------------------------------------------------
                 ICCT stated that the agencies must disclose the basis for the
                ``unrealistically high'' hybrid system cost estimates, such that the
                public can clearly connect the bottom-up cost components to full
                vehicle costs for all vehicle models that have hybrid cost
                applied.\1306\ ICCT stated that hybrid system cost estimates are ``one
                of the most important technology cost estimations to assess the Augural
                standards' compliance cost, as the NPRM projects that 22 percent of
                vehicles will need full hybrid systems to meet the augural standards,''
                and accordingly after disclosing those costs, the agencies must provide
                another opportunity for public comment. Similarly, CARB stated that it
                was unable to decipher the hybrid cost components, and without that
                information could only guess as to why the costs increased relative to
                costs in the Draft TAR and EPA's Proposed Determination.\1307\ As such,
                CARB stated they could not make a conclusion as to whether improper
                battery resizing, incorrectly modeled batteries, or oversized electric
                motors contributed to the overestimation of costs for strong hybrid
                systems.
                ---------------------------------------------------------------------------
                 \1306\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741.
                 \1307\ California Air Resources Board, NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 The agencies believe comparing the retail price of P2 or PS hybrid
                to conventional vehicles could be misleading. Even though hybrid
                vehicles may have higher direct manufacturing costs, manufacturers may
                choose not to price it higher than the conventional version of the
                vehicle. In other words, manufacturers may choose to subsidize the cost
                of hybrid technologies to gain overall credit for fleetwide compliance.
                Therefore, the agencies believe that comparing retail price between
                hybrid and conventional vehicles should be done only when other sources
                of information are available to corroborate the differences in retail
                price.
                 The agencies also referred to an EPA-sponsored teardown and cost
                estimate report as suggested by HDS. Table VI-114 shows the absolute
                cost of P2 and PS hybrid systems as estimated in the EPA sponsored
                teardown report and the absolute cost estimated in the final rule in
                2018$. As indicated above, the absolute cost in the final rule includes
                the cost of transmissions for the PS and P2 hybrid systems. The EPA
                teardown cost estimate includes the cost of the eCVT for the PS hybrid
                systems only. The P2 hybrid system costs do not include the cost of
                engine and transmission in the table below.
                 Although ICCT suggested that the agencies cap the maximum cost
                premium for full hybrids of $2,500 in 2017 and linearly decrease the
                cost to $1,400 by 2025, ICCT did not provide any supporting material to
                suggest that maximum upper limit of $2,500 for full hybrid is
                economically feasible, nor did they provide an example of an existing
                full hybrid vehicle in the marketplace with a technology increase of
                $2,500 in 2017. ICCT also did not make it clear if the costs suggested
                would be applicable to P2 or PS hybrid architecture.
                 Based on the comments, the agencies reassessed SHEVP2 and SHEVPS
                cost estimates for the final rule. As discussed above, the agencies
                referred to U.S. DRIVE's October 2017 report, ``Electrical and
                Electronics Technical Team Roadmap'' \1308\ to estimate the cost of
                motors and inverters. The agencies also agreed with commenters and
                referenced the MY 2016 Chevrolet teardown report by UBS to estimate the
                cost of other hybrid components such as wiring harness, cables,
                voltage-step-down DC to DC converters, and on-board chargers. Per
                Section VI.C.3.e)(2) Non-battery Electrification Component Costs, for
                the final rule, the cost of non-battery hybrid system components
                includes the cost of traction motor, motor/generators, high voltage
                cables and connectors, charging cord, and on-board chargers. The cost
                of the planetary gear set is also included in the cost of non-battery
                components. Per Section VI.B.4 Technology Costs, for the final rule,
                the cost of hybrid systems is presented as absolute cost, and not as an
                incremental to some previous technology (absolute cost includes the
                retail price equivalent). The agencies used the cost of the AT8L2
                transmission as a cost proxy for the planetary gear set in P2 hybrid
                systems, and used the cost of CVTL2 transmission as a cost proxy for
                planetary gear set for PS hybrid systems. It should also be noted the
                costs shown here do not include the cost of engine coupled to the
                hybrid system.
                ---------------------------------------------------------------------------
                 \1308\ U.S. DRIVE, Electrical and Electronics Technical Team
                Roadmap (October 2017), https://www.energy.gov/sites/prod/files/2017/11/f39/EETT%20Roadmap%2010-27-17.pdf.
                ---------------------------------------------------------------------------
                 The agencies reviewed the FEV 2010 Ford Fusion HEV teardown report,
                Light Duty Technology Cost Analysis, Power-
                [[Page 24523]]
                Split and P2 HEV Case Studies.\1309\ In a Split-HEV architecture, there
                are two motors; one motor provides torque while the other motor act as
                a generator to recapture the energy during regenerative braking. The
                report does not capture the cost of motor-generator and the cost of the
                DC to DC converter. The report did not include an extensive teardown of
                a P2 hybrid vehicle, but rather made a cost adjustment for the PS motor
                and inverter to reflect additional cost. Table VI-114 shows the
                breakdown of cost estimates for the electric machine in the 2010 Ford
                Fusion HEV.\1310\ Since the costs were developed in 2009$, the cost
                estimates for the same components are presented in 2018$. Table VI-115
                shows the cost estimate for electric machines for a midsize passenger
                car for MY 2017 in 2018$.\1311\ The cost is estimated using the EETT
                Roadmap report as explained earlier. Since EPA uses indirect cost
                multiplier (ICM) to determine the final retail price, and ICMs vary for
                different technologies, the agencies compared the direct manufacturing
                cost from report to the direct cost estimate in the final rule.
                ---------------------------------------------------------------------------
                 \1309\ Light Duty Technology Cost Analysis, Power-Split and P2
                HEV Case Studies, EPA-420-R-11-015 (November 2011), available at
                https://nepis.epa.gov/Exe/ZyPDF.cgi/P100EG1R.PDF?Dockey=P100EG1R.PDF.
                 \1310\ Table D-4 (components considered are transmission, power
                distribution cables and Inverter). The cost of inverter is from
                Table D-11.
                 \1311\ Average peak power for the traction motor used in this
                final rule is 72kW, and 37kW continuous power for the generation
                motor.
                ---------------------------------------------------------------------------
                 The direct manufacturing cost estimated in the Light Duty
                Technology Cost Analysis, Power-Split and P2 HEV Case Studies published
                for EPA is $3,689.28 in 2018$, and direct manufacturing cost estimated
                for electric machines in this final rule is $4,355.82. As mentioned
                before, the cost of the motor-generator and the cost of the DC to DC
                converter is not captured in that report.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.247
                [GRAPHIC] [TIFF OMITTED] TR30AP20.248
                (d) PHEV Cost
                 Plug-in hybrid vehicles' costs were developed similar to strong
                hybrids for the NPRM analysis and the final rule analysis. The plug-in-
                hybrid system components were optimized, per Section VI.C.3.d)(2)
                Modeling and Simulating Vehicles with Electrified Powertrains in
                Autonomie and the resultant systems were used to determine costs, per
                Battery Pack Modeling and Non-battery Electrification Component Costs.
                Per Section VI.C.3.c) Electrification Adoption Features, the agencies
                used one engine technology and one transmission technology per plug-in
                hybrid architecture type.
                 For PHEVs following SHEVP2 on the hybrid/electric architecture
                path, per Section VI.C.3.a)(1) Electrification technologies, the total
                cost of the technology package was determined from summing the costs of
                the TURBO1 engine, the AT8L2 transmission, and the battery and non-
                battery electrification technology components. For PHEVs following
                SHEVPS on the hybrid/electric architecture path, per Section
                VI.C.3.a)(1) Electrification technologies, the total cost of the
                technology package was determined from summing the costs of the
                Atkinson engine, the CVT transmission, and the battery and non-battery
                electrification technology components.
                 CARB provided observations about non-battery component costs for
                PHEVs,
                [[Page 24524]]
                arguing that what the agencies asserted for the incremental costs of a
                PHEV over a strong hybrid vehicle are not supported in the
                market.\1312\ CARB cited the Toyota Prius Prime and Hyundai Sonata as
                examples of vehicles that share most of their components with their
                non-plug-in hybrid counterparts, with components like the on-board
                charger and higher voltage, larger energy capacity battery pack
                excepted. CARB stated the agencies' lack of discussion about how non-
                battery component costs were developed made it ``virtually impossible
                to understand what the drivers are for the increases in costs relative
                to the Agencies' previous analysis for the 2016 Draft TAR and EPA's
                Proposed Determination.'' CARB concluded that the available PHEV market
                offerings do not support the higher costs relative to the Draft TAR and
                EPA's Proposed Determination analyses, and no justification was
                provided for the change.
                ---------------------------------------------------------------------------
                 \1312\ California Air Resources Board, NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 The agencies agree with CARB that the incremental costs of PHEV
                over strong hybrid costs were too high, and that values were not
                supported by the market. In response to this comment, the agencies
                updated the non-battery component costs as well as the battery costs to
                better reflect the market values. In addition, the agencies have
                optimized the Autonomie modeling in a way to maintain the same engine,
                transmission and other components from a SHEVP2 or SHEVPS moving to a
                PHEV20/50 or PHEV20T/50T.\1313\ For further discussions on PHEV
                modeling and updates, see Section VI.C.3.a)(1) Electrification
                technologies and Section VI.C.3.d) Modeling and Simulating Vehicles
                with Electrified Powertrains in Autonomie. The updates discussed here
                and applied to the final analysis resulted in values that more
                accurately represented PHEV technology costs.
                ---------------------------------------------------------------------------
                 \1313\ I.e., a SHEVP2 with a turbocharged engine may adopt
                PHEV20T or PHEV50T technology, but a SHEVPS will only ever adopt
                PHEV20 or PHEV50 technology, as the SHEVPS do not use turbocharged
                engines.
                ---------------------------------------------------------------------------
                (e) BEV Cost
                 For the NPRM and this final rule analysis, the total costs of BEVs
                included optimized battery pack and electric machine costs. Like the
                other electrified powertrains, Autonomie optimized both the size of the
                battery pack and electric machine to fulfill the performance neutrality
                requirements for each vehicle. Further discussion on electrification
                technology component sizing and optimization is provided in Section
                VI.C.3.d) Modeling and Simulating Vehicles with Electrified Powertrains
                in Autonomie. Discussion on electrification component costing is
                provided in Battery Pack Modeling and Non-battery Electrification
                Component Costs. When computing the total cost of a vehicle, the
                agencies remove the costs of the IC engines and transmission when a
                conventional or hybridized powertrain adopts BEV technologies. In
                Section VI.C.1 Engines Path and Section VI.C.22 Transmission, the
                agencies discussed the absolute costs used for engine and transmission
                technologies in the final rule analysis.
                 ICCT stated that if the agencies had considered BEV battery and
                other component costs correctly, cost parity would be reached with
                conventional combustion vehicles in the 2025-2027 timeframe.\1314\ ICCT
                went on to allege that if the agencies removed all constraints on
                electric vehicles,\1315\ they would appropriately realize that the 2025
                standards are more cost-effective if electric vehicles are included.
                ---------------------------------------------------------------------------
                 \1314\ International Council on Clean Transportation, NHTSA-
                2018-0067-11741.
                 \1315\ As discussed above, the agencies believe that ICCT
                misunderstood the agencies' statutory obligations and the
                differences between the standard setting modeling scenario and the
                ``real-world'' modeling scenario. The agencies did not apply
                additional constraints on BEVs in the NPRM analysis.
                ---------------------------------------------------------------------------
                 The agencies disagree with ICCT's statement that BEVs would reach
                parity to IC engines by the 2025-2027 timeframe. For this final rule
                analysis, the agencies have updated the battery pack costs, electric
                machine costs, and excluded costs of IC engines and transmission when a
                vehicle was converted to a BEV. However, the costs still did not reach
                parity within the rulemaking time frame. Furthermore, NHTSA notes that
                the decision to exclude BEV technology from the CAFE program standard-
                setting analysis is not a choice made by the agency, but a statutory
                requirement.\1316\
                ---------------------------------------------------------------------------
                 \1316\ See 49 U.S.C. 32902(h).
                ---------------------------------------------------------------------------
                (f) FCV Cost
                 For the NPRM and the final rule analysis the agencies considered
                fuel cell vehicle technology advancements in hydrogen storage tanks,
                sensors and control systems, and market penetration.\1317\ The agencies
                are also considered the availability of hydrogen refueling stations
                across the country and cost of compressed hydrogen.1318 1319
                Although the agencies did not receive any comments on the cost of fuel
                cell vehicles, the agencies updated the cost of hydrogen storage tanks
                and fuel cells based on a cost analysis from Department of Energy
                (DOE), Office of Energy Efficiency and Renewable Energy (EERE), Fuel
                Cell Technologies Office.\1320\
                ---------------------------------------------------------------------------
                 \1317\ The agencies referenced EPA's 2018 Automotive Trends
                Report, available at https://nepis.epa.gov/Exe/ZyPDF.cgi/P100W5C2.PDF?Dockey=P100W5C2.PDF, for information about FCV market
                penetration.
                 \1318\ MIT Energy Initiative. Insights into Future Mobility
                (2019). Cambridge, MA: MIT Energy Initiative. http://energy.mit.edu/insightsintofuturemobility.
                 \1319\ U.S. Department of Energy, Alternative Fuels Data Center:
                Alternative Fueling Station Counts by State: https://afdc.energy.gov/stations/states (last visited January 3, 2020).
                 \1320\ James et al., Final Report: Hydrogen Storage System Cost
                Analysis (September 2016), available at https://www.osti.gov/servlets/purl/1343975.
                ---------------------------------------------------------------------------
                 The DOE estimates that the cost of a compressed gas storage system
                is around $28/kWh (assumed rate of production of 10,000 units per
                year). The hydrogen fuel price ranges from $12.85 to $16 per kilogram,
                which translates to approximately $5.60 per gallon on an equivalent
                energy basis.\1321\
                ---------------------------------------------------------------------------
                 \1321\ California Fuel Cell Partnership: https://cafcp.org/content/cost-refill (last visited January 3, 2020).
                ---------------------------------------------------------------------------
                 Table VI-116 shows the evolution of the fuel cell vehicle costs
                from the Draft TAR to final rule (costs include the fuel cell, control
                systems, motors, inverters, hydrogen storage tanks, wiring harness,
                hydrogen fuel sending lines, safety systems, sensors and hardware for
                mounting and installation). The cost of the battery pack and battery
                management system is not included in the cost of the fuel cell vehicle.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.249
                [[Page 24525]]
                4. Mass Reduction
                 Mass reduction is a relatively cost-effective means of improving
                fuel economy and reducing CO2 emissions, and vehicle
                manufacturers are expected to apply various mass reduction technologies
                to meet fuel economy and CO2 standards. Reducing vehicle
                mass can be accomplished through several different techniques, such as
                modifying and optimizing vehicle component and system designs, part
                consolidation, and adopting lighter weight materials (advanced high
                strength steel, aluminum, magnesium, and plastics including carbon
                fiber reinforced plastics). The cost for mass reduction depends on the
                type and amount of materials used, the manufacturing and assembly
                processes required, and the degree to which changes to plants and new
                manufacturing and assembly equipment is needed. In addition,
                manufacturers may develop expertise and invest in certain mass
                reduction strategies that may affect the approaches for mass reduction
                they consider and the associated costs. Manufacturers may also consider
                vehicle attributes like noise-vibration-harshness (NVH), ride quality,
                handling, and various acceleration metrics when considering how to
                implement any mass reduction strategy. See Section VI.B.3.a)(5)
                Maintaining Vehicle Attributes for more details.
                 The automotive industry uses different metrics to measure vehicle
                weight. Some commonly used measurements are vehicle curb weight,\1322\
                gross vehicle weight (GVW),\1323\ gross vehicle weight rating
                (GVWR),\1324\ gross combined weight (GCVW),\1325\ and equivalent test
                weight (ETW),\1326\ among others.
                ---------------------------------------------------------------------------
                 \1322\ This is the weight of the vehicle with all fluids and
                components but without the drivers, passengers, and cargo.
                 \1323\ This weight includes all cargo, extra added equipment,
                and passengers aboard.
                 \1324\ This is the maximum total weight of the vehicle,
                passengers, and cargo to avoid damaging the vehicle or compromising
                safety.
                 \1325\ This weight includes the vehicle and a trailer attached
                to the vehicle, if used.
                 \1326\ For the EPA two-cycle regulatory test on a dynamometer,
                an additional weight of 300 lbs. is added to the vehicle curb
                weight. This additional 300 lbs. represents the weight of the
                driver, passenger, and luggage. Depending on the final test weight
                of the vehicle (vehicle curb weight plus 300 lbs.), a test weight
                category is identified using the table published by EPA according to
                40 CFR 1066.805. This test weight category is called ``Equivalent
                Test Weight'' (ETW).
                ---------------------------------------------------------------------------
                 The vehicle curb weight is the most commonly used measurement when
                comparing vehicles. A vehicle's curb weight is the weight of the
                vehicle including fluids, but without a driver, passengers, and cargo.
                 A vehicle's glider weight, which is vehicle curb weight minus the
                powertrain weight, is used to track the potential opportunities for
                weight reduction not including the powertrain. A glider's subsystems
                may consist of the vehicle body, chassis, interior, steering,
                electrical accessory, brake, and wheels systems. However, as noted in
                the PRIA, the definition of a glider may vary from study to study (or
                even simulation to simulation).
                 Each of the subsystems presents an opportunity for weight
                reduction; however, some weight reduction is dependent on the weight
                reduction of other subsystems. The agencies characterize mass reduction
                as either primary mass reduction or secondary mass reduction. Primary
                mass reduction involves reducing mass of components that can occur
                independent from the mass of other components. For example, reducing
                the mass of a hood (e.g., replacing a steel hood with an aluminum hood)
                or reducing the mass of a seat are examples of primary mass reduction
                because each can be implemented independently. Other components and
                systems that may contribute to primary mass reduction include the
                vehicle body, chassis, and interior components.
                 When significant primary mass reduction occurs, other components
                designed based on the mass of primary components may be redesigned as
                well. An example of a subsystem where secondary mass reduction can be
                applied is the brake system. If the mass of primary components is
                reduced sufficiently, the resulting lighter weight vehicle could safely
                maintain braking performance and attributes with a lighter weight brake
                system. Other examples of components where secondary mass reduction can
                be applied are wheels and tires.
                 For this analysis, the agencies consider mass reduction
                opportunities from the glider subsystems of a vehicle first, and then
                consider associated opportunities to downsize the powertrain, which are
                accounted for separately.\1327\ As explained later, in the Autonomie
                simulations, the glider system includes both primary and secondary
                systems from which a percentage of mass is reduced for different glider
                weight reduction levels; specifically, the glider includes the body,
                chassis, interior, electrical accessories, steering, brakes and wheels.
                The model sizes the powertrain based on the glider weight and the mass
                of some of the powertrain components in an iterative process. The mass
                of the powertrain depends on the powertrain size. Therefore, the weight
                of the glider impacts the weight of the powertrain.\1328\ See Section
                VI.B.3.a)(3) Vehicle models for Autonomie and Section VI.B.3.a)(4) How
                Autonomie Sizes Powertrains for Full Vehicle Simulation for more
                details.
                ---------------------------------------------------------------------------
                 \1327\ When the mass of the vehicle is reduced by an appropriate
                amount, the engine may be downsized to maintain performance. See
                Section VI.B.3.a)(5) Maintaining Vehicle Attributes] and Section
                VI.B.3.a)(6) Performance Neutrality for more details.
                 \1328\ Since powertrains are sized based on the glider weight
                for the analysis, glider weight reduction beyond a threshold amount
                during a redesign will lead to re-sizing of the powertrain. For the
                analysis, the glider was used as a base for the application of any
                type of powertrain. A conventional powertrain consists of an engine,
                transmission, exhaust system, fuel tank, radiator and associated
                components. A hybrid powertrain also includes a battery pack,
                electric motor(s), generator, high voltage wiring harness, high
                voltage connectors, inverter, battery management system(s), battery
                pack thermal system, and electric motor thermal system.
                ---------------------------------------------------------------------------
                 The agencies use glider weight to apply non-powertrain mass
                reduction technology, and use Autonomie simulations to determine the
                size of the powertrain and corresponding powertrain weight for the
                respective glider weight. The combination of glider weight (after mass
                reduction) and re-sized powertrain weight equal the vehicle curb
                weight. See Section VI.C.4.d)(1) glider mass and mass reduction
                subsection below for more detail on glider mass and glider mass
                reduction.
                (a) Mass Reduction in the CAFE Model
                 Several studies have explored the amount of vehicle mass reduction
                that is feasible in the rulemaking timeframe and the cost for that mass
                reduction.1329 1330 1331 1332 Those studies were sponsored
                by the agencies, CARB, ICCT, the automotive industry, and material
                manufacturers, and are discussed in Section VI.C.4.e)(1), below. All of
                the studies showed that the maximum feasible amount of mass reduction
                that can be applied in the rulemaking timeframe is around 20 percent of
                a baseline vehicle's curb weight. The National Academies of Sciences
                similarly concluded, based on some of these same studies along with
                other information, that it is feasible to
                [[Page 24526]]
                reduce up to 20 percent of the mass of the vehicle.\1333\
                ---------------------------------------------------------------------------
                 \1329\ DOT HS 811 692: Investigation of Opportunities for
                Lightweight Vehicles Using Advanced Plastics and Composites.
                 \1330\ A Review of the Safety of Reduced Weight Passenger Cars
                and Light Duty Trucks by Michigan Manufacturing Technology Center,
                October 2018.
                 \1331\ ATG Silverado Body Light weighting Study, Aluminum
                Technology Group, January 2017.
                 \1332\ 2013 NanoSteel Intensive Body-In-White, EDAG and
                NanoSteel Company Inc.
                 \1333\ Cost, Effectiveness and Deployment of Fuel Economy
                Technologies for Light-Duty Vehicles, National Academy of Sciences,
                2015, at 212 .
                ---------------------------------------------------------------------------
                 As discussed in Section VI.C.4.e), the mass reduction studies show
                that the cost for mass reduction increases progressively as the amount
                of mass reduction increases. In other words, lower levels of mass
                reduction are more cost effective than higher levels of mass reduction.
                As in past rulemakings, the agencies have considered multiple levels of
                mass reduction to provide options similar to what manufacturers could
                consider at vehicle redesigns.
                 For the NPRM, the agencies included five levels of mass reduction
                with a maximum of 20 percent glider mass reduction, corresponding to 10
                percent curb mass reduction, using the assumption that the glider was
                50 percent of curb weight. Table VI-117 shows the glider and curb
                weight mass reduction levels for each level of mass reduction
                considered in the NPRM analysis.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.250
                 The agencies received a number of comments suggesting that the
                amount of mass reduction allowed should be 20 percent of curb weight,
                as well as suggestions that the agencies should assume the glider
                represents 75 percent of the vehicle's curb weight. These comments are
                addressed in more detail in Section VI.C.4.d) below, but some
                understanding of how the glider share assumption affects the maximum
                amount of mass reduction allowed in the CAFE model is required here.
                 Several commenters stated that the agencies should allow further
                levels of mass reduction technology improvements in the CAFE model. For
                example, ICCT commented that the agencies must revise their treatment
                of mass reduction because studies have demonstrated that at least 20%
                mass reduction of curb weight is available for adoption across vehicle
                classes by 2025. \1334\ ICCT stated that based on these studies, the
                agencies must increase the maximum available mass reduction potential
                levels to include up to 20% and 25% mass reduction of curb weight, as
                the industry ``will cost-effectively deploy at least 15% vehicle curb
                mass reduction in the 2025 timeframe at net zero cost.'' ICCT caveated
                that amount of mass reduction seems less likely in smaller cars, which
                typically employ lower levels of mass reduction, so a constraint of 7.5
                percent mass reduction as was applied in the Draft TAR would be
                appropriate for those vehicles.
                ---------------------------------------------------------------------------
                 \1334\ NHTSA-2018-0067-11741. ICCT also alleged that the
                agencies intentionally disregarded the studies that presented this
                result; those comments are discussed in Section VI.C.4.e) Mass
                Reduction Costs, below.
                ---------------------------------------------------------------------------
                 ICCT also commented that there were numerous material improvements
                in development that were not considered in the rule, including but not
                limited to higher strength aluminum, improved joining techniques for
                mixed materials, third-generation steels with higher strength and
                enhanced ductility, a new generation of ultra-high strength steel cast
                components, and metal/plastic hybrid components, among other
                technologies mentioned in ICCT's working paper on light-weighting.
                 In assessing these comments, the agencies reconsidered the mass
                reduction studies and available reports and agreed that additional
                levels of mass reduction should be available for the final rule
                analysis. In response to comments, the agencies made two adjustments to
                allow higher levels of mass reduction in the analysis. First, as
                explained in Section VI.C.4.d)(1), below, the agencies increased the
                glider percentage of vehicle curb weight used for the analysis from 50
                percent to 71 percent. As explained in that section, increasing the
                glider percentage also increases the amount of curb weight reduction
                for all levels of mass reduction. Second, the agencies created another
                level of mass reduction (MR6) in the CAFE model, which represents a
                significant application of carbon fiber in the vehicle to achieve
                nearly 30 percent reduction in glider weight (which approximately
                translates to 20 percent reduction in vehicle curb weight). For
                example, incorporating a carbon fiber tub,\1335\ or a carbon fiber
                monocoque with aluminum sub frame in the front and back,\1336\ or a
                carbon fiber splitter and carbon fiber wheels,\1337\ allows for greater
                levels of mass reduction, albeit at a very high cost. These
                technologies are not ready for high volume production vehicles.
                ---------------------------------------------------------------------------
                 \1335\ The BMW i3 and BMW i8, which are about 20 percent lighter
                than an average MY 2017 vehicle, use a carbon fiber tub.
                 \1336\ The Alfa Romeo 4c/4c Spider, which is about 20 percent
                lighter than an average MY 2017 vehicle, uses this design.
                 \1337\ The Ford Shelby GT350R which is about 20 percent lighter
                than an average MY 2017 vehicle, uses this design.
                ---------------------------------------------------------------------------
                 Table VI-118 shows the levels of mass reduction technology
                available for application in the final rule analysis, with the
                associated glider weight percentage reduction and the percentage curb
                weight reductions for passenger cars and light trucks. As discussed in
                Section VI.C.4.c) below, the agencies declined to place a constraint on
                the amount of mass reduction technology that smaller cars could adopt.
                [[Page 24527]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.251
                 The agencies continue to believe the maximum feasible mass
                reduction levels identified in comprehensive design studies, such as
                those discussed in Section VI.C.4 Mass Reduction Costs are the most
                reliable for projecting the maximum amount of mass reduction in the
                rulemaking timeframe, and therefore have determined MR6 is the highest
                level that should be used for the final rule analysis. While the
                information provided by ICCT on newer materials and manufacturing and
                assembly methodology is interesting and relevant, this information, by
                itself, is insufficient to assess the amount of mass reduction that is
                feasible and the cost for the mass reduction. ICCT did not provide a
                comprehensive analysis showing a design concept that maintains vehicle
                attributes and performance, such as noise, vibration and harshness,
                stiffness, handling, compliance with NHTSA safety standards, good
                performance under NHTSA NCAP and IIHS rating systems, and other
                criteria. The various studies in Section VI.C.4.e) Mass Reduction
                considered those factors to varying degrees. Without that rigorous
                analysis, the actual amount of mass reduction that could be enabled
                through the use of those materials and methods described by ICCT, and
                the cost of achieving that mass reduction, would be highly speculative.
                As explained in Section VI.C.4.e) Mass Reduction below, the agencies
                determined the NHTSA-sponsored design studies remain a reasonable basis
                for estimating a feasible amount of mass reduction and the cost for
                mass reduction in the rulemaking timeframe, because those studies
                considered a wide range of materials (including advanced materials) and
                design solutions.
                (b) Analysis Fleet Mass Reduction Assignments
                 The agencies included an estimated level of mass reduction
                technology for each vehicle model in the MY 2016 analysis fleet for the
                NPRM, and have updated the estimates for the MY 2017 analysis fleet for
                the final rule analysis. The methodology used to provide each vehicle
                model an appropriate initial mass reduction technology level for
                further improvements was described in detail in the Draft TAR (when
                NHTSA first employed this methodology), in the PRIA accompanying the
                NPRM, and is reproduced here, in part, to provide additional context to
                the agencies' responses to comments on analysis fleet mass reduction
                assignments. The methodology used in this final rule was unchanged from
                the NPRM.
                 For the Draft TAR, NHTSA/Volpe Center staff developed regression
                models to estimate curb weights based on other observable attributes.
                With regression outputs in hand, Volpe evaluated the distribution of
                vehicles in the analysis fleet. In addition, vehicle platforms were
                evaluated based on the sales-weighted residual of actual vehicle curb
                weights versus predicted vehicle curb weights. Based on the actual curb
                weights relative to predicted curb weights, platforms (and the
                subsequent vehicles) were assigned a baseline mass reduction level (MR0
                through MR6). For the NPRM and final rule analysis, the agencies
                followed a similar procedure for the MY 2016 and MY 2017 analysis
                fleets.
                 To develop the curb weight regressions, the agencies grouped
                vehicles into three separate body design categories for analysis: 3-
                Box, 2-Box, and Pick-up.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.252
                 For the NPRM and final rule analysis, the agencies retained the MY
                2015 regressions for 3-Box and 2-Box vehicles, however the pickup
                category regression was updated in response to comments on the Draft
                TAR. The
                [[Page 24528]]
                agencies trained a new regression with EPA MY 2014 data and added pick-
                up bed length as an independent variable. As a result of stepping back
                to MY 2014 data for the pick-up regression, the training data did not
                include the all-aluminum body Ford F-150 in the calculation of the
                baseline. The advanced F-150 in the MY 2015 pick-up regression
                meaningfully affected Draft TAR regression statistics because the F-150
                accounted for a large portion of observations in the analysis fleet,
                and the F-150 included advanced weight savings technology.
                 The agencies leveraged many documented variables in the analysis
                fleet as independent variables in the regressions. Continuous
                independent variables included footprint (wheelbase x track width) and
                powertrain peak power. Binary independent variables included strong HEV
                (yes or no), PHEV (yes or no), BEV or FCV (yes or no), all-wheel drive
                (yes or no), rear-wheel drive (yes or no), and convertible (yes or no).
                In addition, for PHEV and BEV/FCV vehicles, the capacity of the battery
                pack was included in the regression as a continuous independent
                variable. In some body design categories, the analysis fleet did not
                cover the full spectrum of independent variables. For instance, in the
                pickup body style regression, there were no front-wheel drive vehicles
                in the analysis fleet, so the regression defaulted to all-wheel drive
                and left an independent variable for rear-wheel drive.
                 Furthermore, the agencies evaluated alternative regression
                variables in response to comments from vehicle manufacturers on the
                NHTSA/Volpe analysis in the Draft TAR.\1338\ The agencies evaluated
                regressions including overall dimensions of vehicles, such as height,
                width, and length, instead of and in addition to just wheelbase and
                track width. The experimental regression variables only marginally
                changed predicted curb weight residuals as a percentage of predicted
                curb weight, at an industry level and for most manufacturers. The
                results were not significantly different, and therefore the agencies
                opted not to add these variables to regressions or replace independent
                variables presented in Draft TAR with new variables.
                ---------------------------------------------------------------------------
                 \1338\ PRIA at 407.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.253
                [[Page 24529]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.254
                [[Page 24530]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.255
                BILLING CODE 4910-59-C
                 Each of the three regressions produced outputs effective for
                identifying vehicles with a significant amount of mass reduction
                technology in the analysis fleet. Many coefficients for independent
                variables provided clear insight into the average weight penalty for
                the utility feature. In some cases, like battery size, the relatively
                small sub-sample size and high collinearity with other variables
                confounded coefficients.
                 By design, no independent variable directly accounted for the
                degree of weight savings technology applied to the vehicle. Residuals
                of the regression captured weight reduction efforts and noise from
                other sources.
                 The agencies received many comments on the Draft TAR encouraging
                the use of observed technologies in each vehicle, and in each vehicle
                subsystem to assign levels of mass reduction technology. As a practical
                matter, the agencies cannot conduct a tear down study and detailed cost
                assessment for every vehicle in every model year. However, upon review
                of many vehicles and their subsystems, the agencies recognized a few
                vehicles with MR0 or MR1 assignments in NHTSA's analysis of the Draft
                TAR that contained some advanced weight savings technologies, yet these
                vehicles and their platforms still produced ordinary residuals.
                Engineers from industry confirmed important factors other than glider
                weight savings and the independent variables considered in the
                regressions may factor into the use of lightweight technologies. Such
                factors included the desire to lower the center of gravity of a
                vehicle, improve the vehicle weight distribution for handling, optimize
                noise-vibration-and-harshness, increase torsional rigidity of the
                platform, offset increased vehicle content, and many other factors. In
                addition, engineers highlighted the importance of sizing shared
                components for the most demanding applications on the vehicle platform;
                optimum weight savings for one platform application may not be suitable
                for all platform applications. For future analysis, the agencies will
                look for practical ways to improve the assessment of mass reduction
                content and the forecast of incremental mass reduction costs for each
                vehicle.
                 Figure VI-44 below shows results from the pickup truck regression
                on predicted curb weight versus actual curb weight. Points above the
                solid regression line represent vehicles heavier than predicted (with
                lower mass reduction technology levels); points below the solid
                regression line represent vehicles lighter than predicted (with higher
                mass reduction technology levels). The dashed lines in the Figure VI-44
                show the thresholds (5, 7.5, 10, 15, 20 and 28 percent of glider
                weight). Final rule glider weight assumption is 71 percent of vehicle
                curb weight.
                BILLING CODE 4910-59-P
                [[Page 24531]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.256
                BILLING CODE 4910-59-C
                 For points with actual curb weight below the predicted curb weight,
                the agencies used the residual as a percent of predicted weight to get
                a sense for the level of current mass reduction technology used in the
                vehicle. Notably, vehicles approaching -20% curb weight widely use
                advanced composites throughout major vehicle systems, and few examples
                exist in the MY 2016 fleet.\1339\
                ---------------------------------------------------------------------------
                 \1339\ This evidence suggests that achieving a 20% curb weight
                reduction for a production vehicle with a baseline defined with this
                methodology is extremely challenging, and requires very advanced
                materials and disciplined design.
                ---------------------------------------------------------------------------
                 Generally, residuals of regressions as a percent of predicted
                weight appropriately stratified vehicles by mass reduction level. Most
                vehicles showed near zero residuals or had actual curb weights close to
                the predicted curb weight. Few vehicles in the analysis fleet were
                identified with the highest levels of mass reduction. Most vehicles
                with the largest negative residuals have demonstrably adopted advanced
                weight savings technologies at the most expensive end of the cost
                curve.
                 To validate the residuals, the agencies estimated the mass
                reduction technology level for several vehicle models in the analysis
                fleet and compared those estimates to the numerical results from the
                regression analysis. To estimate the mass reduction technology level
                for the selected vehicles, the agencies conducted an in-depth review of
                available information on the materials, design, and last redesign year
                for those vehicle models, and compared that information with the
                designs and materials used in the mass reduction feasibility and cost
                studies summarized in Section VI.C.4.e), below.
                [[Page 24532]]
                That comparison showed good agreement with the technology levels from
                the regression analysis.
                 The agencies believe the regression methodology is a technically
                sound methodology for estimating mass reduction levels in the analysis
                fleet.
                 As part of their comments stating the NPRM modeling reflected
                reality better than the Draft TAR and Proposed Determination analyses,
                Toyota commented broadly that the MY 2016 baseline fleet used in the
                NPRM encompassed powertrain and tractive energy (including mass
                reduction) improvements more representative of vehicles on the road
                today.\1340\ Toyota noted that the 2016 baseline fleet generally
                contained higher levels of technology compared to the MY 2014 and MY
                2015 baseline fleets, and included a comparison of its initial fleet
                mass reduction assignments in the Draft TAR and the NPRM. Toyota showed
                how moving further up the technology tree (e.g., starting with a
                baseline that includes higher levels of technology) for certain
                pathways such as mass reduction increased costs exponentially. Toyota
                stated that the NPRM underestimated mass reduction cost values.
                ---------------------------------------------------------------------------
                 \1340\ NHTSA-2018-0067-12098.
                ---------------------------------------------------------------------------
                 While a more specific discussion of costs is located in Section
                VI.C.4.e), the agencies agree with Toyota's assessment that the costs
                for mass reduction technology increase exponentially as progressively
                higher levels of mass reduction are incorporated. Having an accurate
                assessment of baseline technology levels ensures that the subsequent
                application of technology and its associated costs is correctly
                accounted for.
                 C.A.R produced a report in response to the Draft TAR that generally
                agreed with the regression methodology of using observed vehicle
                attributes for estimating mass reduction levels, as opposed to
                comparing vehicle curb weight from a newer model year to a previous
                generation of the same vehicle, pointing to several of the limitations
                discussed above.\1341\
                ---------------------------------------------------------------------------
                 \1341\ EPA Mass Reduction Analysis--Observations and
                Recommendations, Center for Automotive Research, October 2017 (page
                15), available at https://www.cargroup.org/wp-content/uploads/2017/10/EPA-MR-Analysis-Critique_Oct-5_final.pdf.
                ---------------------------------------------------------------------------
                 Both ICCT and H-D Systems commented on the methodology for
                identifying mass reduction technology levels in the analysis fleet,
                with ICCT broadly stating that by placing additional mass reduction
                technology in the baseline, the agencies artificially removed ``the
                most cost-effective lightweighting from future use, which incorrectly
                increases the costs of all subsequent mass-reduction in the compliance
                modeling.'' \1342\
                ---------------------------------------------------------------------------
                 \1342\ NHTSA-2018-0067-11741 full comments.
                ---------------------------------------------------------------------------
                 ICCT claimed that the agencies unjustifiably increased the amount
                of vehicle mass reduction technology present in the 2016 baseline fleet
                from the 2015 baseline used in the Draft TAR, stating that the 2015
                Draft TAR fleet had 26 percent of vehicles sold with some level of mass
                reduction applied (MR1 or a higher level), whereas the 2016 NPRM fleet
                had 47 percent of vehicles sold with some level of mass reduction
                applied. In addition to faulting the agencies for not acknowledging the
                change and not attempting to justify it, ICCT stated that the 2016
                analysis fleet mass reduction assignments were overstated, as ``it
                appears that the agencies have applied mass reduction technology to
                vehicles in the model that did not have mass reduction applied in the
                real world.'' ICCT stated that the effect of this change was to
                ``render[] unavailable mass reduction technologies for these vehicles
                in the model,'' causing the model to select less cost-effective
                technologies instead and driving the modeled compliance costs higher.
                 ICCT argued that to substantiate the changes made to the baseline
                fleet mass reduction assignments, the agencies must show data on how
                these improvements are evident in the fleet and to quantify and include
                their realized benefits in the analysis, including a detailed and
                justified explanation of all mass reduction technologies deemed already
                to have been applied to the MY 2016 analysis fleet. More specifically,
                ICCT stated that the agencies ``must clearly and precisely share their
                estimated percent (and absolute pounds) mass reduction amount for each
                vehicle make and model in the baseline fleet (rather than simply
                showing binned categories), and their technical justification for each
                value,'' and ``[t]o not do so obscures the agencies' new methods and
                data sources from public view, rendering their lightweighting
                calculations a black box.''
                 In addition, ICCT recommended that the agencies conduct two
                sensitivity analyses, one assuming that every baseline make and model
                has not yet applied any lightweighting (setting the baseline to 0% mass
                reduction), and one assuming that each vehicle model has applied Draft
                TAR baseline mass reduction assignments, to demonstrate how much the
                agencies' decision to load up more baseline technology affects the
                compliance scenarios.
                 ICCT concluded that because the changes in baseline mass reduction
                assignments from prior analyses to the NPRM ``are opaquely buried in
                the agencies' datafiles and unexplained, we believe the agencies have
                to reissue a new regulatory analysis and allow an additional comment
                period for review of their methods and analysis.''
                 To address ICCT's comment, it is important to understand the mass
                reduction baseline technology assignment methodology previously used by
                EPA in the Draft TAR and Proposed Determination.\1343\ As stated in the
                Draft TAR, the curb weight of each vehicle model in the MY 2008
                analysis fleet (used for the 2012 rulemaking to establish MYs 2017-2025
                standards) was assumed to be at a baseline MR0 level. The mass
                reduction technology level in the MY 2014 analysis fleet was determined
                by comparing the curb weight of the MY 2014 vehicle to the most similar
                vehicle in the MY 2008 analysis fleet.\1344\ The curb weight of the
                newer model year vehicle was adjusted to account for changes in the
                vehicle footprint and changes in mass due to added safety technology.
                If a vehicle did not have a previous generation vehicle, then the sales
                weighted average percent mass reduction over the manufacturer's name
                plate product line was used to represent the expectation of mass
                reduction technology available within the vehicle.
                ---------------------------------------------------------------------------
                 \1343\ Draft TAR at 5-395.
                 \1344\ Draft TAR at 5-395.
                ---------------------------------------------------------------------------
                 EPA listed some limitations to this methodology in the Draft
                TAR,\1345\ and others are also addressed here. First, assuming that
                every vehicle started with MR0 technology did not account for the
                actual varying levels of mass reduction technology that existed in the
                MY 2008 fleet. Second, for each vehicle model, there was no accounting
                for the mass associated with different powertrain configurations. This
                was particularly problematic because the method did not account for
                light weight technology already available in the vehicle structure to
                counter the increased mass associated with more advanced powertrains,
                such as HEV, PHEV, and EV technologies.\1346\ Third, there was no
                sales-weight accounting for the various configurations in estimating
                the vehicle model mass reduction technology level, meaning that if a
                high-sales-volume vehicle employed significant mass reduction
                technology, that vehicle was not credited as such in the analysis
                [[Page 24533]]
                fleet. Fourth, there was no accounting for mass increases due to the
                addition of future regulatory requirements like potential safety
                regulations. Fifth, there was no accounting for mass associated with
                changes in vehicle attributes and utility, such as the addition of
                infotainment systems and crash avoidance technologies. These
                limitations all individually had the effect of overestimating mass
                reduction technology effectiveness and undercounting mass reduction
                technology costs across the fleet, and accordingly their combined
                effect was significant. The lack of controls for these items introduced
                errors into the mass reduction technology level effectiveness
                estimates.
                ---------------------------------------------------------------------------
                 \1345\ Draft TAR at 5-395.
                 \1346\ PRIA at 413.
                ---------------------------------------------------------------------------
                 After considering the comments, the agencies determined the use of
                the regression method, based on observable attributes, is the best
                available methodology to provide a reasonable estimate of mass
                reduction technology for the analysis fleet. The agencies believe that,
                contrary to ICCT's assertion, the regression methodology used in the
                NHTSA Draft TAR, NPRM, and final rule analyses provides a more
                transparent method for calculating baseline mass reduction technology
                assignments. The methodology was fully explained in the Draft TAR and
                PRIA, and avoided the limitations identified by EPA by using data from
                the analysis fleet, and not requiring the use of or assumptions about
                the exact mass reduction levels of vehicles in a prior model year
                fleet. In addition, the regression accounted for differences in
                powertrains between trim levels, including non-ICE powertrains by
                accounting for these factors in the regression analysis.
                 Also, because manufacturers generally apply mass reduction
                technology at a vehicle platform level (i.e., using the same components
                across multiple vehicle models that share a common platform) to
                leverage economies of scale and to manage component and manufacturing
                complexity, conducting the regression analysis at the platform level
                leads to more accurate estimates for the real-world vehicle platform
                mass reduction levels. The platform approach also addresses the impact
                of potential weight variations that might exist for specific vehicle
                models, as all of the individual vehicle models are aggregated into the
                platform group, and are effectively averaged using sales weighting,
                which minimizes the impact of any outlier vehicle configurations.
                 The agencies also disagree that the changes in baseline mass
                reduction assignments were unexplained. The PRIA discussed reasons that
                baseline mass reduction assignments differed from prior analyses,
                including that, ``[s]ince the Draft TAR, many platforms have not been
                redesigned, but in some cases the sales-weighted residuals for
                carryover platforms have moved. In the case of 2-Box and 3-Box
                vehicles, the analysis attributes such changes to differences in sales
                mix year-over-year and other updates to reported curb weights and
                platform designations. In the case of platforms with pick-up trucks,
                the analysis updated the pick-up regression since the Draft TAR, so
                that may be a contributing factor.'' \1347\
                ---------------------------------------------------------------------------
                 \1347\ PRIA at 424.
                ---------------------------------------------------------------------------
                 To the extent that the NPRM glider weight assumption impacted the
                NPRM MY 2016 analysis fleet baseline mass reduction assignment values,
                the agencies presented a table in the PRIA showing how different glider
                weight assumptions impacted mass reduction technology levels for the
                analysis fleet.\1348\ The following Table VI-123 recreates that table
                in part, with updates based on the glider weight values used for the
                final rule.
                ---------------------------------------------------------------------------
                 \1348\ PRIA at 422.
                ---------------------------------------------------------------------------
                 For example, from the regression analysis, the Ford F-150 has a
                predicted curb weight (residual) of 12.4 percent of the actual curb
                weight. If the glider weight assumption is 50 percent of the vehicle
                curb weight (like in NPRM), then the agencies would assign MR5 as an
                initial mass reduction assignment in the analysis fleet. With this high
                level of mass reduction technology already applied, the opportunity for
                further mass reduction would be limited. However, if the glider weight
                is assumed to be 71 percent of the vehicle curb weight, then Ford F-150
                would be assigned MR4, and would have an opportunity to apply another
                level of mass reduction albeit at higher cost.
                [[Page 24534]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.257
                 The agencies also disagree that the amount of vehicle mass
                reduction technology present in the 2016 baseline fleet was
                ``unjustifiably increased'' from the 2015 baseline used in the Draft
                TAR. Table VI-124 shows the percent mass reduction technology used in
                Draft TAR, NPRM, and in final rule. It is clear from the table below
                that total percentage of MY 2016 vehicle fleet used in the NPRM had
                nearly the same level of some mass reduction technology applied
                compared to the Draft TAR. Similar to ICCT's observations, 28 percent
                of the MY 2015 vehicle fleet used in the Draft TAR had some level of
                mass reduction technology (MR1 to MR5) and 26 percent of MY 2016
                vehicle fleet had some mass reduction technology applied. Since the
                agencies assumed a reduced glider share in the NPRM, the percentage of
                vehicles assigned a MR4 or MR5 technology level increased compared to
                Draft TAR. In addition, for this final rule, the agencies observed that
                many of the vehicles in the MY 2017 fleet had been redesigned, which
                provided the opportunity to incorporate additional mass reduction
                technologies.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.258
                 The agencies considered a sensitivity case that assumed no mass
                reduction, rolling resistance, or aerodynamic improvements had been
                made to the MY 2017 fleet (i.e., setting all vehicle road levels to
                zero--MRO, AERO and
                [[Page 24535]]
                ROLL0), in response to ICCT's comment. While this is an unrealistic
                characterization of the initial fleet, the agencies conducted a
                sensitivity analysis to understand any affect it may have on technology
                penetration along other paths (e.g., engine and hybrid technology).
                Under the CAFE program, the sensitivity analysis shows a slight
                decrease in reliance on engine technologies (HCR engines, turbocharge
                engines, and engines utilizing cylinder deactivation) and hybridization
                (strong hybrids and plug-in hybrids) in the baseline (relative to the
                central analysis). The consequence of this shift to reliance on lower-
                level road load technologies is a reduction in compliance cost in the
                baseline of about $300 per vehicle (in MY 2026). As a result, cost
                savings in the preferred alternative are reduced by about $200 per
                vehicle. Under the CO2 program, the general trend in
                technology shift is less dramatic (though the change in BEVs is larger)
                than the CAFE results. The cost change is also comparable, but slightly
                smaller ($200 per vehicle in the baseline) than the CAFE program
                results. Cost savings under the preferred alternative are further
                reduced by about $100. With the lower technology costs in all cases,
                the consumer payback periods decreased as well. These results are
                consistent with the approach taken by manufacturers who have already
                deployed many of the low-level road load reduction opportunities to
                improve fuel economy.
                 Second, as discussed above, EPA's Draft TAR baseline mass reduction
                assignments had identified limitations that the regression methodology
                has addressed. Moreover, as discussed above, the regression methodology
                was updated from the Draft TAR to characterize data better on pickup
                trucks. The agencies do not believe that conducting sensitivity
                analyses using these outdated or limited assumptions would be useful
                for this final rule.
                 More narrowly, HDS commented that while the regression coefficients
                between 2-box and 3-box vehicles for footprint seemed consistent, the
                regression coefficients for horsepower between the 2-box and 3-box
                vehicles seemed incorrect because both types of vehicles use similar
                engines.\1349\ HDS stated that ``[c]ollinearity between footprint and
                HP or other effects caused by having electric vehicles (with electric
                motor HP ratings) in the regression data is the probable cause of these
                inconsistent coefficients for HP, but this cannot be confirmed without
                access to the same database used by NHTSA.'' HDS concluded that
                ``[r]evisions to the regression could have a significant effect on the
                baseline assignment of vehicles, as the current assignment for vehicles
                like the 2016 Mazda MX5 as having the highest level of weight reduction
                technology (MR5) and the 2016 Chevy Malibu as having MR4 technology
                appear incorrect as their curb weights are comparable to other similar
                MY 2016 vehicles in their respective class.''
                ---------------------------------------------------------------------------
                 \1349\ H-D Systems, NHTSA-2018-0067-11985.
                ---------------------------------------------------------------------------
                 While many of the vehicles share same the same powertrain for
                passenger cars and SUVs or for cars and pickup trucks, the utility and
                functionality of the vehicle in SUVs and pickup trucks (2-box) is
                different than passenger cars (3-box). The presence of additional
                structure for towing or higher capacity towing, rear cross member,
                higher capacity suspension, and other differences, enable SUVs and
                pickup trucks to have towing and heavier payload capability. For
                example, Ford uses the nearly similar displacement and horsepower
                engines in Mustang Ecoboost Coupe and in F150 2WD XL, Regular Cab, Long
                Box. However, the curb weight for the pickup truck is higher than the
                Mustang. Directionally, this supports that the 2-box weight per
                horsepower coefficient should be greater than the 3-box coefficient,
                just as it is in the for the regression. The coefficient for passenger
                cars and SUVs has not changed since the Draft TAR (based on MY2015
                vehicle fleet). Based on the comments to Draft TAR, for the NPRM, a new
                set of coefficients were generated for pickups using the MY 2014
                vehicle fleet. This was done so that coefficients were not skewed due
                to presence of the aluminum intensive Ford F150 pickup truck. Hence,
                the agencies believe the coefficients used in the regression analysis
                are directionally correct and disagree with HDS's assertion. The
                agencies further note that HDS did not suggest any alternate
                methodology or specific coefficients to use in the regression analysis.
                (c) Mass Reduction Technology Adoption Features
                 The agencies described in the NPRM that given the degree of
                commonality among the vehicle models built on a single platform,
                manufacturers do not have complete freedom to apply unique technologies
                to each vehicle that shares the platform: while some technologies
                (e.g., low rolling resistance tires) are very nearly ``bolt-on''
                technologies, others involve substantial changes to the structure and
                design of the vehicle, and therefore often necessarily affect all of
                the vehicle models that share that platform. In most cases, mass
                reduction technologies are applied to platform level components and
                therefore the same design and components are used on all of the vehicle
                models that share the platform.
                 As discussed in Section Analysis Fleet, above, each vehicle in the
                analysis fleet is associated with a specific platform. Similar to the
                application of engine and transmission technologies, the CAFE model
                defines a platform ``leader'' as the vehicle variant of a given
                platform that has the highest level of observed mass reduction present
                in the analysis fleet. If there is a tie, the CAFE model begins mass
                reduction technology on the vehicle with the highest sales in model
                year 2017. If there remains a tie, the model begins by choosing the
                vehicle with the highest Manufacturer Suggested Retail Price (MSRP) in
                MY 2017. As the model applies technologies, it effectively levels up
                all variants on a platform to the highest level of mass reduction
                technology on the platform. So, if the platform leader is already at
                MR3 in MY 2017, and a ``follower'' starts at MR0 in MY 2017, the
                follower will get MR3 at its next redesign (unless the leader is
                redesigned again before that time, and further increases the mass
                reduction level associated with that platform, then the follower would
                receive the new mass reduction level).
                 Important for analysis fleet mass reduction assignments (discussed
                above), and for understanding adoption features as well, is the
                agencies' handling of vehicles that traditionally operated on the same
                platform but had a mix of old and new platforms in production when the
                analysis fleet was created. As described in the PRIA, the Honda Civic
                and Honda CR-V traditionally share the same platform. In MY 2016, Honda
                redesigned the Civic and updated the platform to include many mass
                reduction technologies. Also in MY 2016, Honda continued to build the
                CR-V on the previous generation platform--a platform that did not
                include many of the mass reduction technologies on the all new MY 2016
                Civic. In MY 2017, Honda launched the new CR-V that incorporated
                changes to the Civic platform, and the Civic and CR-V again shared the
                same platform with common mass reduction technologies. The NPRM and
                final rule analyses treat the old and new platforms separately to
                assign technology levels in the baseline, and the CAFE model brings
                vehicles on the old platform up to the level of mass reduction
                technology on the new shared platform at the first available redesign
                year.
                 Furthermore, as stated in the NPRM and PRIA, unlike the analysis
                presented in the Draft TAR that restricted high
                [[Page 24536]]
                levels of mass reduction for cars to show a safety neutral pathway to
                compliance, the NPRM analysis did not artificially restrict mass
                reduction to achieve a safety neutral outcome.\1350\ The NPRM CAFE
                model considered MR0 through MR5 for all vehicles at redesign, and
                similarly for the final rule, the CAFE model considers MR0 through MR6
                for all vehicles at redesign.
                ---------------------------------------------------------------------------
                 \1350\ PRIA at 494.
                ---------------------------------------------------------------------------
                 Ford commented in support of the removal of ``previously applied
                modeling rules that disallowed the mass reduction technology pathway
                for certain vehicle classes since this restriction was not supported by
                an adequate technical justification.'' \1351\ ICCT commented that a
                constraint of 7.5 percent mass reduction to smaller cars, as was
                applied in the Draft TAR, would be appropriate for those vehicles.
                ---------------------------------------------------------------------------
                 \1351\ NHTSA-2018-0067-11928.
                ---------------------------------------------------------------------------
                 The agencies considered ICCT's comment that mass reduction on small
                passenger cars should be limited to 7.5 percent, and Ford's comment
                supporting the removal of ``previously applied modeling rules that
                disallowed the mass reduction technology pathway for certain vehicle
                classes.'' Neither CAFE standards nor this analysis mandate mass
                reduction, or mandate that mass reduction occur in any specific manner.
                The mass reduction cost subsection below shows mass reduction is a
                cost-effective technology for improving fuel economy and CO2
                emissions. The steel, aluminum, plastics, composite, and other material
                industries are developing new materials and manufacturing equipment and
                facilities to produce those materials. In addition, suppliers and
                manufacturers are optimizing designs to maintain or improve functional
                performance with lower mass. Manufacturers have stated that they will
                continue to reduce vehicle mass to meet more stringent standards, and
                therefore, this expectation is incorporated into the modeling analysis
                supporting the standards to: (1) Determine capabilities of
                manufacturers; and (2) predict costs and fuel consumption effects of
                CAFE standards. The CAFE and CO2 rulemakings in 2012, and
                the Draft TAR and EPA Proposed Determination, imposed an artificial
                constraint that limited vehicle mass reduction in some small vehicles
                to achieve a desired safety-neutral outcome. For the current
                rulemaking, this artificial constraint is eliminated so the analysis
                reflects manufacturers' applying the most cost effective technologies
                to achieve compliance with the regulatory alternatives and the final
                standards; this approach allows mass reduction to be applied across the
                fleet. This approach is consistent with industry trends. To the extent
                that mass reduction is only cost-effective for the heaviest vehicles,
                the CAFE model would create the outcome predicted by commenters. In
                reality, however, mass reduction is a cost-effective means of improving
                fuel economy and does take place across vehicles of all sizes and
                weights. Accordingly, the model reflects that manufacturers may reduce
                vehicle mass--regardless of vehicle class--when doing so is cost
                effective.
                 The agencies have included one additional mass reduction level for
                the final rule in response to comments by ICCT and others, and to
                account for carbon fiber use in vehicles. For the NPRM, the maximum
                level of mass reduction was limited to 10 percent of a vehicle's curb
                weight, and that amount of mass reduction could be applied during the
                rulemaking timeframe. For the final rule, based on the current state of
                mass reduction technology and the application rate of different levels
                of mass reduction technologies, the agencies applied phase-in caps for
                MR5 and MR6 (15 percent and 20 percent reduction of a vehicle's curb
                weight, respectively). The agencies applied a phase-in cap for MR5
                level technology so that 15 percent of the vehicle fleet starting in
                2016 employed the technology, and the technology could be applied to
                100 percent of the fleet by MY 2022. This cap is consistent with the
                NHTSA lightweighting study that found that a 15 percent curb weight
                reduction for the fleet is possible within the rulemaking
                timeframe.\1352\ The agencies also applied a phase in cap for MR6
                technology so that one percent of the vehicle fleet starting in MY2016
                employed the technology, and the technology could be applied to 13
                percent of the fleet by MY2025. The agencies believe that this phase-in
                cap appropriately functions as a proxy for the cost and complexity
                currently required (and that likely will continue to be required until
                manufacturing process evolve) to produce carbon fiber components.
                Again, MR6 technology in this analysis reflects the use of a
                significant share of carbon fiber content, as seen through the BMW i3
                and Alfa Romeo 4c as discussed above.
                ---------------------------------------------------------------------------
                 \1352\ DOT HS 811 666: Mass Reduction for Light Duty Vehicles
                for Model Years 2017-2025: Figure 397 at page 356.
                ---------------------------------------------------------------------------
                (d) Mass Reduction Technology Effectiveness
                 As discussed in Section VI.B.3, Argonne developed a database of
                vehicle attributes and characteristics for each vehicle technology
                class that included over 100 different attributes like frontal area,
                drag coefficient, fuel tank weight, transmission housing weight,
                transmission clutch weight, hybrid vehicle component weights, and
                weights for components that comprise engines and electric machines,
                tire rolling resistance, transmission gear ratios, and final drive
                ratio. Argonne used these attributes to ``build'' each vehicle that it
                used for the effectiveness modeling and simulation. Important for
                precisely estimating the effectiveness of different levels of mass
                reduction is an accurate list of initial component weights that make up
                each vehicle subsystem, from which Autonomie considered potential mass
                reduction opportunities.
                 As stated above, glider weight, or the vehicle curb weight minus
                the powertrain weight, is used to determine the potential opportunities
                for weight reduction irrespective of the type of powertrain.\1353\ This
                is because weight reduction can vary depending on the type of
                powertrain. For example, an 8-speed transmission may weigh more than a
                6-speed transmission, and a basic engine without variable valve timing
                may weigh more than an advanced engine with variable valve timing.
                Autonomie simulations account for the weight of the powertrain system
                inherently as part of the analysis, and the powertrain mass accounting
                is separate from the application and accounting for mass reduction
                technology levels (MR0-MR6) that are applied to the glider in the
                simulations. Similarly, Autonomie also accounts for battery and motor
                mass used in hybrid and electric vehicles separately. This secondary
                mass reduction is discussed further, below.
                ---------------------------------------------------------------------------
                 \1353\ Depending on the powertrain combination, the total curb
                weight of the vehicle includes glider, engine, transmission and/or
                battery pack and motor(s).
                ---------------------------------------------------------------------------
                 Accordingly, in the Autonomie simulation, mass reduction technology
                is simulated as a percentage of mass removed from the specific
                subsystems that make up the glider, as defined for that set of
                simulations (including the non-powertrain secondary mass systems such
                as the brake system).
                (1) Glider Mass and Mass Reduction
                 Autonomie accounts for the mass of each subsystem that comprises
                the glider. For the NPRM, the glider subsystems included the vehicle
                body and the chassis, but did not include mass from subsystems such as
                the interior system, brake system, electrical accessory system, and
                steering and
                [[Page 24537]]
                wheel systems. The agencies described in the PRIA that based on
                advances in active and passive safety technologies that add some mass
                to the interior system, certain subsystems were not considered for
                potential light-weighting to maintain safety performance.\1354\ For the
                NPRM, the A2Mac1 database was used to estimate the average mass of each
                subsystem considered as part of the glider based on the subsystem
                assumptions, and to compute the average glider share of vehicle curb
                weight.\1355\ That analysis showed the glider accounted for 50 percent
                of the vehicle curb weight. The agencies solicited comment on whether
                systems or components beyond the vehicle body and chassis should be
                included as part of the glider, and also indicated that the glider
                weight assumption might increase for the final rule based on further
                research.
                ---------------------------------------------------------------------------
                 \1354\ PRIA at 411-12.
                 \1355\ The A2Mac1 database was used and this analysis was
                presented in ANL report docketed here: NHTSA-2018-0067-1490. The
                mass data in the database were obtained from vehicle teardown
                studies.
                ---------------------------------------------------------------------------
                 The agencies received several comments on the NPRM glider weight
                assumptions, with the overarching theme of the comments being that the
                NPRM did not include all systems and components that should be
                included, and if those systems and components were included, the glider
                share would be higher. Commenters also stated that the 50 percent
                glider share value used for the NPRM reduced the amount of mass
                reduction that could be applied to vehicles in the analysis.
                 UCS stated that representing the glider as a reduced fraction of
                the curb weight caused the agencies significantly to underestimate the
                potential for mass reduction. UCS noted that because mass reduction is
                applied at the glider level, reducing the share of the glider
                inherently caps the potential reduction in the curb weight, and this
                single change cut the potential improvement from mass reduction by one-
                third. Similarly, CARB stated that the updated glider weight assumption
                severely limited the effectiveness of mass reduction, as the most
                aggressive mass reduction category of 15 to 20 percent mass reduction
                can only reduce the vehicle curb weight by 10 percent.
                 UCS cited previous agency analyses and analyses from other
                organizations that stated the total potential for mass reduction by
                2025 is between 15.8 and 32 percent of curb weight, contrasted to the
                NPRM assumption of a maximum 10 percent reduction.\1356\ UCS also cited
                industry data which showed that the glider represented a higher share
                of vehicle curb weight than was assumed in the Draft TAR analysis, and
                both UCS and CARB cited to industry data from vehicles like the Ford F-
                150, which UCS stated was able to achieve the NPRM maximum achievable
                mass reduction through the deployment of aluminum alone.\1357\ UCS
                concluded that by capping the total potential for mass reduction at
                such a low level, the agencies artificially reduced the potential for
                the cost-effective technology, which increased the use of more
                expensive and more advanced technologies. CARB concluded that the
                agencies' 10 percent restriction means that real-world improvements
                that have already happened on production vehicles were not considered
                feasible in the NPRM analysis.
                ---------------------------------------------------------------------------
                 \1356\ NHTSA-2018-0067-12039 (citing Caffrey et al. 2013,
                Caffrey et al. 2015, Lotus 2012, NAS 2015, Singh et al. 2012, Singh
                et al. 2016, Singh et al. 2018).
                 \1357\ NHTSA-2018-0067-12039. See also NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 Several commenters also stated that the 50 percent glider weight
                assumption was unexplained and unjustified, and argued that the
                agencies' own studies showed that the glider weight percentage should
                range from 75-80 percent.\1358\ UCS stated that both the NHTSA-
                sponsored 2011 Honda Accord study, which showed the glider making up 79
                percent of the vehicle, and the NHTSA-sponsored 2014 Chevrolet
                Silverado study, which showed the glider making up 73.6 percent, showed
                values substantially higher than the 50 percent value, and were in line
                with the agencies' prior analyses.\1359\ As part of its comments that
                key assumptions about mass reduction changed from the Draft TAR without
                any supporting rationale, CARB stated that EPA had previously relied on
                four studies (two contracted for by EPA and two contracted for by
                NHTSA), and for the NPRM analysis the agencies only cited two of those
                studies.\1360\ Moreover, ICCT commented that the agencies' previous
                studies showed a glider fraction greater than 75 percent even with
                numerous safety features considered. Accordingly, ICCT stated that the
                agencies must specifically identify the ``safety components'' referred
                to in the NPRM and justify the limitations placed on light weighting in
                response. ICCT affirmatively concluded that the agencies must re-adopt
                the Draft TAR methodology in which glider mass is assumed to be 75
                percent of vehicle mass, or provide detailed justification and evidence
                supporting the new value of 50 percent.\1361\
                ---------------------------------------------------------------------------
                 \1358\ NHTSA-2018-0067-11985; NHTSA-2018-0067-12039; NHTSA-2018-
                0067-11873.
                 \1359\ NHTSA-2018-0067-12039.
                 \1360\ NHTSA-2018-0067-11873.
                 \1361\ NHTSA-2018-0067-11741.
                ---------------------------------------------------------------------------
                 The agencies carefully considered these comments and reexamined
                available data and information. The NHTSA-sponsored passenger car light
                weighting study showed a glider mass of 79 percent, and the NHTSA-
                sponsored light duty truck light weighting study showed a glider mass
                of 73.6 percent, and the 75 percent value used for the Draft TAR was a
                value between the values from these two studies. The agencies
                determined it would be more rigorous to consider data from a broader
                array of vehicles with various powertrain combinations and trim levels
                to assess the glider share for the final rule, considering that the
                vehicle fleet analyzed in this rule consists of over 2900 vehicle
                models.
                 The agencies examined glider weight data available in the A2Mac1
                database.\1362\ The A2Mac1 database tool is widely used by industry and
                academia to determine the bill of materials and mass of each component
                in the vehicle system.\1363\ The A2Mac1 database has been used by the
                agencies to inform past CAFE and CO2 rulemakings. The
                agencies analyzed a total of 147 MY 2014 to 2016 vehicles, covering 35
                vehicle brands with different powertrain options representing a wide
                array of vehicle classes to determine the glider weight for the final
                rule analysis.\1364\
                ---------------------------------------------------------------------------
                 \1362\ A2Mac1: Automotive Benchmarking. (n.d.). Retrieved from
                https://a2mac1.com.
                 \1363\ Bill of material (BOM) is a list of the raw materials,
                sub-assemblies, parts and quantities needed to manufacture an end
                product.
                 \1364\ The agencies presented this material for comments in the
                ANL report posted in the docket NHTSA-2018-0067-1490.
                ---------------------------------------------------------------------------
                 The agencies also considered that the NHTSA passenger car and light
                truck light-weighting studies examined mass reduction in the body,
                chassis, interior, brakes, steering, electrical accessory, and wheels
                subsystems and had developed costs for light weighted components in
                those subsystems. As a result, the agencies determined it is
                appropriate to include all of those subsystems as available for mass
                reduction as part of the glider. Therefore, all of these systems were
                included for the analysis of glider weight using the A2Mac1 database.
                Table VI-125 shows the average mass for each subsystem and the glider
                share for each of the vehicle classes for all powertrain combinations.
                BILLING CODE 4910-59-P
                [[Page 24538]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.259
                 This data was also compared with the glider weight measured in the
                NHTSA MY 2014 Chevrolet Silverado light weighting study,\1365\ and the
                glider weight data range was similar to the analysis results. Based on
                the comments and the agencies' updated assessment, the agencies have
                increased the glider weight assumption to 71 percent of the vehicle
                curb weight for the final rule.
                ---------------------------------------------------------------------------
                 \1365\ DOT HS 812 487: Mass Reduction for Light-Duty Vehicles
                for Model Years 2017-2025.
                ---------------------------------------------------------------------------
                 As stated above, for the NPRM, the interior, brake system,
                electrical accessory system, and steering and wheel systems were not
                included as part of the glider. The decision not to include the
                interior system was based on an assumption at that time that interior
                system mass reduction might adversely impact safety. In addition, the
                decision not to include the brake system was based on an assumption at
                that time that there would be little or no opportunity for downsizing
                and reducing mass based on the reduced weight from body and chassis
                only. As a result, brake systems were not considered as part of the
                glider in the NPRM. For the final rule, the agencies included the
                interior system based on market observations that light-weighted seats,
                side door trim, frontal dash, and others interior components have been
                incorporated on production vehicles that meet FMVSSs and perform well
                on voluntary NCAP and IIHS safety tests. The agencies also considered
                that interior, brakes, steering, wheel and electrical subsystems were
                included in the NHTSA light weighting studies. By adding the interior,
                steering, wheel subsystems and electrical subsystems as part of glider,
                the agencies believe light weighting the glider increases the
                opportunity for brake system optimization and mass reduction.
                Similarly, there is increased opportunity for mass reduction for wheels
                using gauge optimization, resulting from including more subsystems in
                the glider.
                 By including the interior, brake, steering, electrical accessory,
                and wheel subsystems in addition to the body and chassis subsystems in
                the definition of what subsystems comprise the glider, the agencies
                increased the glider weight from 50 percent of the vehicle curb weight
                to 71 percent of the vehicle curb weight. This increase in turn means
                that the potential for vehicle mass reduction was increased from 10
                percent of the vehicle curb weight to 20 percent of the vehicle curb
                weight. Table VI-126 shows the percent of light truck glider weight
                reduction and the corresponding vehicle curb weight reduction for each
                level of mass reduction for the glider shares used in the Draft TAR (75
                percent), NPRM (50 percent), and final rule (71 percent)
                analyses.\1366\
                ---------------------------------------------------------------------------
                 \1366\ Table 6-57 in PRIA showed the vehicle curb weight changes
                for different glider weight assumptions.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-C
                [[Page 24539]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.260
                2) Powertrain Mass Reduction
                 As explained above, any mass reduction due to powertrain
                improvements is accounted for separately from glider mass reduction.
                Autonomie considers several components for powertrain mass reduction,
                including engine downsizing, and transmission, fuel tank, exhaust
                systems, and cooling system lightweighting.
                 The 2015 NAS report suggested an engine downsizing opportunity
                exists when the glider mass is lightweighted by at least 10%. The 2015
                NAS report also suggested that 10% lightweighting of the glider mass
                alone would boost fuel economy by 3% and any engine downsizing
                following the 10% glider mass reduction would provide an additional 3%
                increase in fuel economy.\1367\ The agencies' lightweighting studies
                applied engine downsizing (for some vehicle types but not all) when the
                glider weight was reduced by 10 percent. Accordingly, the NPRM analysis
                limited engine resizing to several specific incremental technology
                steps; \1368\ important for this discussion, engines in the analysis
                were only resized when mass reduction of 10% or greater was applied to
                the glider mass, or when one powertrain architecture was replaced with
                another architecture.
                ---------------------------------------------------------------------------
                 \1367\ National Research Council. 2015. Cost, Effectiveness, and
                Deployment of Fuel Economy Technologies for Light-Duty Vehicles.
                Washington, DC--The National Academies Press. https://doi.org/10.17226/21744.
                 \1368\ 83 FR 43027.
                ---------------------------------------------------------------------------
                 Argonne performed a regression analysis of engine peak power versus
                weight for the NPRM based on attribute data taken from the A2Mac1
                benchmarking database, to account for the difference in weight for
                different engine types. For example, to account for weight of different
                engine sizes like 4-cylinder versus 8-cylinder, Argonne developed a
                relationship curve between peak power and engine weight based on the
                A2Mac1 benchmarking data. For the NPRM analysis, this relationship was
                used to estimate mass for all engine types regardless of technology
                type (e.g., variable valve lift and direct injection). Weight
                associated with changes in engine technology was applied by using this
                linear relationship between engine power and engine weight from the
                A2Mac1 benchmarking database. When a vehicle in the analysis fleet with
                an 8-cylinder engine adopted a more fuel efficient 6-cylinder engine,
                the total vehicle weight would reflect the updated engine weight with
                two less cylinders based on the peak power versus engine weight
                relationship.
                 When Autonomie selects a powertrain combination for a lightweighted
                glider, the engine and transmission are selected such that there is no
                degradation in the performance of the vehicle relative to the baseline
                vehicle. The resulting curb weight is a combination of the
                lightweighted glider with the resized and potentially new engine and
                transmission. This methodology also helps in accurately accounting for
                the cost of the glider and cost of the engine and transmission in the
                CAFE model. This is one of the fundamental differences between the
                analysis for this rulemaking the analysis for the Proposed
                Determination. For the Proposed Determination, the cost for mass
                reduction included mass reduction and cost reduction for one specific
                engine downsizing, and applied it to all vehicle classes without regard
                for performance and utility. There also was no accounting for the mass
                of other applied powertrains and the associated effectiveness impacts.
                 As explained in the introduction, secondary mass reduction is
                possible from some of the components in the glider after mass reduction
                has been incorporated in primary subsystems (body, chassis, and
                interior). Similarly, engine downsizing and powertrain secondary mass
                reduction is possible after certain level of mass reduction is
                incorporated in the glider. For the analysis, the agencies include both
                primary mass reduction, and when there is sufficient primary mass
                reduction, additional secondary mass reduction. The Autonomie
                simulations account for the aggregate of both primary and secondary
                glider mass reduction, and separately for powertrain mass.
                 The agencies received several comments about secondary mass
                reduction and powertrain mass reduction. Broadly, CARB commented that
                the agencies did not include powertrain downsizing and associated
                secondary mass reduction, which was a departure from the analysis done
                by
                [[Page 24540]]
                EPA for the Draft TAR.\1369\ CARB stated that the agencies
                ``inexplicably'' did not consider secondary mass reduction
                opportunities ``including but not limited to drive axles, suspension,
                and braking components (as a result of the overall vehicle being
                lighter); fuel tank (and corresponding weight of fuel during
                certification testing); powertrain (lighter engine and transmission
                needed to power the lighter vehicle); and thermal systems.'' CARB cited
                both EPA and NHTSA light weighting studies for the proposition that
                there are significant opportunities for secondary mass reduction that
                lead to additional cost savings. As a result, CARB stated that the
                agencies inflated the cost of mass reduction as well as the amount of
                mass reduction that is feasible and cost-effective, leading to an over
                estimate in the technology costs to meet the existing standards.
                ---------------------------------------------------------------------------
                 \1369\ NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 As CARB correctly noted, the NHTSA-sponsored studies have taken
                into consideration secondary mass reduction benefits such as radiator
                engine support, and optimized engine cradles, wheels, and suspension
                systems. As discussed above, in response to comments, the agencies have
                included additional subsystems such as brakes, wheels, steering,
                electrical, and interior systems to the glider for the final rule
                analysis, thereby accounting for mass reduction opportunities for these
                systems.
                 Also, as discussed further in Section VI.C.4.e), below, secondary
                mass reduction is integrated into the mass reduction cost curves.
                Specifically, the NHTSA studies, upon which the cost curves were built,
                first generated costs for lightweighting the vehicle body, chassis,
                interior, and other primary components, and then calculated costs for
                lightweighting secondary components. Accordingly, the cost curves
                reflect that, for example, secondary mass reduction for the brake
                system is only applied after there has been sufficient primary mass
                reduction to allow the smaller brake system to provide safe braking
                performance and to maintain mechanical functionality.
                 CARB appears to have misunderstood how the analysis accounts for
                powertrain mass reduction. The agencies described in the PRIA that the
                Autonomie simulations recognize that many powertrain packages have
                different weights for each vehicle class; for example, an eight-speed
                transmission may weigh more than a six-speed transmission, and a basic
                engine with variable valve timing may weigh more than a basic engine
                without variable valve timing.\1370\ Autonomie varies the weight of
                these powertrain systems as part of the analysis, and these changes are
                done separately from the glider mass reduction technology levels (MR0
                to MR6) in the simulations. Accordingly, accounting for powertrain mass
                reduction as part of the mass reduction technology analysis would
                double count impacts. The use of separate accounting assures that the
                analysis accounts for mass associated with secondary mass reduction
                from glider, and engine downsizing, as well as mass associated with
                each individual engine, transmission, and electrification technology.
                These mass changes were not accounted for in the Draft TAR and Proposed
                Determination analyses. Moreover, these are accounted for separately in
                the cost accounting, which is discussed further in the Section
                VI.C.4.e), below.
                ---------------------------------------------------------------------------
                 \1370\ PRIA at 418.
                ---------------------------------------------------------------------------
                 HDS commented that some assumptions in the Autonomie modeling
                related to engine weight appeared incorrect, such as the assumption
                that a turbocharged 4-cylinder engine weighed the same as a DOHV V6
                engine with 1.5 times the 4-cylinder's displacement, when in fact that
                engine is often 75 to 100 lbs. lighter.\1371\
                ---------------------------------------------------------------------------
                 \1371\ NHTSA-2018-0067-11985.
                ---------------------------------------------------------------------------
                 HDS also noted that ``mass reduction assumes no reduction of
                powertrain weight for mass reduction levels of 2.5% and 5%. Mass
                reduction effectiveness therefore are somewhat more appropriate for
                reductions over 5% which apparently include some powertrain weight
                reduction. More transparency in the PRIA regarding powertrain weight
                changes will allow more detailed comment on engine weight assumptions
                used.''
                 We agree with the comment that certain advanced engines could be
                lighter than a basic engine. For the final rule, the estimated mass
                levels for engines were updated, as discussed in Section VI.B.3 Tech
                Effectiveness, based on the A2Mac1 database and other sources that
                provided more precise mass data for powertrain technologies. Also, the
                agencies improved upon the precision of estimated engine weights by
                creating two curves to represent separately naturally aspirated engine
                designs and turbocharged engine designs.\1372\ This update resulted in
                two benefits. First, small naturally aspirated 4-cylinder engines that
                adopted turbocharging technology reflected the increased weight of
                associated components like ducting, clamps, the turbocharger itself, a
                charged air cooler, wiring, fasteners, and a modified exhaust manifold.
                Second, larger cylinder count engines like naturally aspirated 8-
                cylinder and 6-cylinder engines that adopted turbocharging and
                downsized technologies would have lower weight due to having fewer
                engine cylinders. For the final rule analysis, a naturally aspirated 8-
                cylinder engine that adopts turbocharging technology and is downsized
                to a 6-cylinder turbocharged engine appropriately reflects the added
                weight of the turbocharging components, and the lower weight of fewer
                cylinders. These refinements address the issues identified in HDS's
                comments.
                ---------------------------------------------------------------------------
                 \1372\ ANL Final Model Documentation for final rule analysis
                Chapter 5.2.9 Engine Weight Determination.
                ---------------------------------------------------------------------------
                 Regarding HDS's second comment, as discussed in the NPRM, to
                address product complexity and economies of scale, engine resizing is
                limited to specific incremental technology changes that would typically
                be associated with a major vehicle or engine redesign.\1373\ As
                discussed further in Section VI.B.3.a)(6) Performance Neutrality, the
                NPRM also referred to the 2015 NAS report conclusion that ``[f]or small
                (under 5 percent [of curb weight]) changes in mass, resizing the engine
                may not be justified, but as the reduction in mass increases (greater
                than 10 percent [of curb weight]), it becomes more important for
                certain vehicles to resize the engine and seek secondary mass reduction
                opportunities.'' \1374\ In consideration of both the NAS report and
                comments received from manufacturers, the agencies determined it would
                be reasonable to allow allows engine resizing upon adoption of 7.1%,
                10.7%, 14.2%, and 20% curb weight reduction, but not at 3.6% and
                5.3%.\1375\ Resizing is also allowed upon changes in powertrain type or
                the inheritance of a powertrain from another vehicle in the same
                platform. The increments of these higher levels of mass reduction, or
                complete powertrain changes, more appropriately match the typical
                engine displacement increments that are available in a manufacturer's
                engine portfolio.
                ---------------------------------------------------------------------------
                 \1373\ See 83 FR 43027 (Aug. 24, 2018).
                 \1374\ National Research Council. 2011. Assessment of Fuel
                Economy Technologies for Light-Duty Vehicles. Washington, DC--The
                National Academies Press. http://nap.edu/12924.
                 \1375\ These curb weight reductions equate to the following
                levels of mass reduction as defined in the analysis: MR3, MR4, MR5
                and MR6, but not MR1 and MR2; additional discussion of engine
                resizing for mass reduction can be found in Section VI.B.3
                Technology Effectiveness.
                ---------------------------------------------------------------------------
                [[Page 24541]]
                3) Summary of Final Rule Mass Reduction Technology Effectiveness
                 Figure VI-45 below shows the range of incremental effectiveness
                used for the NPRM analysis. The chart lumps all of the vehicle classes
                for each of the technology types.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.261
                BILLING CODE 4910-59-C
                 Figure VI-46 below shows the range of incremental effectiveness
                improvement from full vehicle modeling when mass reduction technologies
                were applied to vehicles for the final rule analysis.
                BILLING CODE 4910-59-P
                [[Page 24542]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.262
                BILLING CODE 4910-59-C
                e) Mass Reduction Costs
                 The PRIA described the decision to use NHTSA's passenger car light
                weighting study based on a MY 2011 Honda Accord and NHTSA's full-size
                pickup truck light weighting study based on a MY 2014 Chevrolet
                Silverado to derive the estimated cost for each of the mass reduction
                technology levels.\1376\ The agencies relied on the results of those
                studies because they considered an extensive range of material types,
                material gauge, and component redesign while taking into account real
                world constraints such as manufacturing and assembly methods and
                complexity, platform-sharing, and maintaining vehicle utility,
                functionality and attributes, including safety, performance, payload
                capacity, towing capacity, handling, NVH, and other characteristics. In
                addition, the agencies described that the baseline vehicles assessed in
                the NHTSA-sponsored studies were reasonably representative of baseline
                vehicles in the MY 2016 analysis fleet.\1377\ The agencies also noted
                they made the decision to rely on these studies after reviewing other
                agency, CARB, ICCT and industry studies.\1378\ The other studies often
                did not consider important factors, made unrealistic assumptions about
                key vehicle systems, and/or applied secondary mass reduction
                inappropriately, resulting in unrealistically low costs. The PRIA also
                described how the cost estimates derived from the NHTSA lightweighting
                studies were adjusted to reflect the NPRM glider share
                assumption.\1379\
                ---------------------------------------------------------------------------
                 \1376\ PRIA at 391; Table 6-38 and Table 6-41 in PRIA.
                 \1377\ PRIA at 403.
                 \1378\ As described in the PRIA at 390-91, studies by EPA, CARB,
                Transport Canada, the American Iron and Steel Institute (AISI), the
                Aluminum Association, and the American Chemistry Council were all
                reviewed for potential incorporation into the analysis.
                 \1379\ See PRIA at 396, Tables 6-38 and 6-39; PRIA at 401,
                Tables 6-41 and 6-42. See also PRIA at 391 (``While the definitions
                of glider may vary from study to study (or even simulation to
                simulation), the agencies referenced the same dollar per pound of
                curb weight to develop costs for different glider definitions. In
                translating these values, the agencies took care to track units ($/
                kg vs. $/lb.) and the reference for percentage improvements (glider
                vs. curb weight).'').
                ---------------------------------------------------------------------------
                 Furthermore, the agencies changed the cost of mass reduction
                accounting from a curb weight basis in the Draft TAR to glider weight
                basis in the NPRM.\1380\ Because the mass reduction studies provide
                mass reduction costs for the glider, this change enabled more direct
                use of cost curve data from the studies in the CAFE model. This change
                also allowed independent accounting for powertrain mass, which enabled
                the CAFE model to account more accurately for the unique mass of each
                of the powertrains that are available in each vehicle model. The cost
                of the engine, transmission, and electrification are accounted for
                separately from the glider in the CAFE model.
                ---------------------------------------------------------------------------
                 \1380\ In the Draft TAR, the agencies presented the cost
                estimates from mass reduction studies sponsored by both NHTSA and
                EPA. EPA presented the cost of mass reduction as function of vehicle
                curb weight. To harmonize the cost estimates with EPA, NHTSA also
                presented the cost of mass reduction as a function of vehicle curb
                weight.
                ---------------------------------------------------------------------------
                 The agencies received several comments on the mass reduction costs
                used in the NPRM. FCA commented that the costs and benefits used the
                CAFE model were overly optimistic,
                [[Page 24543]]
                stating that although its Ram 1500 pickup truck achieved several
                hundred pounds of weight reduction, the cost of achieving that weight
                reduction was greater than that used in the CAFE model.\1381\
                Similarly, as mentioned above, Toyota commented that mass reduction
                cost values were underestimated.\1382\ Conversely, CARB, UCS, and the
                City of Oakland in California commented that the costs used for mass
                reduction in the NPRM overstated the cost of mass reduction. The
                agencies also received several comments relating to the studies used to
                develop the mass reduction cost curves, how the values from those
                curves were applied in the CAFE model, and costs for secondary mass
                reduction; those comments are discussed in turn.
                ---------------------------------------------------------------------------
                 \1381\ NHTSA-2018-0067-11943.
                 \1382\ NHTSA-2018-0067-12098.
                ---------------------------------------------------------------------------
                (1) Studies Used To Develop Mass Reduction Cost Curves
                 The agencies described in the PRIA that since the 2012 final rule,
                both agencies conducted lightweighting studies to assess the technical
                feasibility and cost of mass reduction.\1383\ The agencies also stayed
                apprised of studies performed by other agencies, manufacturers, and
                industry trade associations, and reviewed them in development of
                lightweighting assumptions used in the NPRM and final rule
                analysis.\1384\ Among the several lightweighting studies, the agencies
                used NHTSA's passenger car lightweighting study, based on a MY 2011
                Honda Accord, and NHTSA's full-size pickup truck lightweighting study,
                based on a MY 2014 Chevrolet Silverado, to derive the cost estimates to
                achieve different levels of mass reduction for the NPRM and final rule.
                ---------------------------------------------------------------------------
                 \1383\ PRIA at 390.
                 \1384\ PRIA at 403.
                ---------------------------------------------------------------------------
                 The agencies described that the decision to rely on those studies
                included that those studies considered materials, manufacturing,
                platform-sharing, functional attribute, performance, and noise-
                vibration- and harshness (NVH), among other constraints pertaining to
                cost, effectiveness, and safety considerations, in addition to that
                these vehicles were a reasonable representation of the baseline
                vehicles in the MY 2016 compliance simulation.\1385\ Specifically in
                regards to safety, the agencies described a preference to use studies
                that considered small overlap impact tests conducted by the Insurance
                Institute for Highway Safety (IIHS) and not all studies took that test
                into account. In regards to maintaining vehicle functionality, the
                agencies described that the NHTSA pickup truck study accounted for
                vehicle functional performance for attributes including towing, noise
                and vibration, and gradeability, in addition to considering platform
                sharing constraints.
                ---------------------------------------------------------------------------
                 \1385\ PRIA at 403.
                ---------------------------------------------------------------------------
                 In contrast, the agencies explained that the other studies often
                did not consider many important factors, or those studies made
                unrealistic assumptions about key vehicle systems through secondary
                downsizing, resulting in unrealistically low costs. Specifically, the
                agencies referenced EPA's past analysis of a MY 2010 Toyota Venza as an
                example of a study that employed overly aggressive secondary mass
                reduction, which translated into cost savings for the initial 10% mass
                reduction.\1386\
                ---------------------------------------------------------------------------
                 \1386\ PRIA at 391.
                ---------------------------------------------------------------------------
                 The agencies received several comments on the studies used to
                generate the mass reduction cost curves. Ford commented in support of
                the agencies' decision to exclude mass reduction studies that were
                misaligned with tear-down studies.\1387\ Ford cited the MY 2010 Toyota
                Venza Phase II study used to establish the mass reduction cost values
                used for the Draft TAR and Proposed Determination that suggested the
                first 7-10% of mass reduction could be accomplished with zero or
                reduced cost,\1388\ which Ford characterized as ``a gross
                underestimation of industry investment and material costs associated
                with any weight reduction.''
                ---------------------------------------------------------------------------
                 \1387\ NHTSA-2018-0067-11928.
                 \1388\ EPA-420-R-16-021: Proposed Determination Technical
                Support Document at 2-158, November 2016.
                ---------------------------------------------------------------------------
                 ICCT commented that The National Academies of Science
                ``specifically endorsed tear-down studies as the most appropriate way
                to get at vehicle technology costs, [as those] studies are typically
                more accurate and far more transparent than the older method of
                surveying manufacturers, and such whole-vehicle studies are key to
                capturing holistic vehicle level mass-reduction technology costs.''
                ICCT noted that there are many peer-reviewed tear-down studies that
                demonstrate that at least 20 percent mass reduction is available for
                adoption across vehicle classes by 2025, including studies by EDAG,
                FEV, Ford, and Lotus Engineering; however, ICCT alleged that the
                agencies ``have either incorrectly interpreted or invalidly nullified
                the most relevant detailed engineering teardown studies on mass-
                reduction technology.'' ICCT noted that the agencies were ``well
                aware'' of these studies, as they were performed by CARB in conjunction
                with the agencies, however, ICCT alleged that the agencies
                ``reinterpreted the results of the main study relied upon in the TAR in
                order to inflate costs,'' and that the ``technical assessment by the
                agencies has a clear technical bias towards reducing CAFE and GHG
                standards.'' ICCT concluded that ``[e]xcluding these studies amounted
                to intentionally disregarding the most pertinent and rigorous
                engineering studies that are applicable to the rulemaking timeframe.''
                 ICCT recommended the agencies adjust their technology cost inputs
                to reflect the ``best-available technology studies.'' ICCT stated that
                the correct cost assumption from these studies is that ``a 5-10% mass
                reduction by 2025 reduces vehicle cost, and the auto industry will
                cost-effectively deploy at least 15% vehicle curb mass reduction in the
                2025 timeframe at near zero net cost (and consistently less than
                $500).''
                 CARB asserted that the agencies inflated the costs of mass
                reduction in the NPRM analysis by only considering NHTSA-sponsored
                studies and improperly excluding the effects of secondary mass
                reduction as documented in those studies.\1389\ CARB provided a table
                of studies that largely mirrored the tables of studies the agencies
                considered in the Draft TAR and PRIA,\1390\ and also included the
                associated mass reduction costs in $/kg included in each study, noting
                that for all excluded studies cited in the table, all mass reduction
                costs were substantially lower than the values used in the agencies'
                analysis.\1391\ Similarly, UCS commented that while the PRIA did state
                that additional studies ``often did not consider many important factors
                or . . . made unrealistic assumptions about key vehicle systems,'' the
                agencies did not specifically identify the factors and assumptions that
                merited disregarding those studies, which were included previously in
                agency analysis as part of the record when deriving previous estimates
                for the costs of mass reduction.\1392\
                ---------------------------------------------------------------------------
                 \1389\ NHTSA-2018-0067-11873.
                 \1390\ Draft TAR at 5-168; PRIA at 404-05.
                 \1391\ NHTSA-2018-0067-11873.
                 \1392\ NHTSA-2018-0067-12039.
                ---------------------------------------------------------------------------
                 The agencies agree with ICCT that peer-reviewed tear-down studies
                present an appropriate method to capture holistic vehicle-level mass
                reduction technology costs. The agencies also agree with ICCT that the
                agencies were well aware of studies conducted by EDAG, FEV, Ford, and
                Lotus Engineering; in fact, the agencies
                [[Page 24544]]
                presented a table listing several of those studies in the PRIA with the
                qualification that those studies were reviewed in developing
                lightweight assumptions for the analysis, but those studies did not
                consider important factors, or those studies made unrealistic
                assumptions about key vehicle systems through secondary downsizing,
                resulting in unrealistically low costs.
                 The agencies also agree with UCS' comment that the language could
                have been more specific about identifying the factors and assumptions
                that merited disregarding studies that were previously included as part
                of the record when deriving previous estimates for the costs of mass
                reduction. The following discussion briefly summarizes the record since
                the Draft TAR and differences between NHTSA's and other lightweighting
                studies' approach to factors listed in the PRIA. Important for this
                discussion is an understanding of primary versus secondary mass
                reduction; as described above, when there is sufficient primary mass
                reduction, other components that are designed based on the mass of
                primary components may be redesigned and have lower mass. Recall the
                braking system example used throughout this section; mass reduction in
                the braking system is secondary mass reduction because it requires
                primary mass reduction before it can be incorporated. If the mass of
                primary components is reduced sufficiently, the resulting lighter
                weight vehicle could maintain braking performance, attributes, and
                safety with a lighter weight brake system.
                 Several studies were referenced in the Draft TAR that either used
                tear-down analyses and computer-aided engineering (CAE) to produce a
                future engineered lightweight vehicle, or considered future
                technologies and processes for lightweighting vehicle components.\1393\
                ---------------------------------------------------------------------------
                 \1393\ Draft TAR at 5-158 through 5-197.
                ---------------------------------------------------------------------------
                 EPA developed cost curves for cars and CUVs based on the MY 2010
                Toyota Venza study, and pickup truck cost curves based on the MY 2011
                Chevrolet Silverado study.\1394\ The other studies were considered by
                EPA, but not used to develop the Draft TAR, Proposed Determination and
                Final Determination cost curves. In brief, EPA described that the
                Toyota Venza Phase I was a mass reduction opportunity study only, and
                the Phase II study was a holistic vehicle study that examined nearly
                every component in the vehicle for mass reduction potential and
                calculated a related cost and mass saved for each. For the cost curve,
                EPA applied the individual components in sequence from largest cost per
                kilogram savings to largest cost per kilogram increase. For example,
                the cost curve for the Draft TAR and Proposed Determination applied
                engine downsizing and transmission system mass reduction first, and
                before lightweighting the body, chassis, doors and other
                components.\1395\ EPA stated this methodology was chosen based on the
                understanding that OEMs will choose the cost saving technologies first
                and that some cost mass reduction technologies will be paid for by the
                cost save mass reduction technologies, citing a 2016 publication by CAR
                and a GM presentation that stated over $2,000,000,000 was saved in
                material costs through various lightweighting approaches.\1396\
                ---------------------------------------------------------------------------
                 \1394\ Draft TAR at 5-367.
                 \1395\ EPA-420-R-16-021: Proposed Determination Technical
                Support Document at 2-161 and 2-162
                 \1396\ Draft TAR at 5-172 (citing ``Identifying Real world
                Barriers to Implementing Lightweighting Technologies and Challenges
                in Estimating the Increase in Costs,'' Center for Automotive
                Research, Jay Baron, Ph.D., January 2016 http://www.cargroup.org/?module=Publications&event=View&pubID=128; General Motors, ``General
                Motors 2015 Global Business Conference,'' Presentation, October 1,
                2015, Slides 43-45 in document, https://www.gm.com/content/dam/gm/events/docs/5194074-596155-ChartSet-10-1-2015.).
                ---------------------------------------------------------------------------
                 The NHTSA cost curves were developed by rearranging the
                lightweighted components from the MY 2011 Honda Accord and MY 2014
                Chevrolet Silverado studies based on cost effectiveness, assuming the
                vehicle body, chassis, interior, and other primary components were
                lightweighted first, followed then by lightweighting powertrain
                components and other secondary systems.\1397\ The cost curves based on
                the NHTSA studies reflect that, returning to this example, secondary
                mass reduction for the brake system is only applied after there has
                been sufficient primary mass reduction to allow the smaller brake
                system to provide safe braking performance and to maintain mechanical
                functionality.\1398\
                ---------------------------------------------------------------------------
                 \1397\ Draft TAR at 5-421 (``The powertrain components which
                include engine, transmission, and fuel systems such as fuel filler
                pipe, fuel tank, fuel pump, etc., exhaust systems and cooling
                systems were not considered for application of primary mass
                reduction but benefits of secondary mass reduction were accounted
                for. These powertrain components are assumed to be downsized only
                after the primary vehicle structural components (Body-In-White)
                achieve certain level of mass reduction.'').
                 \1398\ Draft TAR at 5-422.
                ---------------------------------------------------------------------------
                 The EPA and NHTSA studies took fundamentally different approaches
                to accounting for the costs of mass reduction technology, and
                accordingly, EPA needed to translate the cost curves from the NHTSA
                studies to use a similar methodology as the cost curves from the EPA
                studies.\1399\ To ``normalize'' the NHTSA studies with the EPA's
                studies, EPA listed components identified for lightweighting in the
                NHTSA studies and reorganized those components from the lowest cost to
                highest cost along with associated mass reduction per the ``whole
                vehicle'' approach mentioned above, distributed mass savings from
                secondary mass reduction to all points along the cost curve, and
                included the mass saved from engine downsizing without taking into
                consideration the cost of added engine technology. This resulted in
                lower-cost secondary mass reduction opportunities being considered
                before primary mass reduction opportunities, which in turn resulted in
                artificially low $/kg costs for mass reduction.
                ---------------------------------------------------------------------------
                 \1399\ Draft TAR at 5-369.
                ---------------------------------------------------------------------------
                 For the NPRM and final rule, the agencies simply used the original
                ordered list of components from the MY 2011 Honda Accord study and MY
                2014 Chevrolet Silverado study, arranged sequentially for cost
                effectiveness based on primary then secondary mass reduction
                opportunities, to generate the cost curves for passenger cars and light
                trucks. Accordingly, the agencies did not ``reinterpret'' the results
                of studies used in the Draft TAR in the NPRM, as ICCT alleged, but
                rather appropriately represented how primary and secondary mass
                reduction opportunities are implemented in the real world (to the
                extent that ICCT is referring to the translation of the study costs to
                the NPRM glider weight assumptions, that is discussed in Section
                VI.C.4.e)(1), below). To maintain utility and performance in the real
                world, primary components must be lightweighted first before the engine
                and transmission can be resized. Moreover, as described in the Draft
                TAR, NHTSA's mass reduction studies did not ``improperly exclude'' the
                effects of secondary mass reduction, rather those effects were
                appropriately accounted for after primary components achieved certain
                levels of mass reduction. As discussed in Section VI.B.3.a)(6)
                Performance Neutrality, this approach aligned with the NAS approach to
                consider powertrain downsizing only after the vehicle structural
                components achieved 10 percent mass reduction.
                 OEMs have also disagreed with the conclusion that mass reduction
                could come at a cost savings. For instance, Ford characterized the
                Toyota Venza studies, which concluded the first 7-10% of mass reduction
                could come at a negative cost as ``a gross
                [[Page 24545]]
                underestimation of industry investment and material costs associated
                with any weight reduction.'' The agencies believe that the approach to
                secondary mass reduction followed in the NHTSA passenger car and pickup
                truck lightweighting studies appropriately incorporated both the costs
                and real-world constraints associated with employing primary and
                secondary mass reduction technologies.
                 Aside from the differences in how studies treated secondary mass
                reduction, the agencies opted not to use, or could not use, other
                studies either previously considered in the rulemaking record or
                mentioned by commenters for several reasons:
                 Studies were not comprehensive, and therefore could not be used to
                develop a comprehensive cost curve: Some studies narrowly assessed
                lightweighting of a portion of vehicle, such as the body in white
                subsystem, or as stated in the PRIA,\1400\ were limited to material
                substitution of the vehicle components, such as replacing steel with
                aluminum or replacing mild steel with AHSS or replacing mild steel with
                CFRP in selective components. Factors important to vehicle
                functionality, like material joining techniques and the feasibility of
                manufacturing processes or necessary retooling requirements were not
                considered, and therefore could not be used to develop a comprehensive
                cost curve representative of the costs required to reduce mass in a
                vehicle.\1401\
                ---------------------------------------------------------------------------
                 \1400\ PRIA at 391.
                 \1401\ An Assessment of Mass Reduction Opportunities for a 2017-
                2020 Model Year Vehicle Program, March 2010, Lotus Engineering, at
                p. 6.
                ---------------------------------------------------------------------------
                 Cost curves were not developed or no cost analysis was performed:
                For the CARB Holistic Vehicle Mass Reduction/Cost Study, a cost curve
                was not developed, and the resulting cost per kilogram data points were
                point estimates. The calculated cost per kilogram was used as one data
                point of several to indicate the direction for mass reduction beyond
                EPA's original passenger car/CUV curve.\1402\ Or, as in the case of the
                DOE/Ford/Magna Multi Material Lightweight Vehicle (MMLV) project, no
                cost analysis was performed for the initial project, and later
                project(s) concluded that ``a 37% to 45% mass reduction in a standard
                mid-sized vehicle is within reach if carbon fiber composite materials
                and manufacturing processes are available and if customers are willing
                to accept a reduction in vehicle features and content, as demonstrated
                with the Multi-Materials and Carbon Fiber Composite-Intensive vehicle
                scenarios.'' \1403\
                ---------------------------------------------------------------------------
                 \1402\ Draft TAR at 5-185.
                 \1403\ Draft TAR at 5-194.
                ---------------------------------------------------------------------------
                 Engineered vehicles did not meet functional design or manufacturing
                requirements: As noted in the update to EPA's Light-Duty Vehicle Mass
                Reduction and Cost Analysis for the Toyota Venza, the Phase I
                engineered Venza did not meet the design target of no expected NVH
                degradation.\1404\ The Phase II (High Development) study assumed
                significant cost savings from reduced parts manufacturing, but did not
                appropriately explain the methodology used to conclude that the part
                count reduction was feasible.\1405\
                ---------------------------------------------------------------------------
                 \1404\ Light-Duty Vehicle Mass Reduction and Cost Analysis--
                Midsize Crossover Utility Vehicle, EPA-420-R-12-026 (August 2012).
                 \1405\ Peer Review of Demonstrating the Safety and
                Crashworthiness of a 2020 Model-Year, Mass-Reduced Crossover Vehicle
                (Lotus Phase 2 Report), EPA-420-R-12-028 (September 2012).
                ---------------------------------------------------------------------------
                 In addition, the agencies qualified in the PRIA a preference to use
                studies that considered the small overlap impact test conducted by
                IIHS, and not all studies took that test into account.\1406\ NHTSA's
                ``Update to Future Midsize Lightweight Vehicle Findings in Response to
                Manufacturer Review and IIHS Small-Overlap Testing'' based on the MY
                2011 Honda Accord presented results incorporating suggestions from
                Honda regarding NVH and durability, and updating the engineered vehicle
                to achieve a ``good'' rating in seven crash safety tests.\1407\ EPA
                studies also accounted for the IIHS small overlap test through an ad
                hoc estimate of mass and cost, unlike the NHTSA update, which
                explicitly modeled to account for NVH performance and to comply with
                the IIHS small overlap test.
                ---------------------------------------------------------------------------
                 \1406\ PRIA at 391.
                 \1407\ Singh, H., Kan, C-D., Marzougui, D., & Quong, S. (2016,
                February). Update to future midsize lightweight vehicle findings in
                response to manufacturer review and IIHS small-overlap testing
                (Report No. DOT HS 812 237). Washington, DC: National Highway
                Traffic Safety Administration.
                ---------------------------------------------------------------------------
                 The agencies continue to believe that the MY 2011 Honda Accord and
                MY 2014 Chevrolet Silverado lightweighting studies are the best studies
                upon which to estimate the costs of mass reduction in the rulemaking
                timeframe.
                (2) How the Cost Curves Are Applied in the Model
                 Commenters also submitted comments on how the cost curves were
                applied in the model, including that the studies the agencies relied
                upon to generate cost curves, discussed above, did not support the 50
                percent glider share assumption used in the NPRM, and the agencies did
                not correctly scale the costs to match the glider share assumption.
                 UCS commented that the agencies based the costs for mass reduction
                on glider weight reduction, however, the need for more expensive
                materials and more advanced engineering and design strategies only
                results from the need for greater levels of absolute mass reduction on
                the vehicle.\1408\ UCS stated that the cost curves had effectively been
                derived from the assumption of reductions as great as 16.8 percent
                reduction in curb weight in the case of the Silverado (Singh et al.
                2018) and as great as 18 percent reduction in curb weight in the case
                of the Honda Accord (Singh et al. 2016), but applied to curb weight
                reductions approximately two-thirds that magnitude. UCS stated that
                approach was ``completely invalid and significantly overstates the
                costs for mass reduction.'' UCS also commented that the agencies
                incorrectly scaled the cost curves based on the agencies' mass
                reduction studies, which refer to direct manufacturing costs as a
                function of vehicle curb weight, not just glider weight. UCS stated
                this incorrectly yielded the same costs for two-thirds the amount of
                mass reduction.
                ---------------------------------------------------------------------------
                 \1408\ NHTSA-2018-0067-12039.
                ---------------------------------------------------------------------------
                 CARB similarly commented that the mass reduction costs assigned to
                both passenger cars and light trucks in the CAFE model were
                inappropriately inflated based on incorrect scaling from the glider
                share assumptions used in the Honda Accord and Chevy Silverado studies
                to the NPRM glider share value.\1409\ CARB analyzed two tables in the
                PRIA that showed the agencies' translation of cost numbers derived from
                the two studies to the cost numbers used in the CAFE model, and
                asserted that the agencies improperly used costs from the upper end of
                the mass reduction range rather than the midpoint of the range, leading
                to cost overestimation.
                ---------------------------------------------------------------------------
                 \1409\ NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 Similarly, HDS commented that the PRIA passenger car cost curves
                used data that were not in agreement with the study that they were
                based upon, noting that the Honda Accord study showed the glider
                accounting for 78% of curb weight, and this limited absolute weight
                reduction.\1410\ HDS noted that the truck weight reduction cost data
                were closer to those cited in the Chevy Silverado teardown study,
                although the glider share for that study was also 73.6% of vehicle curb
                weight.
                ---------------------------------------------------------------------------
                 \1410\ NHTSA-2018-0067-11985.
                ---------------------------------------------------------------------------
                 HDS also commented that although the agencies relied on the same
                Honda Accord study that was used in the Draft
                [[Page 24546]]
                TAR, ``the costs have been changed significantly [from the Draft TAR]
                for unexplained reasons.'' \1411\ HDS stated that the PRIA showed
                average costs for mass reduction, whereas earlier studies showed the
                cost increment for each 5% mass reduction, noting that with increasing
                incremental cost with increased mass reduction, average cost will
                always be lower than incremental cost. HDS claimed that it was
                ``unusual'' for the Draft TAR incremental costs to decrease between 11%
                and 19% mass reduction but increase elsewhere, but also noted the
                unexplained increase in cost, specifically a $536 cost for 175kg weight
                reduction, shown in the PRIA.
                ---------------------------------------------------------------------------
                 \1411\ NHTSA-2018-0067-11985.
                ---------------------------------------------------------------------------
                 HDS also compared manufacturing costs from the Draft TAR to the
                PRIA analysis, noting that the direct manufacturing cost was found to
                be negative (i.e., a cost saving) in the Draft TAR analysis for mass
                reduction up to 15 percent, but EPA assumed the indirect costs were
                positive so that the total cost was a sum of positive and negative
                costs--meaning the total cost could be positive or negative. In
                contrast, HDS noted there were no negative costs in the cost curves
                used for the PRIA analysis, resulting in a very large differential
                between the costs of mass reduction, with the 2018 average cost being
                higher than even the 2016 marginal costs.
                 Three notable changes from the NHTSA Draft TAR to NPRM and final
                rule analysis impacted how the cost curves for mass reduction are
                applied in the CAFE Model.
                 First, the Draft TAR considered mass reduction in the glider and
                powertrain together, and calculated the percentage mass reduction on a
                vehicle curb weight basis. In the Draft TAR, only one engine and
                transmission combination were considered to account for the mass change
                associated with downsizing the engine, and the cost estimates for mass
                reduction for this one powertrain combination was applied to all
                powertrain combinations. This approach did not account for the mass
                changes associated with the application of powertrain technologies
                (engine, transmission and electrification) technologies, and did not
                account for the corresponding change in glider mass needed to offset
                the powertrain mass change and to achieve the specified curb weight
                mass reduction level. This approach did not reflect the real world,
                where there are many vehicles with different body styles and powertrain
                combinations, and therefore did not account for differences in mass for
                different engines, transmissions, or electrification.
                 Accordingly, for the NPRM and final rule, the cost of mass
                reduction was calculated on a glider weight basis so that the weight of
                each powertrain configuration could be directly and separately
                accounted for. This approach provides the true cost of mass reduction
                without conflating the mass change and costs associated with downsizing
                a powertrain or adding additional advanced powertrain technologies.
                Hence, the mass reduction costs in the NPRM reflect the cost of mass
                reduction in the glider and do not include the mass reduction
                associated with engine downsizing, and therefore appear to be higher
                than the cost estimates in the Draft TAR.
                 Second, the glider share of curb weight changes from the Draft TAR
                to NPRM and from the NPRM to the final rule analysis also affected the
                absolute amount of curb weight reduction that was applied, and
                therefore for cost per pound for the mass reduction changes with the
                change in the glider share. The cost for removing 20 percent of the
                glider weight when the glider represents 75% of a vehicle's curb weight
                is not the same as the cost for removing 20 percent of the glider
                weight when the glider represents 50% of the vehicle's curb weight. For
                example, the glider share of 79 percent of a 3,000-pound curb weight
                vehicle is 2,370 pounds, while the glider share of 50 percent of a
                3,000-pound curb weight vehicle is 1,500 pounds, and the glider share
                of 71 percent of a 3,000-pound curb weight vehicle is 2,130 pounds. The
                mass change associated with 20 percent mass reduction is 474 pounds for
                79 percent glider share (=[3,000 pounds x 79% x 20%]), 300 pounds for
                50 percent glider share (=[3,000 pounds x 50% x 20%]), and 426 pounds
                for 71 percent glider share (=[3,000 pounds x 71% x 20%]). The mass
                reduction cost studies show that the cost for mass reduction varies
                with the amount of mass reduction. Therefore, for a fixed glider mass
                reduction percentage, different glider share assumptions will have
                different costs.
                 To further illustrate, Table VI-127 and Table VI-128 below shows
                the associated curb weight percentage mass reduction and the associated
                average cost per pound for different glider weight assumptions for each
                glider mass reduction technology level used in the final rule analysis.
                For reference, the costs from the passenger car light weighting study
                are presented.\1412\ These costs were the basis for deriving the costs
                for each mass reduction technology level in the Draft TAR, NPRM, and
                final rule analyses, using the unique glider share values for each of
                those analyses. In the light weighting study, NHTSA applied the mass
                reduction technologies identified for the exemplar vehicle on other
                vehicle(s) and vehicle types to understand the level of mass reduction
                that could be achieved. In the case of passenger cars, the maximum
                level of mass reduction was around 15% of the vehicle curb weight if
                all the mass reduction technologies are applied. In other words,
                achieving mass reduction greater than 10% of the curb weight for
                passenger cars will require extensive use of advanced materials such as
                high strength aluminum and carbon fiber composite material.
                ---------------------------------------------------------------------------
                 \1412\ Table 6-39 in PRIA.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24547]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.263
                [GRAPHIC] [TIFF OMITTED] TR30AP20.264
                BILLING CODE 4910-59-C
                 Finally, as explained earlier, to determine the mass reduction
                technology levels for the NPRM 2016 analysis fleet, a distribution of
                the residuals from the regression using 50 percent glider weight
                generally showed a greater percentage of vehicles achieving higher
                levels of mass reduction. With this high level of mass reduction
                already achieved, the opportunities for further mass reduction would be
                limited and have higher costs. For the final rule, since the agencies
                updated the glider share to 71 percent of the vehicle curb weight, the
                distribution of residuals from the regression shifted some vehicles to
                lower baseline mass reduction
                [[Page 24548]]
                technology levels, providing more opportunity for further mass
                reduction, on average. Even as some of the vehicles start further up on
                the mass reduction cost curve due to higher levels of mass reduction
                technology (MR3, MR4) already present in the vehicles, there are
                additional opportunities for further mass reduction to achieve MR5 and
                above.
                 Table VI-127 and Table VI-128 show that for the final rule, cost
                estimates with the 71 percent glider share come closer to the cost
                estimates used in Draft TAR, which assumed a 79 percent glider share.
                (3) Secondary Mass Reduction Costs
                 As discussed above, the agencies changed the cost of mass reduction
                calculation from a curb weight basis in the Draft TAR to a glider
                weight basis in the NPRM.\1413\ This change allowed us to estimate the
                cost of mass reduction independently of the cost associated with
                downsized advanced engines and advanced transmissions, as the cost of
                downsized advanced engines and transmissions are accounted for
                separately in the CAFE model.
                ---------------------------------------------------------------------------
                 \1413\ In the Draft TAR, the agencies presented the cost
                estimates from mass reduction studies sponsored by both NHTSA and
                EPA. EPA presented the cost of mass reduction as function of vehicle
                curb weight. To harmonize the cost estimates with EPA, NHTSA also
                presented the cost of mass reduction as a function of vehicle curb
                weight.
                ---------------------------------------------------------------------------
                 The MY 2011 Honda Accord and MY 2014 Chevy Silverado studies used
                to develop the NPRM and final rule cost curves for mass reduction
                technologies include some non-powertrain secondary mass reduction
                technologies such as brakes and wheels. The agencies presented the list
                of mass reduction technologies in NPRM.\1414\ Following the publication
                of NHTSA's light weighting studies, peer reviewers and manufacturers
                commented that many components such as drive axles, engine cradles, and
                radiator engine support that are considered to be non-powertrain
                secondary mass reduction opportunities cannot be downsized, as the same
                components are used across many vehicles with different powertrain
                options. Even though some of these components may provide opportunities
                for additional mass reduction, NHTSA agreed with peer reviewers and
                manufacturers that retaining a common design for all powertrain options
                provides for cost reductions due to economies of scale.
                ---------------------------------------------------------------------------
                 \1414\ Table 6-37 and Table 6-40 in PRIA.
                ---------------------------------------------------------------------------
                 Commenters faulted the agencies for a perceived lack of accounting
                for the cost decreases from secondary mass reduction. ICCT commented
                although the agencies relied on the Honda Accord study, which
                considered cost savings from downsizing the powertrain, in the NPRM
                only glider weight reduction was ever considered without the cost-
                offsetting engine downsizing.\1415\ ICCT stated that this omission had
                two effects, first that accounting for associated powertrain weight
                reductions would have allowed for more mass reduction, thus allowing
                for greater efficiency benefits at a lower cost, and second, that
                vehicle performance was erroneously improved, contrary to the agencies'
                assertion that the analysis assumed a level of performance neutrality.
                ICCT concluded that it was unclear if and how costs were reduced for
                powertrain downsizing, as well as the precise changes to fuel
                efficiency.
                ---------------------------------------------------------------------------
                 \1415\ NHTSA-2018-0067-11741.
                ---------------------------------------------------------------------------
                 CARB faulted the agencies for not including secondary mass
                reduction in the NPRM analysis, and stated that by failing to account
                for secondary mass reduction as was done in the Draft TAR, the agencies
                inflated the costs for mass reduction as well as the amount of mass
                reduction that is feasible and cost-effective leading to an
                overestimate in the technology costs needed to meet the existing
                standards.
                 The agencies note that the cost curves used for the NPRM and this
                final rule do in fact include secondary mass reduction. The cost curves
                reflect secondary mass reduction applied when there is sufficient
                primary mass reduction to implement secondary mass reduction without
                degrading function and safety. Specifically, the NHTSA studies, upon
                which the cost curves were built, first generated costs for
                lightweighting the vehicle body, chassis, interior, and other primary
                components, and then calculated costs for lightweighting secondary
                components. Accordingly, the cost curves reflect that, for example,
                secondary mass reduction for the brake system is only applied after
                there has been sufficient primary mass reduction to allow the smaller
                brake system to provide safe braking performance and to maintain
                mechanical functionality.
                 In addition, CARB stated that the 2011 Honda Accord and the 2014
                Chevrolet Silverado studies had ``markedly'' lower costs than the
                proposal when secondary mass reduction is included. Again, the agencies
                believe these comments resulted from a lack of understanding about how
                the analysis considers primary and secondary mass reduction, even
                though the NPRM and PRIA explicitly stated how costs are accounted for
                separately.\1416\ Also, as discussed above, engine mass reduction
                enabled by mass reduction in the glider is accounted for separately and
                therefore not included as part of glider mass reduction technology, as
                doing so would result in double counting the impacts.
                ---------------------------------------------------------------------------
                 \1416\ PRIA at 413.
                ---------------------------------------------------------------------------
                (4) Summary of Final Rule Mass Reduction Costs
                 For the final rule, the agencies continue to use multiple mass
                reduction technology levels and costs based on the lightweighting
                studies that were presented in PRIA.\1417\ Since the agencies have
                changed the glider share of curb weight assumption from 50 percent in
                NPRM to 71 percent in the final rule, the mass reduction costs reflect
                the updated glider share. Table VI-129 and Table VI-130 show mass
                reduction costs used in the CAFE model for passenger car and light
                trucks.
                ---------------------------------------------------------------------------
                 \1417\ Table 6-37 and 6-40 in PRIA.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24549]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.265
                [GRAPHIC] [TIFF OMITTED] TR30AP20.266
                BILLING CODE 4910-59-C
                5. Aerodynamics
                 The energy required to overcome aerodynamic drag accounts for a
                significant portion of the energy consumed by a vehicle, and can become
                the dominant factor for a vehicle's energy consumption at high speeds.
                Reducing aerodynamic drag can, therefore, be an effective way to reduce
                fuel consumption and emissions.
                 Aerodynamic drag is proportional to the frontal area (A) of the
                vehicle and coefficient of drag (Cd), such that aerodynamic
                performance is often expressed as the product of the two values,
                CdA, which is also known as the drag area of a vehicle. The
                coefficient of drag (Cd) is a dimensionless value that
                essentially represents the aerodynamic efficiency of the vehicle shape.
                The
                [[Page 24550]]
                frontal area (A) is the cross-sectional area of the vehicle as viewed
                from the front. It acts with the coefficient of drag as a sort of
                scaling factor, representing the relative size of the vehicle shape
                that the coefficient of drag describes. The force imposed by
                aerodynamic drag increases with the square of vehicle velocity,
                accounting for the largest contribution to road loads' higher speeds.
                 Aerodynamic drag reduction can be achieved via two approaches,
                either by reducing the drag coefficient or reducing vehicle frontal
                area, with two different categories of technologies, passive and active
                aerodynamic technologies. Passive aerodynamics refers to aerodynamic
                attributes that are inherent to the shape and size of the vehicle,
                including any components of a fixed nature. Active aerodynamics refers
                to technologies that variably deploy in response to driving conditions.
                These include technologies such as active grille shutters, active air
                dams, and active ride height adjustment. It is important to note that
                manufacturers may employ both passive and active aerodynamic
                technologies to achieve aerodynamic drag values.
                 The greatest opportunity for improving aerodynamic performance is
                during a vehicle redesign cycle when significant changes to the shape
                and size of the vehicle can be made. Incremental improvements may also
                be achieved during mid-cycle vehicle refresh using restyled exterior
                components and add-on devices. Some examples of potential technologies
                applied during mid-cycle refresh are restyled front and rear fascia,
                modified front air dams and rear valances, addition of rear deck lips
                and underbody panels, and low-drag exterior mirrors. While
                manufacturers may nudge the frontal area of the vehicle during
                redesigns, large changes in frontal area are typically not possible
                without impacting the utility and interior space of the vehicle.
                Similarly, manufacturers may improve Cd by changing the
                frontal shape of the vehicle or lowering the height of the vehicle,
                among other approaches, but the form drag of certain body styles and
                airflow needs for engine cooling often limit how much Cd may
                be improved.
                 During the vehicle development process, manufacturers use various
                tools such as Computational Fluid Dynamics (CFD), scaled clay models,
                and full size physical prototypes for wind tunnel testing and
                measurements to determine aerodynamic drag values and to evaluate
                alternate vehicle designs to improve those values.
                 The agencies presented a table in the PRIA showing aerodynamic drag
                improvements from individual technologies based on wind-tunnel testing
                for a study commissioned by Transport Canada, which is reproduced in
                Table VI-131 below.\1418\ The individual technologies are present in
                many of the 2016 and 2017 vehicles in the fleet. Table VI-131 shows the
                list of aerodynamic technologies and corresponding aero drag
                improvements.
                ---------------------------------------------------------------------------
                 \1418\ Table 6-63 in PRIA.
                 [GRAPHIC] [TIFF OMITTED] TR30AP20.267
                
                 As discussed in the PRIA and further below, the agencies made
                several notable changes for modeling aerodynamic improvement
                technologies from the Draft TAR to the NPRM. First, the agencies
                revised the aerodynamic
                [[Page 24551]]
                improvements from two levels in the Draft TAR (10% and 20% improvement
                over the baseline) to four levels (5%, 10%, 15% and 20% aerodynamic
                drag improvement values over the baseline). This change provided the
                improved granularity to bin the vehicles with different aerodynamic
                improvements more appropriately. Next, the agencies assigned levels of
                aerodynamic technology to the MY 2016 fleet on a relative basis based
                on confidential business information submitted by the manufacturers,
                taking steps to verify information submitted by manufactures with other
                sources, and making changes particularly for vehicles that showed large
                improvements over baseline values. Third, the agencies limited the
                maximum level of aerodynamic improvements that certain body styles
                (pickup trucks, minivans) could achieve and limited the maximum level
                of improvements that cars and SUVs with more than 405 horsepower could
                achieve, based on the agencies' assessment of industry comments.
                Finally, the agencies updated the cost for aerodynamic improvements
                based on the assessment of comments that the National Academy of
                Sciences (NAS) cost estimates used in the Draft TAR underestimated the
                cost for aerodynamic improvements.
                 Broadly, Ford commented in support of the approach to aerodynamic
                improvement modeling in the NPRM, stating that the rule recognized
                potential constraints like consumer needs and preferences regarding
                vehicle styling, vehicle utility, and interior space, by among other
                things, recognizing that the potential for aerodynamic drag differs
                among different vehicle body styles and vehicle classes.\1419\ Ford
                stated that these are major factors considered by customers when
                comparing competing vehicles, and the failure of a manufacturer to
                deliver in these areas can lead to the production of non-competitive,
                poor-selling vehicles.
                ---------------------------------------------------------------------------
                 \1419\ NHTSA-2018-0067-11928.
                ---------------------------------------------------------------------------
                 On the other hand, ICCT claimed that the agencies greatly limited
                the availability of many load reduction technologies (i.e., mass
                reduction improvements, aerodynamic improvements, and rolling
                resistance improvements) by pushing very large amounts of these
                technologies into the 2016 model year baseline fleet, thereby making
                the technologies unavailable for use in future years.\1420\ ICCT
                commented that these improvements in the analysis fleet would
                ostensibly amount to massive efficiency improvements, however, these
                assumed changes were not substantiated as resulting in any test-cycle
                efficiency improvements in the model year 2016 fleet versus the 2015
                fleet. ICCT concluded that the adjusted baseline had been developed and
                presented opaquely, apparently based primarily upon estimations from
                automaker-supplied data, without critical analysis, vetting, or sharing
                of the necessary data to substantiate the changes and real-world
                benefits by the agencies.
                ---------------------------------------------------------------------------
                 \1420\ NHTSA-2018-0067-11741 full comments.
                ---------------------------------------------------------------------------
                 As discussed further in Section VI.C.5.b) AERO drag analysis fleet
                assignments below, the agencies believe the updated analysis fleet
                aerodynamic technology level assignments in the NPRM analysis represent
                an improvement over the MY 2015 assignments in the Draft TAR, as the
                updated assignments are based on precise values, not estimated from
                road load coefficients, and have been corroborated by observed
                improvements on actual production vehicles. Accordingly, the agencies
                carried over the NPRM approach for determining the aerodynamic
                technology levels for the analysis fleet to the final rule.
                a) Aerodynamics Drag Reduction Modeling in the CAFE Model
                 The agencies summarized in the PRIA that the Draft TAR aerodynamic
                improvement levels were binned into two groups, AERO1 and AERO2.
                However, market observations showed that many vehicles had aero
                improvements from 0% to 10%, and some vehicles showed improvements from
                10% to 20%.\1421\ Based on industry feedback and market observations,
                the agencies revised the aerodynamic improvements from two levels in
                the Draft TAR (10% and 20% improvement over the baseline) to four
                levels (5%, 10%, 15% and 20% aerodynamic drag improvement values over
                the baseline). This revision provided the necessary granularity to bin
                the vehicles with different aerodynamic improvements appropriately.
                ---------------------------------------------------------------------------
                 \1421\ PRIA at 437.
                ---------------------------------------------------------------------------
                 ICCT commented that to model appropriately the baseline standards,
                the agencies would need to include increasing use of aerodynamic off-
                cycle technology credits across all companies through 2025. ICCT stated
                that it appeared that the agencies did not use EPA's engineering
                expertise or compliance data, where EPA would be able to advise better
                based on their certification data from the off-cycle program.
                 As discussed further in Sections VI.A and VI.C.8, the NPRM analysis
                carried forward manufacturers' off-cycle fuel consumption improvement
                values (FCIVs) at MY 2016 levels unless an explicitly simulated off-
                cycle technology, like start-stop systems, was added to a vehicle in
                the simulation modeling. Specific to aerodynamic improvements, active
                grille shutters were assumed to be applied at the 20 percent
                aerodynamic improvement (AERO20) level. For the final rule analysis,
                based on the assessment of comments that the application of off-cycle
                technologies in the analysis was too conservative, the agencies agreed
                and increased each manufacturers' application of off-cycle technologies
                so that 10 g/mi of technology was applied by 2023, using an
                extrapolated increase in levels in MYs 2017-2023 based on EPA
                compliance data.\1422\ This approach did not assume any specific mix of
                off-cycle technologies that would be used by manufacturers to achieve
                the 10 g/mi off-cycle improvement, because manufactures currently use a
                variety of technologies, and different manufacturers likely would
                implement unique combinations of technologies. It is expected that
                aerodynamic off-cycle technologies would be included in the mix of off-
                cycle technologies.
                ---------------------------------------------------------------------------
                 \1422\ The 2018 EPA Automotive Trends Report, https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends.
                ---------------------------------------------------------------------------
                 Table VI-132 and Table VI-133 show aerodynamic technologies that
                could be used to achieve 5%, 10%, 15% and 20% aero improvements in
                passenger cars, SUVs, and pickup trucks.\1423\ The agencies developed
                these potential combinations of technologies using aerodynamic data
                from a National Research Council (NRC) of Canada sponsored wind tunnel
                testing program that included an extensive review of production
                vehicles utilizing these technologies, and industry
                comments.1424 1425 These technology combinations are
                intended to show a potential way for a manufacturer to achieve each
                aerodynamic improvement level; however, in the real world,
                manufacturers may implement different combinations of aerodynamic
                technologies to achieve a percentage
                [[Page 24552]]
                improvement over their baseline vehicles.
                ---------------------------------------------------------------------------
                 \1423\ Table 6-67 and Table 6-68 in PRIA.
                 \1424\ Larose, G., Belluz, L., Whittal, I., Belzile, M. et al.,
                ``Evaluation of the Aerodynamics of Drag Reduction Technologies for
                Light-duty Vehicles--a Comprehensive Wind Tunnel Study,'' SAE Int.
                J. Passeng. Cars--Mech. Syst. 9(2):772-784, 2016, https://doi.org/10.4271/2016-01-1613.
                 \1425\ Larose, Guy & Belluz, Leanna & Whittal, Ian & Belzile,
                Marc & Klomp, Ryan & Schmitt, Andreas. (2016). Evaluation of the
                Aerodynamics of Drag Reduction Technologies for Light-duty
                Vehicles--a Comprehensive Wind Tunnel Study. SAE International
                Journal of Passenger Cars--Mechanical Systems. 9. 10.4271/2016-01-
                1613.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.268
                [[Page 24553]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.269
                BILLING CODE 4910-59-C
                b) Aerodynamic Drag Reduction Analysis Fleet Assignments
                 The agencies described in the PRIA that for the 2015 analysis fleet
                used in the Draft TAR, the agencies received Cd values for
                the MY 2015 vehicles' baseline assignments from manufacturers, or used
                estimated Cd values. In response, the industry commented
                that Cd values often varied by measurement approach and,
                therefore, it was important to account for differences in the
                methodologies used to estimate those values. For instance, aerodynamic
                drag coefficients for the same vehicle often vary significantly from
                wind-tunnel to wind-tunnel, complicating cross-comparison and cross-
                referencing.\1426\ The industry commented that, on average, the
                manufacturer-reported Cd values are nine percent lower than
                the values reported by USCAR.\1427\ For reference, USCAR follows the
                SAE J2881 test procedure. However, because Cd values are not
                required to be reported for compliance, manufacturers can and do choose
                different methods to estimate the Cd values. Therefore, the
                industry commented that assigning baseline aerodynamic improvement
                levels should not simply be comparing the lowest reported Cd
                value in a vehicle segment to other reported Cd values. The
                industry commented that such a comparison would not reflect the
                plausible amount of aerodynamic drag improvement that could be
                achieved. Accordingly, the industry suggested that the analysis should
                normalize manufacturer-reported Cd values using SAE J2881.
                ---------------------------------------------------------------------------
                 \1426\ PRIA at 435.
                 \1427\ Footnote in PRIA at 435: FCA Draft TAR comments. Docket
                ID: NHTSA-2016-0068-0082.
                ---------------------------------------------------------------------------
                 The commenters stated manufacturers have the option to use other
                methods (apart from coast down testing) to estimate the Cd
                values such as wind tunnel testing, cross referencing the Cd
                value from other vehicles with similar frontal design and aero
                technologies deployed. Since manufacturers do not have to specify the
                methodology used to estimate the Cd value, the agencies have
                limited capability to make accurate comparisons of the Cd
                value estimates from different testing methods. As a result, the
                agencies determined using average(s) of the fleet provide a better
                estimate of Cd levels than using the lowest Cd
                value in the fleet to assign aerodynamic improvement levels. The
                agencies determined it is appropriate to continue to use the NPRM
                approach for the final rule.
                 The NPRM and final rule analysis used a relative performance
                approach to assign the current aerodynamic technology level to a
                vehicle. Different body styles offer different utility and have varying
                levels of baseline form drag. In addition, frontal area is a major
                factor in aerodynamic forces, and the frontal area varies by vehicle.
                This analysis considered both frontal area and body style as utility
                factors affecting aerodynamic forces; therefore, the analysis assumed
                all reduction in aerodynamic drag forces come from improvement in the
                Cd. Per the process outlined in NHTSA's section of the Draft
                TAR,\1428\ the agencies computed an average Cd for each body
                style segment in the MY 2015 analysis fleet from drag coefficients
                published by manufacturers. By comparing the Cd among
                vehicles sharing body styles, this allowed the agencies to estimate the
                level of aerodynamic improvement present on specific vehicles.
                ---------------------------------------------------------------------------
                 \1428\ Draft TAR at 4-80.
                ---------------------------------------------------------------------------
                 While some small differences existed between the aggregate MY 2015
                and MY 2016 data, the agencies retained the NHTSA-calculated MY 2015
                average Cd as the baseline drag coefficient for nearly all
                body styles. For pickup trucks, the agencies assigned a baseline drag
                coefficient of 0.42, considering that a large portion of the pickups
                sold in MY 2015 already included aerodynamic features assumed for
                advanced levels of aero. The agencies harmonized the Autonomie
                simulation baselines with
                [[Page 24554]]
                the analysis fleet assignment baselines to the fullest extent
                possible.\1429\
                ---------------------------------------------------------------------------
                 \1429\ Often, vehicles assigned to technology classes do not
                perfectly match up with simulated vehicles, but in most cases this
                analysis assumed the aerodynamic effects and other specifications
                were comparable and appropriate for use as proxies.
                ---------------------------------------------------------------------------
                 The agencies assigned levels of aerodynamic technology to the MY
                2016 fleet based on confidential business information submitted by
                manufacturers on aerodynamic drag coefficients, and from other
                information sources such as in product release information. The
                analysis referenced manufacturer-submitted data (if that data was
                supplied), and the agencies took industry comments to Draft TAR into
                account and closely reviewed the manufacturer-submitted Cd
                data. In the few cases that manufacturers did not submit Cd
                values as confidential business information, the agencies estimated the
                Cd based vehicle attributes, design, and aero technologies
                applied to that vehicle. The agencies noted that the Cd
                values reported by some manufacturers showed high levels of improvement
                relative to the previous model year or previous generation. In some
                cases, the agencies contacted the manufacturers to further discuss
                differences in Cd estimation methodologies. Where
                appropriate, the agencies adjusted MY 2016 fleet Cd values
                after consultation with the manufacturers and used these values to
                assign baseline technology levels for each vehicle in the NPRM CAFE
                model simulation.
                 The Alliance commented that the NPRM analysis fleet had more
                appropriately assigned aerodynamic technology levels, and the
                assignments were more accurate than the Draft TAR, where vehicles were
                generally considered to have little aerodynamic improvement technology,
                and the CAFE model would add aerodynamic improvement despite the fact
                that manufacturers had already made significant improvements and there
                was little opportunity remaining for more.\1430\ The Alliance concluded
                that the Draft TAR approach ultimately led the CAFE model to under-
                predict how much powertrain technology was required for compliance. The
                Alliance also commented that it is possible to estimate aerodynamic
                features of a vehicle using road load coefficients, but the process
                requires various assumptions and is not very accurate. The Alliance
                concluded that the agencies' use of CBI to assign initial aerodynamic
                improvement values is an accurate and practical solution to support
                correct baseline assignments.
                ---------------------------------------------------------------------------
                 \1430\ NHTSA-2018-0067-12039 at 136.
                ---------------------------------------------------------------------------
                 Ford commented that the use of actual data, like manufacturer
                confidential information or other sources, to characterize better the
                aerodynamic improvements already incorporated into the baseline fleet
                is a substantial improvement over previous analyses that either assumed
                no aero improvement due to insufficient data, or attempted to infer
                Cd from the road load coefficients.\1431\ Ford stated that
                attempting to infer Cd from road load coefficients is not
                sufficiently accurate for a vehicle-level determination since the
                aerodynamic component of the road load coefficients is inextricably
                confounded with tire, transmission, and other parasitic losses. As part
                of its comments that the proposed rule analysis recognized constraints
                like consumer needs and preferences regarding vehicle styling and
                utility, Ford stated that the baseline Cd for pickup trucks
                properly recognized that these vehicles already include many advanced-
                level aerodynamic technologies. Ford concluded that an accurate
                assessment of the current technological state of the baseline fleet is
                critical to ensuring that the benefits of technological improvements
                are not ``double-counted'' in the modeling.
                ---------------------------------------------------------------------------
                 \1431\ NHTSA-2018-0067-11928.
                ---------------------------------------------------------------------------
                 On the other hand, ICCT commented that the agencies artificially
                limited the availability of aerodynamic technologies in the CAFE model
                in future years by assigning approximately three times as many
                aerodynamic technology packages in the 2016 analysis fleet as they did
                in the 2015 baseline fleet used in the Draft TAR.\1432\ ICCT noted that
                the 2015 Draft TAR fleet had about 8 percent vehicles with one of the
                aerodynamic packages, whereas the NPRM's 2016 fleet had about 53
                percent, and argued that the agencies did not justify the increase with
                data to show that automakers actually deployed the technology. ICCT
                pointed to the agencies' introduction of intermediate aerodynamic
                improvement steps as the justification for the change, which ICCT
                argued ``redistributes the baseline fleet into more advanced
                aerodynamic levels without observing or verifying real-world
                aerodynamic improvements.''
                ---------------------------------------------------------------------------
                 \1432\ NHTSA-2018-0067-11741 full comments.
                ---------------------------------------------------------------------------
                 ICCT argued that if an improvement of this magnitude were true, it
                would be evident in fleet level miles-per-gallon and CO2
                levels (e.g., in EPA's Trends and Manufacturer Performance reports),
                but none of the quantifiable mpg or CO2 benefits that would
                be associated with these additional aerodynamic improvements were
                reflected in any real-world evidence in the model year 2016 fleet. ICCT
                stated that to show the automakers deployed this level of aerodynamic
                improvements, the agencies needed to show data on how these
                improvements are evident in the fleet and delivering benefits.
                Specifically, ICCT stated that the agencies must share the basis for
                any aerodynamic calculation and exact estimated percent improvement
                (rather than binned percentage categories) for each vehicle make and
                model in the baseline and future modeled fleet, and their technical
                justification for each value, arguing that not doing so would obscure
                the agencies' methods. In addition, ICCT stated that the agencies must
                conduct two sensitivity analysis cases that assume that every baseline
                make and model is set to 0 percent aerodynamic improvement and set to
                the previous baseline aerodynamic levels (i.e., from TAR) to
                demonstrate how much the agencies' decision to load up more baseline
                technology affects the compliance scenarios. ICCT concluded that
                because changes in aerodynamic improvement assumptions ``are opaquely
                buried in the agencies' datafiles and unexplained,'' the agencies must
                issue a new regulatory analysis and allow an additional comment period
                for review of the methods and analysis.
                 ACEEE asserted, as part of its comments that the MY 2016 analysis
                fleet assignments appeared to contain errors, that the assignment of
                AERO10 for the MY 2016 Toyota Tundra pickup truck was in error.\1433\
                ACEEE stated that Tundra pickup trucks have had similar specs from MY
                2011 to today, and the Cd for all Tundra models has been
                0.37 or 0.38 for 2WD and 4WD, respectively, since MY 2011. ACEEE noted
                that this is higher than the AERO10 Cd cut off value of
                0.355 for pickups, as shown in the 2016 Draft TAR and referenced in the
                PRIA.
                ---------------------------------------------------------------------------
                 \1433\ NHTSA-2018-0067-12122, at 6.
                ---------------------------------------------------------------------------
                 As described above, the agencies assigned levels of aerodynamic
                technology to the NPRM MY 2016 analysis fleet on a relative basis based
                on confidential business information submitted by the manufacturers on
                aerodynamic drag coefficients and other information sources such as in
                product release information. In addition, based on the Draft TAR
                comments, the agencies verified wherever possible the information
                submitted by manufactures with other sources (product release
                information and cross referencing with vehicles with similar design
                features and aero technologies), and made
                [[Page 24555]]
                changes particularly for vehicles which showed large improvements over
                baseline values. Figure 6-175 in PRIA presented the distribution of
                different levels of aerodynamic drag improvements in MY 2016 vehicle
                fleet in NPRM relative to MY 2015 vehicle fleet used in Draft TAR. The
                distribution shows that 46 percent of the MY 2016 vehicle fleet was
                assigned AERO0 (0 percent improvement), 31 percent of the fleet was
                assigned AERO5 (5% improvement), and 15 percent of the vehicle fleet
                was assigned AERO10 (10 percent improvement). This distribution clearly
                shows that there is substantial opportunity for additional aerodynamic
                drag improvements in the vehicle fleet.
                 Regarding comments by ACEEE on Toyota Tundra pickup trucks, as just
                stated, the agencies used manufacturer submitted information and other
                available information to assign aerodynamic technology levels and the
                agencies applied the same process for all of the manufacturers for the
                NPRM and for the final rule. The agencies did assign AERO10 for some
                Toyota Tundra pickups, but not for all as asserted by ACEEE. Some of
                the Toyota Tundra pickups with 2WD and short bed and crew cab or double
                cab were assigned AERO5 and other configurations were assigned
                AER10.\1434\ For reference, the baseline Cd value used in
                the NPRM for pickups is 0.395; a 5 percent improvement in Cd value is
                0.375 and 10 percent improvement in Cd value is 0.355. The agencies
                considered the ACEEE comment and available information and determined
                the aerodynamic assignments for the Toyota Tundra were reasonable for
                the final rule analysis.
                ---------------------------------------------------------------------------
                 \1434\ The variations could be from coast down testing with
                different powertrains and with different pickup bed length and crew
                cab configurations.
                ---------------------------------------------------------------------------
                 Table VI-134 below shows the percentage aerodynamic drag
                improvement assigned to the MY 2015 (Draft TAR), MY 2016 (NPRM) and MY
                2017 (final rule) analysis fleets. It is clear from this table that
                there is natural progression of aero technologies being adopted and the
                vast majority of the MY 2017 vehicle fleet is at or below AERO10
                (81percent).
                [GRAPHIC] [TIFF OMITTED] TR30AP20.270
                 Moreover, notable aerodynamic improvements have actually been
                observed on production vehicles. As described in PRIA, EPA observed 76
                vehicles at the 2015 North American International Auto Show in Detroit
                (2015 NAIAS).\1435\ EPA's observations showed that manufacturers have
                widely deployed both active and passive aerodynamic drag reduction
                technologies with significant opportunity remaining to apply aero
                technologies further in more optimized fashion as vehicles enter
                redesign cycles in the future.\1436\ Although EPA did not identify the
                aerodynamic drag coefficient values for these vehicles, Figure 6-167 in
                PRIA showed the distribution of some aero technologies identified by
                EPA during this informal survey.
                ---------------------------------------------------------------------------
                 \1435\ PRIA at 432. See also Docket No. EPA-HQ-OAR-2015-0827.
                 \1436\ Draft TAR at 5-363.
                ---------------------------------------------------------------------------
                 The survey showed that wheel dams and underbody panels are the most
                widely used aero technologies, followed by front bumper air dams and
                active grill shutters. Since this survey, many pickup trucks and
                passenger cars have active grill shutters installed to improve
                aerodynamic drag, and to get off-cycle credit. Table 6-67 in PRIA shows
                the ``active grill shutter'' by itself will improve aerodynamic drag
                reduction improvement by 3 percent. Combined with other aero
                technologies, this can improve the aerodynamic drag reduction values
                significantly in pickup trucks and SUVs. As a result, there has been
                overall fleet wide aerodynamic drag reduction improvement; however, the
                above Table VI-134 shows that only 19 percent (13 percent from AERO10,
                5 percent from AERO15 and 1 percent from AERO20) of the MY 2017 vehicle
                fleet has aerodynamic drag reduction improvement greater than 10
                percent. This shows that there is significant opportunity for the
                vehicle fleet to improve aero technologies by MY 2025.
                 The agencies also described examples of how production vehicles in
                different technology classes improved aerodynamic drag reduction values
                relative to their previous generation model since the 2012 final
                rule.\1437\ The PRIA described how aerodynamic technologies were being
                deployed on production vehicles, using the MY 2015 Nissan Murano and MY
                2015 Ford F150 as examples. For example, MY 2015 Ford F150 has the
                passive and active aerodynamic technologies as shown in Table VI-135.
                ---------------------------------------------------------------------------
                 \1437\ PRIA at 433.
                ---------------------------------------------------------------------------
                 The air curtain technology in the MY 2015 F150 guides the air flow
                across the front wheels to reduce wind turbulence.\1438\ For reference,
                the wind tunnel testing by NRC of the MY 2015 Ford F150 showed a drag
                coefficient value of 0.37 while the coast down testing by EPA pegged
                the drag coefficient value between 0.35 to 0.40 depending on the type
                of powertrain, cab and cargo box combination. The prior generation F150
                was released in 2008 as a MY 2009 and this vehicle had
                [[Page 24556]]
                very few aerodynamic technologies applied. The agencies do not have the
                MY 2009 Cd value to estimate the percentage improvement.
                Since the F150 also included significant light weighting and powertrain
                improvements including a downsized turbocharged engine, the
                effectiveness improvement attributable to aerodynamic technologies is
                uncertain.
                ---------------------------------------------------------------------------
                 \1438\ Ford, How Air Curtains on F-150 Help Reduce Aerodynamic
                Drag and Aid Fuel Efficiency (July 15, 2015), https://media.ford.com/content/fordmedia/fna/us/en/news/2015/07/15/how-air-curtains-on-f-150-help-reduce-aerodynamic-drag.html.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.271
                 The Nissan Murano is an example of a mid-size SUV with greater than
                fifteen percent improvement in aerodynamic drag values compared to the
                previous generation. The SAE paper published in 2015 outlines the
                specifics of aerodynamics in the Nissan Murano,\1439\ and they include
                those listed in Table VI-136 below.
                ---------------------------------------------------------------------------
                 \1439\ Arai, M., Tone, K., Taniguchi, K., Murakami, M. et al.,
                ``Development of the Aerodynamics of the New Nissan Murano,'' SAE
                Technical Paper 2015-01-1542, 2015, https://doi.org/10.4271/2015-01-1542.
                ---------------------------------------------------------------------------
                 The exterior of this vehicle was completely redesigned from the MY
                2013-2014 generation with the goal of minimizing aerodynamic drag by
                combining passive aerodynamic devices with an optimized vehicle shape.
                The primary passive devices employed include optimization of the rear
                end shape to reduce rear end drag, and addition of a large front
                spoiler to reduce underbody air flow and redirect it toward the roof of
                the vehicle, thus augmenting the rear end drag improvements. Other
                passive improvements include plastic fillet moldings at the wheel
                arches, raising the rear edge of the hood, shaping the windshield
                molding and front pillars, engine under-cover and floor cover, and air
                deflectors at the rear wheel wells. An active lower grille shutter also
                redirects air over the body when closed. Together, these measures for
                the MY 2015 model achieved a drag coefficient of 0.31, representing a
                16 to 17 percent improvement over the 0.37 Cd of the
                previous model.
                [[Page 24557]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.272
                [[Page 24558]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.273
                BILLING CODE 4910-59-C
                 A combination of a slightly lighter MY 2015 Nissan Murano (on
                average lighter by 94 lbs. considering all trim levels), relative to
                the previous generation, and engine improvements (comparing 3.5L V6 in
                MY 2014 to 3.5L V6 in MY 2015), and transmission improvements resulted
                in an overall improvement in fuel economy.\1440\ Accordingly, the real-
                world fuel economy improvement directly attributable to the package of
                aerodynamic technologies included on either vehicle is uncertain, as
                each vehicle included other fuel economy improving technologies along
                with the improvements in aerodynamic technologies.
                ---------------------------------------------------------------------------
                 \1440\ https://www.fueleconomy.gov/feg/Find.do?action=sbs&id=34457&id=37198 (last visited 12.12.2019) shows
                20 mpg (combined) in MY2014 Nissan Murano (3.5L VQ35DE V6 with
                Variable gear ratio transmission) and 24 mpg (combined in MY2015
                Nissan Murano (3.5L VQ35DE V6 with Automatic AV S7 transmission)).
                ---------------------------------------------------------------------------
                 The agencies considered a sensitivity case that assumed no mass
                reduction, rolling resistance, or aerodynamic improvements had been
                made to the MY 2017 fleet (i.e., setting all vehicle road levels to
                zero--MRO, AERO and ROLL0), in response to ICCT's comment. While this
                is an unrealistic characterization of the initial fleet, the agencies
                conducted a sensitivity analysis to understand any affect it may have
                on technology penetration along other paths (e.g., engine and hybrid
                technology). Under the CAFE program, the sensitivity analysis shows a
                slight decrease in reliance on engine technologies (HCR engines,
                turbocharge engines, and engines utilizing cylinder deactivation) and
                hybridization (strong hybrids and plug-in hybrids) in the baseline
                (relative to the central analysis). The consequence of this shift to
                reliance on lower-level road load technologies is a reduction in
                compliance cost in the baseline of about $300 per vehicle (in MY 2026).
                As a result, cost savings in the preferred alternative are reduced by
                about $200 per vehicle. Under the CO2 program, the general
                trend in technology shift is less dramatic (though the change in BEVs
                is larger) than the CAFE results. The cost change is also comparable,
                but slightly smaller ($200 per vehicle in the baseline) than the CAFE
                program results. Cost savings under the preferred alternative are
                further reduced by about $100. With the lower technology costs in all
                cases, the consumer payback periods decreased as well. These results
                are consistent with the approach taken by manufacturers who have
                already deployed many of the low-level road load reduction
                opportunities to improve fuel economy.
                 Second, as discussed above, EPA's baseline aerodynamic levels in
                the Draft TAR were based on road load coefficients, leading to baseline
                assignments that were not accurate. In the NPRM, the agencies discussed
                in the tradeoffs between building the analysis fleet using confidential
                information from manufacturers and publicly available data on the
                vehicles.\1441\ In the case of drag coefficient values, which cannot be
                gleaned from publicly available information, except in cases where a
                manufacturer chooses to publicly release that data, or by simply
                observing a vehicle, the agencies decided that the improved accuracy
                associated with using manufacturer-provided Cd values
                outweighed the benefits of using publicly releasable Cd
                estimates based on road load coefficients, especially as manufacturer-
                provided Cd values are only used to assign initial
                aerodynamic improvement levels relative to Cd values for
                each body style segment in the analysis fleet.
                ---------------------------------------------------------------------------
                 \1441\ 83 FR 43004.
                ---------------------------------------------------------------------------
                 In addition, manufacturers had submitted comments that the Draft
                TAR approach to baseline fleet assignments had underestimated
                technology already present on vehicles, leading the analysis to apply
                more aerodynamic drag reduction technology than could be applied in the
                real world. In response to those comments, as described in the Proposed
                Determination TSD, EPA stated that they ``agree[ ] with the commenters
                that it is appropriate to account for aerodynamic drag reductions
                already present in the baseline fleet in order to avoid overestimating
                the amount of additional improvement that can be achieved at a given
                cost.'' \1442\ Accordingly, EPA ``applied some level of aerodynamic
                drag reduction to a significant portion of the MY2015 baseline fleet.''
                \1443\ Consequently, the agencies believe that ICCT's statement that if
                aerodynamic improvements between the MY 2015 analysis fleet used in the
                Draft TAR and the MY 2016 analysis fleet were true it would be evident
                in the fleet is incorrect. It is inappropriate to compare the Draft TAR
                MY 2015 analysis fleet, which notably included too few aerodynamic
                technology assignments, with the fleet's achieved fuel economy in the
                real world. The agencies disagree
                [[Page 24559]]
                with ICCT that the availability of aerodynamic technologies was
                artificially limited by appropriately assigning baseline aerodynamic
                technology levels in the analysis fleet.
                ---------------------------------------------------------------------------
                 \1442\ Proposed Determination TSD at 2-406.
                 \1443\ Proposed Determination TSD at 2-408.
                ---------------------------------------------------------------------------
                 This also relates to ICCT's comment that the agencies must share
                the basis for any aerodynamic calculation and exact estimated percent
                improvement (rather than binned percentage categories) for each vehicle
                make and model in the baseline and future modeled fleet, and their
                technical justification for each value. As discussed above, the
                agencies shared the relative performance approach methodology for
                assigning baseline aerodynamic levels to vehicles in the analysis fleet
                in detail in the PRIA,\1444\ and this approach was the basis for the
                aerodynamic calculation performed for every vehicle make and model in
                the analysis fleet. The agencies provided the summary of aerodynamic
                drag coefficients (including averages for MY 2016 vehicles) by vehicle
                body style,\1445\ and the baseline aerodynamic improvement assignments
                for each vehicle model were included in the
                2018_NPRM_market_inputs_ref.xlsx. In addition, because aerodynamic drag
                information from manufacturers is provided as confidential business
                information, the agencies are unable to disclose that specific
                information. However, as discussed above, the agencies are closely
                examining the data provided and comparing it to other available
                information to assess the best estimate for aerodynamic technology for
                each vehicle in the analysis fleet.
                ---------------------------------------------------------------------------
                 \1444\ PRIA at 441.
                 \1445\ PRIA at 443.
                ---------------------------------------------------------------------------
                 For these reasons, the agencies continued to use the NPRM
                methodology to assign aerodynamic drag reduction improvements for the
                MY 2017 vehicle fleet for this final rule.
                c) Aerodynamic Drag Technology Adoption Features
                 As discussed above, the agencies used a relative performance
                approach to assign current aerodynamic technology level to a vehicle.
                For some body styles with different utility, such as pickup trucks,
                SUVs and minivans, frontal area can vary, and this can affect the
                overall aerodynamic drag forces. In order to maintain vehicle utility
                and functionality related to passenger space and cargo space, the
                agencies assumed all technologies that improve aerodynamic drag forces
                would do so through reducing the Cd while maintaining
                frontal area.
                 In the NPRM, the agencies noted that the Proposed Determination
                analysis assumed that some vehicles from all body styles could (and
                would) reduce aerodynamic forces by 20 percent, which in some cases led
                to future pickup trucks having aerodynamic drag coefficients better
                than some of today's typical cars, if frontal area were held constant
                in order to preserve interior space and cargo space. The agencies
                further noted that for some vehicle types, there was limited practical
                capability to significantly improve aerodynamic drag coefficients over
                baseline levels. In those cases, the agencies deemed the most advanced
                levels of aerodynamic drag simulated as not technically practicable
                given the need to maintain vehicle functionality and utility, such as
                interior volume, cargo area, and ground clearance.
                 The industry had also commented in response to EPA's Proposed
                Determination on the difficulty to achieve AERO20 improvements for
                certain body styles. In the NPRM, the agencies considered the industry
                comments along with the observations made in the MY 2016 fleet, and
                tentatively determined the maximum feasible improvement in
                Cd that could be achieved for pickup trucks is AERO15.\1446\
                Similarly, the agencies determined the maximum feasible improvement in
                Cd that could be achieved for minivans is AERO10. Next, the
                NPRM analysis did not apply 15 percent or 20 percent aerodynamic drag
                coefficient reduction to cars and SUVs with more than 405 horsepower.
                The agencies noted that many high-performance vehicles already include
                advanced aerodynamic features despite middling aerodynamic drag
                coefficients. In these high-performance vehicle cases, the agencies
                recognized that manufacturers tune aerodynamic features to provide
                desirable downforce at high speeds and to provide sufficient cooling
                for the powertrain, and, therefore, manufacturers may have limited
                ability to improve aerodynamic drag coefficients for high performance
                vehicles with internal combustion engines without reducing horsepower.
                Accordingly, the agencies did not allow application of AERO15 and
                AERO20 technology for all vehicles with more than 405 HP. Approximately
                400,000 units of volume in the MY 2016 market data file included
                limited application of aerodynamic technologies because of vehicle
                performance. The agencies sought comment on limiting the Cd
                improvement in these circumstances.
                ---------------------------------------------------------------------------
                 \1446\ The agencies noted in the NPRM that although ANL created
                full-vehicle simulations for trucks with 20 percent drag reduction,
                those simulations were not used in the CAFE modeling. The agencies
                concluded that level of drag reduction was likely not
                technologically feasible with today's technology, and the analysis
                accordingly restricted the application of advanced levels of
                aerodynamics in some instances, such as in that case, due to
                bodystyle form drag limitations.
                ---------------------------------------------------------------------------
                 Ford commented in support of the agencies' decision to limit the
                application of AERO20 on pickup trucks, noting that limiting AERO20 on
                pickups is appropriate given the high inherent form drag associated
                with pickups' aerodynamic profile.\1447\
                ---------------------------------------------------------------------------
                 \1447\ NHTSA-2018-0067-11928.
                ---------------------------------------------------------------------------
                 CARB commented that the agencies excluded AERO20 inconsistently
                across the fleet, noting that while some of the restrictions may be
                valid, the broad rule the agencies used resulted in technology being
                inappropriately excluded from too many vehicles.\1448\ Specifically,
                CARB took issue with the majority of luxury sedans and SUVs being
                excluded from AERO20 because they had high horsepower engines, while
                the agencies did assign AERO20 to vehicles like the Tesla Model S and
                Model X SUVs, which have horsepower in excess of 405. CARB stated that
                while electrification provides a higher motivation to minimize road
                load through technologies such as aerodynamic reductions, implementing
                AERO20 reductions on high horsepower sedans and SUVs is clearly
                feasible and should not be artificially restricted in the CAFE model.
                ---------------------------------------------------------------------------
                 \1448\ NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 In addressing these comments, the agencies considered the relative
                cooling requirements for all electric powertrains and for high
                performance internal combustion engine powertrains since airflow
                diverted for cooling adversely impacts a vehicle's Cd. The
                peak heat rejection and engine cooling needs for high performance
                internal combustion engines is significantly higher than for all
                electric powertrains. Internal combustion engines convert a lower
                percentage of energy contained in gasoline into mechanical work (and
                other useful work, such as lighting and sound), and the energy not
                converted into mechanical work (or other useful work) is converted into
                heat. A significant amount of the waste heat must be handled by the
                cooling systems. Battery electric vehicles convert most of the
                electrical energy stored in the battery into mechanical work and other
                useful work, and therefore convert less energy into heat that must be
                handled by the cooling system. Also, electric powertrains can provide a
                degree of electric braking, whereas internal combustion engines
                exclusively use friction braking, which generates heat and requires
                greater cooling,
                [[Page 24560]]
                particularly on vehicles with substantial braking performance
                capabilities. In the case of high-performance BEVs, since the cooling
                needs are not as demanding as with high-performance vehicles that use
                internal combustion engines, manufacturers can (and do, as can be
                observed in the fleet) apply higher levels of aerodynamic technologies.
                The agencies believe it is appropriate to account for these differences
                in considering the amount of aerodynamic improvement that can be
                implemented, and determined there are valid technical reasons for
                allowing BEVs with greater than 405 horsepower to adopt AERO20
                technology.
                d) Aerodynamic Drag Technology Effectiveness
                 The NPRM analysis included four levels of aerodynamic improvements,
                AERO5, AERO10, AERO15, and AERO20, representing 5, 10, 15, and 20
                percent Cd improvements, respectively. Notably, the NPRM
                analysis assumed that aerodynamic drag reduction could only come from
                reduction in the aerodynamic drag coefficient and not from reduction of
                frontal area, to maintain vehicle functionality and utility, such as
                passenger space, ingress/egress ergonomics, and cargo space.\1449\
                ---------------------------------------------------------------------------
                 \1449\ 83 FR 43047.
                ---------------------------------------------------------------------------
                 Ford commented in support of the agencies' decision to consider the
                frontal area and body style as ``utility factors'' and requiring that
                aerodynamic improvements come from reductions in Coefficient of Drag
                (Cd) and not from reductions in frontal area.\1450\
                ---------------------------------------------------------------------------
                 \1450\ NHTSA-2018-0067-11928.
                ---------------------------------------------------------------------------
                 CBD commented that EPA staff had critiqued NHTSA's characterization
                of research on aerodynamic drag coefficients and the NPRM did not
                appear to incorporate or respond to this input.1451 1452
                Specifically, CBD stated that EPA staff had commented in response to
                the characterization that ``[f]or some bodystyles, the agencies have no
                evidence that manufacturers may be able to achieve 15 percent or 20
                percent aerodynamic drag coefficient reduction relative to baseline
                (for instance, with pickup trucks'' and noted that ``[i]n the past, EPA
                has assigned aero tech in the baseline relative to a ``Null'' and then
                applied drag reduction level against that Null in order to ensure that
                the maximum aero level (i.e., 15 or 20 percent) would always be
                achievable for all body styles.'' This comment reflects deliberative,
                in-process input from EPA staff. In fact, the NPRM text was developed
                by the agencies with the benefit of this and other input from EPA
                staff, and the NPRM clarified that reducing frontal area would likely
                degrade other utility features like interior volume or ingress/egress.
                ---------------------------------------------------------------------------
                 \1451\ NHTSA-2018-0067-12000, at 188.
                 \1452\ Docket No. EPA-HQ-OAR-2018-0283-0453, June 29, 2018
                Comments at 93.
                ---------------------------------------------------------------------------
                 CARB commented, as part of its broader comments, that the agencies'
                effectiveness values were reduced relative to what EPA's LPM
                calculated, that the benefits of aerodynamic improvements were
                underestimated.\1453\ Specifically, CARB cited the H-D Systems
                comparison of LPM benefits for AERO10 and AERO20 of 2.1 percent and 4.3
                percent, respectively, compared with Autonomie benefits of 1.51 percent
                and 3.03 percent, respectively, and stated that the agencies' analysis
                provided no description or cited any new data or evidence as to why
                they reduced the projected assumptions compared to what EPA's Lumped
                Parameter Model calculated.
                ---------------------------------------------------------------------------
                 \1453\ NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 HDS also commented that the Autonomie modeling assumed no engine
                change when aerodynamic drag and rolling resistance reductions were
                implemented, as well as no changes to the transmission gear ratios and
                axle ratios, which vary by transmission type but not by the tractive
                load.\1454\ HDS stated that the EPA ALPHA model adjusted for this
                effect, which accounted for the difference in technology effectiveness
                estimates that HDS characterized between the Draft TAR and NPRM. HDS
                provided a ``correct estimate'' for AERO20 effectiveness improvements
                of 4.3 percent, with the justification that there was no gear/axle
                ratio adjustment in the Autonomie analysis.
                ---------------------------------------------------------------------------
                 \1454\ NHTSA-2018-0067-11985.
                ---------------------------------------------------------------------------
                 In response to HDS's comment, the Alliance submitted supplemental
                comments questioning the extent to which aerodynamics (and changes in
                top gear ratio) affect performance metrics held constant in the
                analysis, like low- and high-speed acceleration performance and
                gradeability.\1455\ The Alliance cited a study for the proposition that
                vehicle acceleration is most influenced by engine power and weight, and
                also that bodystyle differences have a lesser impact on acceleration
                performance. The Alliance further commented that ``[r]egarding changes
                in top gear ratios in response to aerodynamic changes, the Alliance is
                not aware of any examples in which a top gear ratio was changed solely
                due to aerodynamic improvements. There may be examples where a
                vehicle's top gear ratio was changed at the same time aerodynamic
                changes were made, but such changes would be made in response to the
                cumulative changes across the entire vehicle, not just aerodynamic
                improvements.'' The Alliance concluded that ``[t]here are also
                practical manufacturing and investment constraints which limit the
                potential for applying engine changes in response to improved vehicle
                aerodynamics,'' citing the agencies decision to only resize engines
                with significant design changes, to account for product complexity and
                economies of scale.
                ---------------------------------------------------------------------------
                 \1455\ NHTSA-2018-0067-12385, at 31-32.
                ---------------------------------------------------------------------------
                 In response to the Alliance's supplemental comment, HDS submitted
                supplemental comments stating that ``[d]rag reduction is usually
                accomplished when a vehicle body is redesigned, so gear and axle ratios
                are typically re-optimized for the entire set of changes, but these
                changes include the drag reduction.'' \1456\ HDS commented that the
                Alliance's comments acknowledged that calibration changes are made in
                response to tractive load changes, while the Autonomie analysis
                recalibrates the powertrain in response only to large mass reduction
                improvements, and not any other vehicle changes that reduce tractive
                load, like aerodynamic improvements, even when those changes would
                result in a greater tractive load reduction than a 10 percent mass
                reduction. HDS reiterated its statement that ``[i]n the real world (and
                as captured in EPA's prior ALPHA model), automakers typically alter
                many vehicle attributes affecting tractive load simultaneously,
                including aerodynamics,'' and the Autonomie outputs underrepresent the
                benefit of tractive load reduction strategies by not optimizing engine
                efficiency after most changes in tractive load because the model
                employees fixed shift points, gear ratios, and axle ratios when drag or
                tire rolling resistance is reduced.
                ---------------------------------------------------------------------------
                 \1456\ NHTSA-2018-0067-12395, at 4-5.
                ---------------------------------------------------------------------------
                 Regarding the first set of comments that the aerodynamic
                effectiveness values were reduced from EPA's values presented in the
                Draft TAR, that results from differences in the two modeling
                approaches. As discussed above, for this analysis the agencies decided
                that aerodynamic drag reduction could only come from reduction in the
                aerodynamic drag coefficient, and not from a reduction in vehicle
                frontal area, at least without reducing other attributes of the
                vehicle. EPA's process for assigning road load technologies to baseline
                vehicles used road load coefficients from coast downs, which aggregated
                individual aero, mass and tire reduction technologies. In contrast,
                [[Page 24561]]
                the CAFE Model and Autonomie used individually assigned road load
                technologies for each vehicle to appropriately assign initial road load
                and to appropriately capture benefits of subsequent individual road
                load technologies. The differences in using road load coefficients from
                coast downs and individually isolating the improvements from existing
                and future road load technologies in the Autonomie modeling resulted in
                the differences noted by commenters. And so, the resulting
                effectiveness from the incremental adoption of individual technologies
                to a newer analysis fleet will have different result than what was
                estimated by the previous analyses. For further discussion of the
                analysis fleet see Section VI.B.1.
                 In Section VI.B.3 Tech Effectiveness and Modeling and Section
                VI.C.2 Transmissions, the agencies provide a full discussion of the
                issues associated with assuming the engine and transmission can be
                optimized for every combination of technologies. It would be
                unreasonable and unaffordable to resize powertrains, including engines
                and transmission and axle ratios, for every unique combination of
                technologies, and exceedingly so for every unique combination
                technologies across every vehicle model due to the extreme
                manufacturing complexity that would be required to do so. Product
                complexity and economies of scale are real, and in the NPRM, engine
                resizing was limited to specific incremental technology changes that
                would typically be associated with a major vehicle or engine
                redesign.\1457\ As noted by HDS, the EPA Draft TAR and Proposed
                Determination analyses adjusted the effectiveness of every technology
                combination, including for aerodynamics technologies, assuming
                performance could be held constant for every combination. However,
                those analyses did not recognize or account for the extreme complexity
                nor the associated costs for that impractical assumption. The NPRM and
                final rule analyses account for these real-world practicalities and
                constraints, and doing so explains some of the effectiveness and cost
                differences between the Draft TAR/Proposed Determination and the NPRM/
                final rule. The agencies believe the NPRM and the final rule approach
                appropriately resizes powertrain components for specific incremental
                technology changes that would typically be associated with a major
                vehicle or engine redesign.
                ---------------------------------------------------------------------------
                 \1457\ See 83 FR 43027 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 For the NPRM, and carried into the final rule analysis, Autonomie
                simulates all road load conditions (e.g., MR, AERO, and ROLL technology
                levels) for each engine and transmission combination. In addition,
                engines are resized for appropriate specific technology changes that
                would be associated with a major vehicle or engine redesign. Also, as
                discussed further in Section VI.C.2 Transmissions, many commenters
                seemed to conflate the practice in the analysis of using a common
                (same) gear set across vehicle configurations (to address manufacturing
                complexity) with using the same shift maps. As commenters stated, they
                assumed the same shift maps were applied across vehicle models.
                However, the shift initializer routine was run for every unique
                Autonomie full vehicle model configuration and generated customized
                shifting maps. The algorithms' optimization was designed to balance
                minimization of energy consumption and vehicle performance. This
                balance was necessary to achieve the best fuel efficiency while
                maintaining customer acceptability by meeting performance neutrality
                requirements. The agencies believe the level of optimization of engine
                size, transmissions, gear ratios and shift schedules reasonably
                approximate what is achievable and what manufacturers actually do.
                 Figure VI-47 below shows the range effectiveness used for AERO
                technologies for the NPRM analysis.
                BILLING CODE 4910-59-P
                [[Page 24562]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.274
                 Figure VI-48 below shows the range of aero effectiveness used for
                the final rule analysis.
                [[Page 24563]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.276
                BILLING CODE 4910-59-C
                e) Aerodynamic Drag Technology Cost
                 For the Draft TAR, the agencies relied on the 2015 NAS report to
                estimate the cost of AERO1 and AERO2 levels of aerodynamic drag
                coefficient improvements. The agencies received several comments
                related to the cost assumptions used in the Draft TAR, mainly that they
                were too low to meet AERO1 and AERO2 levels. The industry submitted
                confidential business information on the costs of passive aerodynamic
                technologies needed to achieve AERO1 (10 percent improvement in drag
                improvement), which showed a significantly higher estimated costs than
                assumed for the Draft TAR. Similarly, the industry submitted
                confidential business information on the costs of active aerodynamic
                technologies, including some high cost technologies. The industry also
                commented that some active aerodynamic technologies could only be
                implemented during vehicle redesigns and not during a mid-cycle vehicle
                refresh.
                 The agencies considered these comments and performed additional
                research to assess the costs for passive and active aerodynamic
                technologies. The agencies revised the cost estimates for the NPRM,
                based in part on confidential information from the automotive industry,
                and from the agencies' own assessment of manufacturing costs for
                specific aerodynamic technologies from available sources. In general,
                the NPRM cost estimates were higher than Draft TAR cost estimates. The
                agencies included a high-level discussion in the PRIA that the cost to
                achieve AERO5 is relatively low, as most of the improvements can be
                made through body styling changes. The cost to achieve AERO10 is higher
                than AERO5, due to the addition of several passive aero technologies,
                and the cost to achieve AERO15 and AERO20 is higher than AERO10 due to
                use of both passive and active aero technologies.
                 The agencies did not receive any comments on the costs of
                aerodynamic improvements, and accordingly, for the final rule, as shown
                in Table VI-137 and Table VI-138 below, the agencies used the same
                aerodynamic improvement costs presented in NPRM.
                [[Page 24564]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.277
                [GRAPHIC] [TIFF OMITTED] TR30AP20.278
                6. Tire Rolling Resistance
                 Tire rolling resistance is a road load force that arises primarily
                from the energy dissipated by elastic deformation of the tires as they
                roll. Tire design characteristics (for example, materials,
                construction, and tread design) have a strong influence on the amount
                and type of deformation and the energy it dissipates. Designers can
                select these characteristics to minimize rolling resistance. However,
                these characteristics may also influence other performance attributes,
                such as durability, wet and dry traction, handling, and ride comfort.
                 Low rolling resistance tires are increasingly specified by OEMs in
                new vehicles and are also increasingly available from aftermarket tire
                vendors. They commonly include attributes such as higher inflation
                pressure, material changes, tire construction optimized for lower
                hysteresis, geometry changes (e.g., reduced aspect ratios), and reduced
                sidewall and tread deflection. These changes are commonly accompanied
                by additional changes to vehicle suspension tuning and/or suspension
                design to mitigate any potential impact on other performance attributes
                of the vehicle.
                 Lower-rolling-resistance tires have characteristics that reduce
                frictional losses associated with the energy dissipated mainly in the
                deformation of the tires under load, thereby improving fuel economy and
                reducing CO2 emissions. The agencies considered two levels
                of improvement for low rolling resistance tires in the analysis: The
                first level of low rolling resistance tires considered reduced rolling
                resistance 10 percent from an industry-average baseline, while the
                second level reduced rolling resistance 20 percent from the baseline.
                 Walter Kreucher commented that the agencies should eliminate low
                rolling resistance tires from the list of viable technologies, in
                recognition of the safety impacts of low rolling resistance tires in
                relation to stopping distance and accident rates.\1458\ Separately, Mr.
                Kreucher argued that the model should reflect the safety impact of low
                rolling resistance tires.
                ---------------------------------------------------------------------------
                 \1458\ NHTSA-2018-0067-0444.
                ---------------------------------------------------------------------------
                 The agencies have been following the industry developments and
                trends in application of rolling resistance technologies to light duty
                vehicles. As stated in the NAP special report on Tires and Passenger
                Vehicle Fuel Economy,\1459\ cited by Mr. Kreucher, national crash data
                does not provide data about tire structural failures specifically
                related to tire rolling resistance, because the rolling resistance of a
                tire at a crash scene cannot be determined. However, other metrics like
                brake performance compliance test data
                [[Page 24565]]
                are helpful to show trends like that stopping distance has not changed
                in the last ten years,\1460\ during which time many manufacturers have
                installed low rolling resistance tires in their fleet--meaning that
                manufacturers were successful in improving rolling resistance while
                maintaining stopping distances through tire design, tire materials,
                and/or braking system improvements. In addition, NHTSA has addressed
                other tire-related issues through rulemaking,\1461\ and continues to
                research tire problems such as blowouts, flat tires, tire or wheel
                deficiency, tire or wheel failure, and tire degradation.\1462\ However,
                there are currently no data connecting low rolling resistance tires to
                accident or fatality rates.
                ---------------------------------------------------------------------------
                 \1459\ Tires and Passenger Vehicle Fuel Economy: Informing
                Consumers, Improving Performance--Special Report 286 (2006),
                available at https://www.nap.edu/read/11620/chapter/6.
                 \1460\ https://one.nhtsa.gov/cars/problems/comply/index.cfm.
                 \1461\ 49 CFR 571.138, Tire pressure monitoring systems.
                 \1462\ Tire-Related Factors in the Pre-Crash Phase, DOT HS 811
                617 (April 2012), available at https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/811617.
                ---------------------------------------------------------------------------
                 With better tire design, tire compound formulations and improved
                tread design, tire manufacturers have tools to balance stopping
                distance and reduced rolling resistance. As stated in one article
                referenced by Mr. Kreucher, tire manufacturers can use ``higher
                performance materials in the tread compound, more silica as reinforcing
                fillers and advanced tread design features'' to mitigate issues related
                to stopping distance.\1463\ The agencies do not believe that there is
                sufficient data or other information to support removing low rolling
                resistance tires as a viable technology considered in the CAFE and
                CO2 analysis at this time.
                ---------------------------------------------------------------------------
                 \1463\ Jesse Snyder, A big fuel saver: Easy-rolling tires (but
                watch braking) (July 21, 2008), https://www.autonews.com/article/20080721/OEM01/307219960/a-big-fuel-saver-easy-rolling-tires-but-watch-braking. Last visited December 3, 2019.
                ---------------------------------------------------------------------------
                 HDS argued, as discussed further below, that based on available
                data on current vehicle models and the likely possibility that there
                would be additional tire improvements over the next decade, the
                agencies should consider ROLL30 technology, or a 30 percent reduction
                of tire rolling resistance over the baseline.\1464\
                ---------------------------------------------------------------------------
                 \1464\ NHTSA-2018-0067-11985.
                ---------------------------------------------------------------------------
                 As stated in Joint TSD for the 2017-2025 final rule, tire
                technologies that enable rolling resistance improvements of 10 and 20
                percent have been in existence for many years.\1465\ Achieving
                improvements of up to 20 percent involves optimizing and integrating
                multiple technologies, with a primary contributor being the adoption of
                a silica tread technology. Tire suppliers have indicated that
                additional innovations are necessary to achieve the next level of low
                rolling resistance technology on a commercial basis, such as
                improvements in material to retain tire pressure, tread design to
                manage both stopping distance and wet traction, and development of
                carbon black material for low rolling resistance without the use of
                silica to reduce cost and weight.\1466\ The agencies are continuously
                monitoring these and other tire technology improvements. The agencies
                believe that the tire industry is in the process of moving automotive
                manufacturers towards the first level of low rolling resistance
                technology across the vehicle fleet (10 percent reduction in rolling
                resistance), and that 20 percent improvement is achievable in the
                rulemaking timeframe. However, the agencies believe that at this time,
                the emerging tire technologies that would achieve 30 percent
                improvement in rolling resistance, like changing tire profile,
                strengthening tire walls, or adopting improved tires along with active
                chassis control,\1467\ among other technologies, will not be available
                for commercial adoption in the fleet during the rulemaking timeframe.
                As a result, the agencies decided not to incorporate 30 percent
                reduction in rolling resistance technology for this final rule.
                ---------------------------------------------------------------------------
                 \1465\ EPA-420-R-12-901, at page 3-210.
                 \1466\ Assessment of Fuel Economy Technologies for Light-Duty
                Vehicles (2011) at page 103.
                 \1467\ Mohammad Mehdi Davari, Rolling resistance and energy loss
                in tyres (May 20, 2015), available at https://www.sveafordon.com/media/42060/SVEA-Presentation_Davari_public.pdf. Last visited
                December 30, 2019.
                ---------------------------------------------------------------------------
                a) Rolling Resistance Modeling in the CAFE Model
                 The two levels of rolling resistance technology considered in the
                analysis include ROLL10 and ROLL20, which represent a 10 percent and 20
                percent rolling resistance reduction from the baseline (ROLL0),
                respectively.
                 To understand the following discussions about rolling resistance
                analysis fleet assignments and effectiveness values, it is important to
                understand how the agencies developed the baseline value (ROLL0) used
                in prior analyses, and how the agencies developed the baseline value
                used in the NPRM and final rule. In the Draft TAR, the agencies used
                unique baseline rolling resistance coefficients for each vehicle class.
                Specifically, the compact car class value was 0.0075, the midsize car
                value was 0.008, the small SUV value was 0.0084, the midsize SUV value
                was 0.0084, and the pickup truck value was 0.009. The PRIA described
                that since the Draft TAR, the agencies had reassessed rolling
                resistance values for contemporary tires through discussions with
                vehicle manufacturers, tire manufactures, and independent bench
                testing. Based on a thorough review of confidential business
                information submitted by industry, and a review of other literature,
                including the CARB/CONTROLTEC study mentioned below, the baseline
                rolling resistance coefficient for all vehicle classes was updated to
                0.009 for the NPRM analysis. The agencies concluded that the updated
                baseline value brought the NPRM simulations into better alignment with
                tires in the MY 2016 analysis fleet. The agencies also discussed that
                updated value was consistent with the findings of the CONTROLTEC study
                on vehicle road loads, sponsored by CARB.\1468\ The following figure
                shows the distribution of estimated tire rolling resistance coefficient
                values for the 1,358 MY 2014 vehicles evaluated in the CONTROLTEC/CARB
                study.
                ---------------------------------------------------------------------------
                 \1468\ Technical Analysis of Vehicle Load Reduction Potential
                for Advanced Clean Cars, https://www.arb.ca.gov/research/apr/past/13-313.pdf, page 39.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24566]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.279
                BILLING CODE 4910-59-C
                 ICCT commented that it was ``quite confusing and perhaps
                troubling'' that the agencies adopted a higher average rolling
                resistance coefficient than that of the Draft TAR, ``as it would imply
                that the fleet rolling resistance got worse, but the agencies are
                deciding to provide baseline credit as if there was more rolling
                resistance technology deployed.'' \1469\ ICCT stated that the change
                appeared to be attributed to the agencies' use of CBI on tire rolling
                resistance received since the Draft TAR.
                ---------------------------------------------------------------------------
                 \1469\ NHTSA-2018-0067-11741 full comments.
                ---------------------------------------------------------------------------
                 As described in the PRIA, the values used in the Draft TAR
                represented the ``Best in Class'' values in each of the vehicle classes
                and this did not necessarily reflect the average ``Rolling Resistance
                Coefficient'' (RRC) of the fleet. For the Draft TAR, the agencies did
                not have access to manufacturer confidential business information and
                relied on estimates from CONTROLTEC. As stated earlier, Figure VI-49
                shows the distribution of the estimated RRC for 1,358 vehicles models.
                The average RRC from the CONTROLTEC study (0.009) aligned with the NPRM
                estimate which was based in part on manufacturer submitted confidential
                business information. CONTROLTEC compared the estimated RRC data with
                the values provided by Rubber Manufacturers Association (renamed as
                USTMA-U.S. Tire Manufacturers Association) for original equipment
                tires. The average RRC from the data provided by RMA was 0.0092,\1470\
                compared to average of 0.009 from CONTROLTEC. CONTROLTEC attributed the
                difference due to analysis assumption, tire loading during coast down
                vs. load during tire testing, inflation pressure during coast down vs.
                inflation pressure during tire testing, coast down test reporting
                issues, tire types represented in the sample, tire break-in, and
                advancement in tire rolling resistance since the time RMA collected the
                data.
                ---------------------------------------------------------------------------
                 \1470\ Technical Analysis of Vehicle Load Reduction by
                CONTROLTEC for California Air Resources Board (April 29, 2015) at
                page 40.
                ---------------------------------------------------------------------------
                 CONTROLTEC also stated that RRC values for some vehicles fell below
                the average RRC (indicating better performance) due to estimation
                assumptions for vehicles where manufacturer data was not available, and
                coast down test reporting issues.\1471\ Further, CONTROLTEC performed a
                sensitivity study by mathematically removing aerodynamic contribution
                from the coast down coefficients. It was observed that the average RRC
                without the aerodynamic contribution is around 0.011. Accordingly, the
                agencies believe that it was reasonable to use 0.009 as the average RRC
                for the fleet for the NPRM and to continue to use that value for the
                final rule, based on the latest available data from manufacturers and
                alignment with the average RRC to the CONTROLTEC study estimate.
                ---------------------------------------------------------------------------
                 \1471\ Technical Analysis of Vehicle Load Reduction by
                CONTROLTEC for California Air Resources Board (April 29, 2015) at
                page 38.
                ---------------------------------------------------------------------------
                 H-D Systems (HDS) commented that the CONTROLTEC/CARB study showed
                that there is a very significant fraction of the fleet with tire
                rolling resistance coefficients above 10kg/1000 kg, and a small
                percentage of vehicles with rolling resistance coefficients already at
                0.05 or 0.06. HDS stated that NHTSA's baseline of 0.09 appeared ``a
                little low but may be appropriate if the distribution was sales
                weighted.'' HDS argued that a number of vehicle models already have
                tires below 0.07, and the likelihood that there would be additional
                tire improvements over the next decade are likely, meaning that ROLL30
                technology--or a 30 percent reduction of the tire rolling resistance
                coefficient to 0.063--is possible and appropriate for MY 2025.
                 Roush commented that rolling resistance is erroneously assumed to
                be the same across different vehicle classes, and that rolling
                resistance would vary depending upon the vehicle size, power,
                acceleration and performance package.\1472\
                ---------------------------------------------------------------------------
                 \1472\ NHTSA-2018-0067-11984.
                ---------------------------------------------------------------------------
                 As explained earlier, the RRC values used in the CONTROLTEC study
                were a combination of manufacturer information, estimates from coast
                down tests for some vehicles, and application of tire RRC values across
                other vehicles on the same platform. CONTROLTEC stated that some RRC
                values were below the estimated average (showing significant
                improvement from the baseline) due to assumptions that were
                [[Page 24567]]
                applied to some vehicles when manufacturer data was not available.
                Further, some of the RRC estimates were based on vehicle coast down
                tests which had errors.\1473\ As a result, some of the RRC values used
                in the Draft TAR showed significant improvements (30 percent reduction
                in rolling resistance relative to baseline), as observed by HDS. Based
                on a review of manufacturer-submitted confidential business information
                and other sources, the agencies are unaware of any tires in production
                which have 30 percent reduction in rolling resistance relative to
                baseline values.
                ---------------------------------------------------------------------------
                 \1473\ Technical Analysis of Vehicle Load Reduction by
                CONTROLTEC for California Air Resources Board (April 29, 2015) at
                page 38.
                ---------------------------------------------------------------------------
                 As stated earlier, the baseline values used for the Draft TAR
                analysis were ``Best in Class'' values from the estimates developed by
                CONTROLTEC and not representative of the average of the fleet or
                average for the vehicle classes. For the NPRM, the agencies revisited
                the ROLL technology assignments based on the RRC values provided by
                manufacturers, and the average RRC for each of the vehicle class was
                near the fleet average (RRC = 0.009). As shown in Figure VI-50, a vast
                majority of the vehicles in the fleet are in the ROLL0 bin across the
                different vehicle class, vehicle size, power, acceleration and
                performance configurations. For these reasons, the agencies will
                continue to use the fleet average of RRC = 0.009 as the baseline value
                to assess ROLL technology improvements.
                b) Rolling Resistance Analysis Fleet Assignments
                 As discussed above, NHTSA's Draft TAR analysis showed little
                rolling resistance technology in the baseline fleet for three reasons:
                the simulations used baseline values already reflecting best-in-class
                tire rolling resistance, credible tire rolling resistance values for
                all vehicles from bench data were not available to the agencies at the
                time of Draft TAR, and few manufacturers submitted rolling resistance
                values for the Draft TAR analysis.
                 For the NPRM, baseline (ROLL0) rolling resistance values were
                updated to 0.009, and any better rolling resistance values were
                assigned based on whether information indicated that vehicle had
                technology at least 10 percent better than baseline (.0081 or better
                for ROLL10), or at least 20 percent better than baseline (.0072 or
                better for ROLL20). The agencies used confidential business information
                provided by manufacturers to assign initial rolling resistance values
                for each vehicle make and model.
                 The Alliance commented that the NPRM MY 2016 analysis fleet had
                been updated with appropriate ratings of rolling resistance
                improvements, compared to the Draft TAR where vehicles were generally
                considered to have unimproved tires (meaning the Draft TAR assumed
                additional improvements were more achievable than in reality).\1474\
                The Alliance noted that the Draft TAR approach led to the CAFE model
                adding additional tire rolling resistance improvements even though
                manufacturers had already made significant improvements with that
                technology. This meant that the real-world fleet had little remaining
                opportunity for additional tire-related improvements, ultimately
                leading to the Draft TAR analysis underpredicting the amount of
                powertrain technology required for compliance.
                ---------------------------------------------------------------------------
                 \1474\ NHTSA-2018-006712039 at 136.
                ---------------------------------------------------------------------------
                 The Alliance noted that it is possible to estimate rolling
                resistance features of a vehicle using road load coefficients, but the
                process requires various assumptions and is not very accurate. The
                Alliance concluded that the agencies' use of CBI to assign baseline
                technology levels correctly was an accurate and practical solution.
                Similarly, Ford commented in support of the agencies' low rolling
                resistance tire assignments in the baseline fleet, stating that the
                accuracy of the baseline fleet assessment had been considerably
                improved using actual tire rolling resistance data.\1475\
                ---------------------------------------------------------------------------
                 \1475\ NHTSA-2018-0067-11928.
                ---------------------------------------------------------------------------
                 HDS commented that the analysis fleet ``accounts for the
                distribution of tires below 0.09 as 19% of vehicles in MY 2016 are
                modeled as having used ROLL10 and 25% of vehicles as having used ROLL20
                in the base year, but there is no accounting for the ~25% of vehicles
                having RRC values 10 to 20% above the 0.09 RRC average.'' \1476\ HDS
                concluded that ``[a] stricter accounting of the baseline and, possibly
                setting specific lower limits for 2025 RRC by vehicle type (as done for
                aero drag in the PRIA) will show significant additional fleetwide
                effectiveness from RRC reduction which is a very cost-effective
                technology.''
                ---------------------------------------------------------------------------
                 \1476\ NHTSA-2018-0067-11985 at 49.
                ---------------------------------------------------------------------------
                 ICCT commented that the agencies made a ``dramatic and
                unjustified'' shift in baseline tire rolling resistance assignments
                from the 2015 fleet used in the Draft TAR to the 2016 fleet used in the
                NPRM.\1477\ ICCT noted that per the agencies' updated baseline value,
                nearly 20 percent of all vehicles in the MY 2016 analysis fleet
                achieved 0.0081 (or better) rolling resistance value, and more than 26
                percent achieve 0.0072 (or better). ICCT argued that rather than
                changing the definition of rolling resistance technology to include
                improvements beyond the baseline, the agencies instead redefined the
                technology available, reducing the number of vehicles that can use tire
                improvements in future compliance years within the modeling framework,
                which artificially forced companies to use other, more expensive
                technologies.
                ---------------------------------------------------------------------------
                 \1477\ NHTSA-2018-0067-11741 full comments.
                ---------------------------------------------------------------------------
                 ICCT stated that to substantiate the baseline rolling resistance
                assignments, the agencies need to show data on how these improvements
                are evident in the fleet and delivering benefits. ICCT alleged that if
                an improvement of that magnitude were true, it would be evident in
                fleet level miles-per-gallon and CO2 levels; however, ``none
                of the quantifiable mpg or CO2 benefits that would be
                associated with these additional rolling resistance improvements were
                reflected with any real-world evidence in the model year 2016 fleet.''
                ICCT stated this seemed to be a case of the agencies ``artificially
                burying efficiency technology in the baseline, rendering it unusable in
                the post model year 2016 compliance scenarios.''
                 ICCT also stated that the agencies must share absolute road load
                coefficients for each vehicle make and model in the baseline fleet, and
                the technical justification for each value, in addition to conducting
                two sensitivity analysis cases ``assum[ing] that every baseline make
                and model is set to 0% rolling resistance improvement and set to the
                previous baseline rolling resistance (from the Draft TAR) to
                demonstrate how much the agencies' decision to load up more baseline
                technology affects the compliance scenarios, as it appears that the
                agencies may have made a unsupportable and non-rigorous assumption
                about rolling resistance technology across the models.'' ICCT concluded
                that because the changes were buried in the datafiles and unexplained,
                the agencies must issue a new regulatory analysis and allow an
                additional comment period for review of the methods and analysis.
                 Based on the comments from HDS and ICCT, the agencies reexamined
                available tire rolling resistance data. The assignment of ROLL20
                technology was revised for some vehicle models based on information on
                the use of common tires across vehicles that shared a platform. As a
                consequence, for the final rule, only 20 percent of the MY2017 vehicle
                fleet is assigned ROLL20. The
                [[Page 24568]]
                agencies will continue to investigate additional methods to improve the
                accuracy of this method, however as the Alliance and Ford noted, the
                accuracy of the baseline levels had been significantly improved over
                prior analyses by using actual tire RRC data. The agencies approach is
                consistent with the NAS recommendation to have two ROLL technology
                levels. The agencies determined that 30 percent rolling resistance
                improvement while maintaining other tire characteristics is unlikely to
                be available in the rulemaking timeframe.
                 The agencies considered a sensitivity case that assumed no mass
                reduction, rolling resistance, or aerodynamic improvements had been
                made to the MY 2017 fleet (i.e., setting all vehicle road levels to
                zero--MRO, AERO and ROLL0), in response to ICCT's comment. While this
                is an unrealistic characterization of the initial fleet, the agencies
                conducted a sensitivity analysis to understand any affect it may have
                on technology penetration along other paths (e.g. engine and hybrid
                technology). Under the CAFE program, the sensitivity analysis shows a
                slight decrease in reliance on engine technologies (HCR engines,
                turbocharge engines, and engines utilizing cylinder deactivation) and
                hybridization (strong hybrids and plug-in hybrids) in the baseline
                (relative to the central analysis). The consequence of this shift to
                reliance on lower-level road load technologies is a reduction in
                compliance cost in the baseline of about $300 per vehicle (in MY 2026).
                As a result, cost savings in the preferred alternative are reduced by
                about $200 per vehicle. Under the CO2 program, the general
                trend in technology shift is less dramatic (though the change in BEVs
                is larger) than the CAFE results. The cost change is also comparable,
                but slightly smaller ($200 per vehicle in the baseline) than the CAFE
                program results. Cost savings under the preferred alternative are
                further reduced by about $100. With the lower technology costs in all
                cases, the consumer payback periods decreased as well. These results
                are consistent with the approach taken by manufacturers who have
                already deployed many of the low-level road load reduction
                opportunities to improve fuel economy.
                 Figure VI-50 shows the distribution of ROLL technology for the
                Draft TAR, NPRM and final rule. For the NPRM, 64 percent of the MY 2016
                vehicle fleet was assigned ROLL0 and for the final rule, 59 percent of
                the MY2017 vehicle fleet is assigned ROLL0. This shows that the
                majority of the fleet is still at the ROLL0 technology level and there
                is still significant opportunity for the vehicle fleet to improve ROLL
                technology.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.280
                BILLING CODE 4910-59-C
                c) Rolling Resistance Adoption Features
                 In some cases, low rolling resistance tires can affect traction,
                which may adversely impact acceleration, braking and handling
                characteristics for some high-performance vehicles. Similar to past
                rulemakings, the agencies recognized in the NPRM that to maintain
                performance, braking and handling functionality, some high-performance
                vehicles would not adopt low rolling resistance tire technology. For
                cars and SUVs with more than 405 horsepower (hp), the agencies
                restricted the application of ROLL20. For cars and SUVs with more than
                500 hp, the agencies restricted the application of any additional
                rolling resistance technology (ROLL10 or ROLL20). The agencies
                developed these cutoffs based on a review of confidential business
                information and the distribution of rolling resistance values in the
                fleet.
                 Ford commented that the NPRM analysis appropriately limited the
                application of ROLL technology where it would be infeasible or would be
                at odds with the vehicles' intended function, characterizing that the
                decision to restrict application of ROLL10 and ROLL20 for high
                performance vehicles as reasonable.\1478\
                ---------------------------------------------------------------------------
                 \1478\ NHTSA-2018-0067-11928.
                ---------------------------------------------------------------------------
                 Accordingly, the agencies continued with the NPRM methodology of
                restricting certain ROLL technology for high performance vehicles. In
                the final rule, the agencies restricted the ROLL technology to ROLL0
                and ROLL10 for vehicles with greater than 405 hp and below 505hp. For
                vehicles greater than 505hp, the agencies restricted the ROLL
                technology to ROLL0.
                d) Rolling Resistance Effectiveness Modeling and Resulting
                Effectiveness Values
                 As discussed above, the agencies updated the baseline rolling
                resistance value to 0.009, based on a thorough review of confidential
                business information submitted by industry, and a review of other
                literature. To achieve ROLL10 in the NPRM and for the final rule
                analysis, the tire rolling resistance must be at least 10 percent
                better than baseline (.0081 or better). To achieve ROLL20, the tire
                rolling resistance must be at least 20 percent better than baseline
                (.0072 or better).
                 HDS commented that the Autonomie modeling assumed no engine change
                when drag and rolling resistance reductions were implemented, as well
                as no change to the transmission gear ratios and axle ratios, which
                vary by transmission type but not by the tractive
                [[Page 24569]]
                load.\1479\ HDS stated that ``reduction in rolling resistance is
                accompanied by axle ratio adjustments so that the engine operates at
                about the same load but at lower RPM. The EPA ALPHA model adjusts for
                this effect, which accounts for the difference in benefit estimates''
                between Autonomie and the ALPHA model simulations.
                ---------------------------------------------------------------------------
                 \1479\ NHTSA-2018-0067-11985.
                ---------------------------------------------------------------------------
                 As stated in Section VI.B.3 Tech Effectiveness and Modeling,
                Autonomie builds performance-neutral vehicle models by resizing
                engines, electric machines, and hybrid electric vehicle battery packs
                only at specific incremental technology steps. To address product
                complexity and economies of scale, engine resizing is limited to
                specific incremental technology changes that would typically be
                associated with a major vehicle or engine redesign.\1480\ Manufacturers
                have repeatedly told the agencies that the high costs for redesign and
                the increased manufacturing complexity that would result from resizing
                engines for small technology changes preclude them from doing so. It
                would be unreasonable and unaffordable to resize powertrains for every
                unique combination of technologies, and exceedingly so for every unique
                combination technologies across every vehicle model due to the extreme
                manufacturing complexity that would be required to do so. The agencies
                explained in the NPRM that the analysis should not include engine
                resizing with the application of every technology or for combinations
                of technologies that drive small performance changes to reflect better
                what is feasible for manufacturers.\1481\
                ---------------------------------------------------------------------------
                 \1480\ See 83 FR 43027 (Aug. 24, 2018).
                 \1481\ For instance, a vehicle would not get a modestly bigger
                engine if the vehicle comes with floor mats, nor would the vehicle
                get a modestly smaller engine without floor mats. This example
                demonstrates small levels of mass reduction. If manufacturers
                resized engines for small changes, manufacturers would have
                dramatically more part complexity, losing economies of scale.
                ---------------------------------------------------------------------------
                 Compliance modeling in the CAFE model also accounts for the
                industry practice of platform, engine, and transmission sharing to
                manage component complexity and associated costs.\1482\ At a vehicle
                refresh cycle, a vehicle may inherit an already resized powertrain from
                another vehicle within the same engine-sharing platform that adopted
                the powertrain in an earlier model year. In the Autonomie modeling,
                when a new vehicle adopts fuel saving technologies (such as ROLL
                technology) that are inherited, the engine is not resized (the
                properties from the baseline reference vehicle are used directly and
                unchanged) and there may be a small change in vehicle performance.
                ---------------------------------------------------------------------------
                 \1482\ Ford EcoBoost Engines are shared across ten different
                models in MY 2019. https://www.ford.com/powertrains/ecoboost/. Last
                accessed Nov. 05, 2019.
                ---------------------------------------------------------------------------
                 Regarding customizing transmission gear ratios as rolling
                resistance changes are implemented, the agencies explained in Section
                VI.C.2 Transmissions that it is an observable practice in industry to
                use a common gear set across multiple platforms and applications. The
                most recent example is the GM 10L90, a 10-speed automatic transmission
                that used the same gear set in both pick-up truck and passenger car
                applications.\1483\ In Autonomie, optimization of transmission
                performance is achieved through shift control logic rather than
                customized hardware (e.g., gear ratios) for each vehicle line. The
                shift initializer routine was run for every unique Autonomie full
                vehicle model configuration to generate customized shifting maps. The
                algorithms' optimization was designed to balance minimization of energy
                consumption against vehicle performance.\1484\ This balance was
                necessary to achieve the best fuel efficiency while maintaining
                customer acceptability by meeting performance neutrality requirements.
                See Section VI.B.3.a)(6) Performance Neutrality for more details. If
                the systems were over-optimized for the agencies' modeling, such as
                applying a unique gear set for each individual vehicle configuration,
                the analysis would likely over-predict the reasonably achievable fuel
                economy improvement for the technology. Over-prediction would be
                exaggerated when applied under real-world large-scale manufacturing
                constraints necessary to achieve the estimated costs for the
                transmission technologies.
                ---------------------------------------------------------------------------
                 \1483\ ``GM Global Propulsion Systems--USA Information Guide
                Model Year 2018'' (PDF). General Motors Powertrain. Retrieved
                September 26, 2019. https://www.gmpowertrain.com/assets/docs/2018R_F3F_Information_Guide_031918.pdf.
                 \1484\ See ANL model documentation for final rule.
                ---------------------------------------------------------------------------
                 As HDS noted, the EPA Draft TAR and Proposed Determination analyses
                performed using the ALPHA model adjusted the effectiveness of every
                technology combination assuming performance could be held constant for
                every combination, and did not recognize or account for the extreme
                complexity nor the associated costs for that impractical assumption.
                The NPRM and final rule analyses account for real-world practicalities
                and constraints related to both engine adoption and transmission
                adoption when other vehicle technologies are implemented, which
                explains some of the effectiveness and cost differences between the
                Draft TAR/Proposed Determination and the NPRM/final rule.
                 Figure VI-51 below shows the range of effectiveness used for the
                NPRM analysis for ROLL technologies.
                BILLING CODE 4910-59-P
                [[Page 24570]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.281
                BILLING CODE 4910-59-C
                 Figure VI-52 below shows the range of effectiveness values used for
                the final rule analysis.
                BILLING CODE 4910-59-P
                [[Page 24571]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.282
                BILLING CODE 4910-59-C
                e) Rolling Resistance Cost
                 For the NPRM, the analysis used DMC for ROLL technology from the
                Draft TAR and updated the values to reflect 2016$ dollars. The agencies
                continued to use the same cost assumptions presented in the NPRM for
                the final rule, and updated the values to 2018$ dollars. Table VI-139
                and Figure VI-53 show the different levels of tire rolling resistance
                technology cost.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.283
                [[Page 24572]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.284
                BILLING CODE 4910-59-C
                7. Other Vehicle Technologies
                 Four other vehicle technologies were included in the analysis--
                electric power steering (EPS), improved accessory devices (IACC), low
                drag brakes (LDB), and secondary axle disconnect (SAX) (which may only
                be applied to vehicles with all-wheel-drive or four-wheel-drive). The
                effectiveness of these technologies was applied directly by the CAFE
                model, with unique effectiveness values for each technology and for
                each technology class. This methodology was used in these four cases
                because the effectiveness of these technologies varies little with
                combinations of other technologies. Also, applying these technologies
                directly in the CAFE model significantly reduces the number of
                Autonomie simulations that are needed.
                a) Electric Power Steering (EPS)
                 Electric power steering reduces fuel consumption and CO2
                emissions by reducing load on the engine. Specifically, it reduces or
                eliminates the parasitic losses associated with engine-driven power
                steering pumps, which pump hydraulic fluid continuously through the
                steering actuation system even when no steering input is present. By
                selectively powering the electric assist only when steering input is
                applied, the power consumption of the system is reduced in comparison
                to the traditional ``always-on'' hydraulic steering system. Power
                steering may be electrified on light duty vehicles with standard 12V
                electrical systems and is also an enabler for vehicle electrification
                because it provides power steering when the engine is off (or when no
                combustion engine is present).
                 Power steering systems can be electrified in two ways.
                Manufacturers may choose to eliminate the hydraulic portion of the
                steering system and provide electric-only power steering (EPS) driven
                by an independent electric motor, or they may choose to move the
                hydraulic pump from a belt-driven configuration to a stand-alone
                electrically driven hydraulic pump. The latter system is commonly
                referred to as electro-hydraulic power steering (EHPS). As discussed in
                the NPRM, manufacturers have informed the agencies that full EPS
                systems are being developed for all types of light-duty vehicles,
                including large trucks.
                 EPS is also discussed in Section VI.C.3.a) Electrification Modeling
                in the CAFE model.
                b) Improved Accessories (IACC)
                 Engine accessories typically include the alternator, coolant pump,
                cooling fan, and oil pump, and are traditionally mechanically-driven
                via belts, gears, or directly by other rotating engine components such
                as camshafts or the crankshaft. These can be replaced with improved
                accessories (IACC) which may include high efficiency alternators,
                electrically driven (i.e., on-demand) coolant pumps, electric cooling
                fans, variable geometry oil pumps, and a mild regeneration
                strategy.\1485\ Replacing lower-efficiency and/or mechanically-driven
                components with these improved accessories results in a reduction in
                fuel consumption, as the improved accessories can conserve energy by
                being turned on/off ``on demand'' in some cases, driven at partial load
                as needed, or by operating more efficiently.
                ---------------------------------------------------------------------------
                 \1485\ IACC in this analysis excludes other electrical
                accessories such as electric oil pumps and electrically driven air
                conditioner compressors.
                ---------------------------------------------------------------------------
                 For example, electric coolant pumps and electric powertrain cooling
                fans provide better control of engine cooling. Flow from an electric
                coolant pump can be varied, and the cooling fan can be shut off during
                engine warm-up or cold ambient temperature conditions,
                [[Page 24573]]
                reducing warm-up time, fuel enrichment requirements, and, ultimately
                reducing parasitic losses.
                 IACC is also discussed in Section VI.C.3.a) Electrification
                Modeling in the CAFE model.
                c) Low Drag Brakes (LDB)
                 Low or zero drag brakes reduce or eliminate brake drag force by
                separating the brake pad from the rotor, either by mechanical or
                electric methods. Conventional disc brake systems are designed such
                that the brake pad is in contact with the brake rotor at all times.
                This is true even when the brakes are not being applied, and although
                the contact pressure is light in this case, this still produces some
                drag force on the vehicle.
                 LDBs have historically employed a caliper and rotor system that
                allows the piston in the caliper to retract,\1486\ in turn pulling the
                brake pads away from the rotor. However, if pads are allowed to move
                too far away from the rotor, the first pedal application made by the
                vehicle operator can feel spongy and have excessive travel. This can
                lead to customer dissatisfaction regarding braking performance and
                pedal feel. For this reason, in conventional hydraulic-only brake
                systems, manufacturers are limited by how much they can allow pads to
                move away from the rotor.
                ---------------------------------------------------------------------------
                 \1486\ The brake caliper pistons are used to push the brake pad
                against the brake rotor, or disc.
                ---------------------------------------------------------------------------
                 Recent developments in braking systems have resulted in brakes with
                the potential for zero drag. In these systems, the pedal feel is
                separated from hydraulics by a pedal simulator. This system is similar
                to the brake systems designed for hybrid and electric vehicles, where
                some of the primary braking is done through the recuperation of kinetic
                energy in the drive system. However, the pedal feel and the
                deceleration the operator experiences is tuned to provide a braking
                experience equivalent to that of a conventional hydraulic brake system.
                These ``brake-by-wire'' systems have highly tuned pedal simulators that
                feel like typical hydraulic brakes and seamlessly transition to a
                conventional system as required by different braking conditions. The
                application of a pedal simulator and brake-by-wire system is new to
                non-electrified vehicle applications. By using this type of system,
                vehicle manufacturers can allow brake pads to move farther away from
                the rotor and still maintain the initial pedal feel and deceleration
                associated with a conventional brake system.
                 In addition to reducing brake drag, the zero drag brake system
                provides ancillary benefits. It allows for a faster brake application
                and greater deceleration than is normally applied by the average
                vehicle operator. It also allows manufacturers to tune the braking for
                different customer preferences within the same vehicle. This means
                manufacturers can provide a ``sport'' mode, which provides greater
                deceleration with less pedal displacement and a ``normal'' mode, which
                might be more appropriate for day-to-day driving.
                 The zero drag brake system also eliminates the need for a brake
                booster. This saves cost and weight in the system. Elimination of the
                conventional vacuum brake booster could also improve the effectiveness
                of stop-start systems. Typical stop-start systems need to restart the
                engine if the brake pedal is cycled because the action drains the
                vacuum stored in the booster. Because the zero drag brake system
                provides braking assistance electrically, there is no need to
                supplement lost vacuum during an engine off event.
                 Finally, many engine technologies being considered to improve
                efficiency also reduce pumping losses through reduced throttling, and
                in turn there is less engine vacuum available to power-assist a
                conventional brake system. The reduction in throttling could require a
                supplemental vacuum pump to provide vacuum for a conventional brake
                system. This is the situation in many diesel-powered vehicles. Diesel
                engines have no throttling and require a supplemental vacuum for
                conventional brake systems. A zero drag brake system both eliminates
                brake drag and avoids the need for a supplemental vacuum pump.
                d) Secondary Axle Disconnect (SAX)
                 All-wheel drive (AWD) and four-wheel drive (4WD) vehicles provide
                improved traction by delivering torque to the front and rear axles,
                rather than just one axle. When a second axle is rotating, it tends to
                consume more energy because of additional losses related to lubricant
                churning, seal friction, bearing friction, and gear train
                inefficiencies.\1487\ \1488\ Some of these losses may be reduced by
                providing a secondary axle disconnect function that disconnects one of
                the axles when driving conditions do not call for torque to be
                delivered to both.
                ---------------------------------------------------------------------------
                 \1487\ Phelps, P. ``EcoTrac Disconnecting AWD System,''
                presented at 7th International CTI Symposium North America 2013,
                Rochester MI.
                 \1488\ Pilot Systems, ``AWD Component Analysis,'' Project
                Report, performed for Transport Canada, Contract T8080-150132, May
                31, 2016.
                ---------------------------------------------------------------------------
                 The terms AWD and 4WD are often used interchangeably, although they
                have also developed a colloquial distinction, and are two separate
                systems. The term AWD has come to be associated with light-duty
                passenger vehicles providing variable operation of one or both axles on
                ordinary roads. The term 4WD is often associated with larger truck-
                based vehicle platforms providing a locked driveline configuration and/
                or a low range gearing meant primarily for off-road use.
                 Many 4WD vehicles provide for a single-axle (or two-wheel) drive
                mode that may be manually selected by the user. In this mode, a primary
                axle (usually the rear axle) will be powered, while the other axle
                (known as the secondary axle) is not. However, even though the
                secondary axle and associated driveline components are not receiving
                engine power, they are still connected to the non-driven wheels and
                will rotate when the vehicle is in motion. This unnecessary rotation
                consumes energy,\1489\ and leads to increased fuel consumption and CO2
                emissions that could be avoided if the secondary axle components were
                completely disconnected and not rotating.
                ---------------------------------------------------------------------------
                 \1489\ Any time a drivetrain component spins it consumes some
                energy, primarily to overcome frictional forces.
                ---------------------------------------------------------------------------
                 Light-duty AWD systems are often designed to divide variably torque
                between the front and rear axles in normal driving to optimize traction
                and handling in response to driving conditions. However, even when the
                secondary axle is not necessary for enhanced traction or handling, in
                traditional AWD systems it typically remains engaged with the driveline
                and continues to generate losses that could be avoided if the axle was
                instead disconnected. The SAX technology observed in the marketplace
                disengages one axle (typically the rear axle) for 2WD operation, but
                detects changes in driving conditions and automatically engages AWD
                mode when it is necessary. The operation in 2WD can result in reduced
                fuel consumption. For example, Chrysler has estimated the secondary
                axle disconnect feature in the Jeep Cherokee reduces friction and drag
                attributable to the secondary axle by 80% when in disconnect
                mode.\1490\
                ---------------------------------------------------------------------------
                 \1490\ Brooke, L. ``Systems Engineering a new 4x4 benchmark,''
                SAE Automotive Engineering, June 2, 2014.
                ---------------------------------------------------------------------------
                e) Analysis Fleet Assignments for Other Vehicle Technologies
                 The agencies described in the PRIA that the aforementioned
                technologies have been applied, to some extent, in the MY 2016 fleet.
                However, these technologies are difficult to observe and
                [[Page 24574]]
                assign to the analysis fleet, and the agencies relied heavily on
                industry engagement and feedback to assign the technologies properly to
                the NPRM analysis fleet vehicles. In the NPRM, the agencies noted that
                the Draft TAR analysis did not properly account for the presence of
                these technologies in the analysis fleet, and far too few were
                assigned. Accordingly, the NPRM analysis reflected higher EPS and IACC
                application rates than the Draft TAR analysis.
                 The agencies received a handful of comments stating that the
                additional technologies were incorrectly applied to the analysis fleet.
                ICCT stated that the inclusion of EPS, IACC, and LDB in the analysis
                fleet was unsubstantiated, and removed the technologies from potential
                use during the subsequent simulated years.\1491\ ACEEE commented that
                IACC should not have been applied to certain vehicles in the analysis
                fleet because those vehicles do not in actuality display the fuel
                consumption reduction that would confirm the presence of these
                additional technologies.\1492\ In addition, ACEEE commented that the
                CAFE model assumes significant baseline SAX penetration that they could
                not corroborate from Ford F-150 product information brochures.\1493\
                HDS compared the available levels of IACC improvements from the Draft
                TAR to the NPRM analysis, noting that the NPRM only employed one level
                of improved accessory technologies.\1494\ HDS stated that this implied
                the effectiveness of what was previously considered IACC1 (the first
                level of IACC technology improvement available in the Draft TAR) was
                completely used up in the 2016 analysis fleet for this rule.
                ---------------------------------------------------------------------------
                 \1491\ International Council on Clean Transportation, Attachment
                3, Docket No. NHTSA-2018-0067-11741, at I-37.
                 \1492\ American Council for an Energy-Efficient Economy,
                Attachment 6, Docket No. NHTSA-2018-0067-12122, at 6.
                 \1493\ American Council for an Energy-Efficient Economy,
                Attachment 6, Docket No. NHTSA-2018-0067-12122, at 7.
                 \1494\ H-D Systems, ``HDS final report,'' Docket No. NHTSA-2018-
                0067-11985, at 21.
                ---------------------------------------------------------------------------
                 As the agencies stated in the PRIA, in part because of the
                difficulty in observing EPS, IACC, LDB, and SAX on actual vehicles, far
                too few of those technologies were assigned to vehicles in the Draft
                TAR analysis fleets. For the final rule, each vehicle in the MY 2017
                analysis fleet was studied using confidential and publicly available
                information to determine whether, as commenters suggested, the agencies
                had improperly applied any of these additional vehicle technologies.
                This resulted in some adjustments in the application of the
                technologies in the analysis fleet. In regard to ACEEE's comment on SAX
                penetration in the analysis fleet, for the NPRM and final rule
                analysis, the agencies considered all 4WD vehicles to have the
                capability manually to disconnect either the front or rear wheel axle
                and associated rotating components, thus shifting to a 2WD mode. When
                4WD operation is required for safety and utility, the consumer can
                enable this feature. As stated above, this capacity to shift between
                2WD and 4WD modes is another form of SAX. For AWD vehicles, publicly
                available manufacturer information was reviewed to identify the
                specific vehicles that have SAX technology. Based on market
                observations and feedback from OEMs, the entire analysis fleet for NPRM
                and the final rule was considered to have a basic level of improved
                accessories (comparable to what Draft TAR referred to as IACC1). The
                application of IACC in the NPRM and final rule analysis fleets
                represents further improvements to accessories such as electric water
                pumps and higher efficiency alternators with mild regeneration
                capacity.
                 The following distribution of technologies in the analysis fleet
                from the NPRM to the final rule analysis shows a slight decrease in the
                portion of total vehicles produced that have EPS and IACC, a very
                slight increase in the portion of total vehicle production that have
                LDB, and a slight increase in the portion of 4WD/AWD vehicles with SAX
                technology.
                BILLING CODE 4910-59-P
                [[Page 24575]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.285
                BILLING CODE 4910-59-C
                f) Effectiveness Estimates for Other Vehicle Technologies
                 The effectiveness estimates for these four technologies rely on
                previous work published as part of the rulemaking process, both for the
                2012 rule for MYs 2017-2025 and the Draft TAR. The effectiveness values
                are unchanged from the Draft TAR.
                 The effectiveness of both EPS and EHPS is derived from the
                decoupling of the pump from the crankshaft, and is considered to be
                practically the same for both. Thus, a single effectiveness value is
                assigned to all vehicles in the analysis fleet that possess either EPS
                or EHPS, and the ``EPS'' designation is applied.
                 For the Draft TAR analysis, two levels of IACC were offered as a
                technology path (a low improvement level and a high improvement level).
                Since much of the market has incorporated some of these technologies in
                the baseline MY 2016 and 2017 fleets, the NPRM and final rule analyses
                assumed all vehicles have incorporated what was previously the low
                level, so only the high level remained as an option for vehicles. The
                figure above shows the distribution of IACC for NPRM and FRM, which is
                the equivalent type of technology as the high-level IACC in the DRAFT
                TAR.
                 The NPRM analysis carried forward work on the effectiveness of SAX
                systems conducted in the Draft TAR and EPA Proposed Determination. This
                work involved gathering information by monitoring press reports,
                holding meetings with suppliers and OEMs, and attending industry
                technical conferences. The resulting effectiveness estimates used in
                the Draft TAR, NPRM, and this final rule are shown below.
                BILLING CODE 4910-59-P
                [[Page 24576]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.286
                BILLING CODE 4910-59-C
                g) Cost Estimates and Learning Rates for Other Vehicle Technologies
                 The cost estimates for these technologies rely on previous work
                published as part of the rulemaking process, both for the 2012 rule for
                MYs 2017-2027 and the Draft TAR. The cost values are from the same
                sources as the Draft TAR and were updated to 2016 dollars for the NPRM
                and 2018 dollars for the final rule analysis. Learning rates for these
                technologies are also unchanged since the NPRM, and can be seen in
                Section VI.B.4.d)(4) Cost Learning as Applied in the CAFE Model.
                 CARB noted that the IACC costs in Tables 6-32 and 6-33 of the PRIA
                did not align with the Technologies central analysis input file.\1495\
                HDS commented, as part of its comparison of IACC penetration in the
                analysis fleet from the Draft TAR to NPRM, that IACC costs were based
                on the difference between IACC1 and IACC2 costs and this appeared to be
                inconsistent with the cost of accessory electrification which is more
                expensive.\1496\
                ---------------------------------------------------------------------------
                 \1495\ CARB, Docket No. NHTSA-2018-0067-12428, at 21.
                 \1496\ H-D Systems, ``HDS final report,'' Docket No. NHTSA-2018-
                0067-11985, at 21.
                ---------------------------------------------------------------------------
                 In the PRIA, the cost of IACC was reported in some tables as an
                absolute cost (the cost of adding IACC to a base vehicle), while the
                NPRM Technologies central analysis input file showed IACC cost
                incremental to EPS. This was necessary in the model input file because
                the accounting method of the NPRM CAFE model utilized incremental
                costs. In contrast, a change in the CAFE model accounting method for
                this final rule allows all costs in the input file to be reported as
                absolute costs, incremental to a base vehicle. It was assumed that EPS
                must be present on a vehicle in order for it to adopt IACC, and as such
                the cost of IACC includes the cost of EPS. For further detail on the
                use of absolute costs in place of incremental costs, see Section
                VI.C.7.g). Although HDS commented that accessory electrification has a
                higher cost than what is being used in the analysis, no specific
                additional input was given; the cost of IACC, as was done for Draft TAR
                (where it was referred to as IACC2), was taken from the 2015 NAS
                Report.\1497\
                ---------------------------------------------------------------------------
                 \1497\ National Research Council. 2015. Cost, Effectiveness, and
                Deployment of Fuel Economy Technologies for Light-Duty Vehicles.
                Washington, DC--The National Academies Press, Table 8A.2a, available
                at https://www.nap.edu/catalog/21744/cost-effectiveness-and-deployment-of-fuel-economy-technologies-for-light-duty-vehicles.
                ---------------------------------------------------------------------------
                 Table VI-141 below shows the absolute costs for these technologies
                for select model years. The FRM Technologies central analysis input
                file shows the costs for all model years.
                [[Page 24577]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.287
                8. Simulating Off-Cycle and A/C Efficiency Technology Adjustments
                 Off-cycle and air conditioning (A/C) efficiency technologies can
                provide fuel economy improvements in real-world vehicle operation, but
                that benefit cannot be adequately captured by the 2-cycle test
                procedures used to demonstrate compliance with fuel economy and
                CO2 emissions standards.\1498\ Off-cycle technologies
                include technologies like high efficiency alternators and high
                efficiency exterior lighting.\1499\ A/C efficiency technologies operate
                mainly by reducing the operation of the compressor, which pumps A/C
                refrigerant around the system loop. The less the compressor operates or
                the more efficiently it operates, the less load the compressor places
                on the engine, resulting in better fuel efficiency and lower
                CO2 emissions.
                ---------------------------------------------------------------------------
                 \1498\ See 49 U.S.C 32904(c) (``The Administrator shall measure
                fuel economy for each model and calculate average fuel economy for a
                manufacturer under testing and calculation procedures prescribed by
                the Administrator. . . . the Administrator shall use the same
                procedures for passenger automobiles the Administrator used for
                model year 1975 (weighted 55 percent urban cycle and 45 percent
                highway cycle), or procedures that give comparable results.'').
                 \1499\ See 83 FR 43057. A partial list of off-cycle technologies
                is included in Tables II-21 and II-22 of the NPRM.
                ---------------------------------------------------------------------------
                 Vehicle manufacturers have the option to generate credits for off-
                cycle technologies and improved A/C systems under the EPA's
                CO2 program and receive a fuel consumption improvement value
                (FCIV) equal to the value of the benefit not captured on the 2-cycle
                test under NHTSA's CAFE program. The FCIV is not a credit in the NHTSA
                CAFE program, but the FCIVs increase the reported fuel economy of a
                manufacturer's fleet, which is used to determine compliance. EPA
                applies FCIVs during determination of a fleet's final average fuel
                economy reported to NHTSA.\1500\ FCIVs are only calculated and applied
                at a fleet level for a manufacturer and are based on the volume of the
                manufacturer's fleet that contain qualifying technologies.\1501\
                ---------------------------------------------------------------------------
                 \1500\ 49 U.S.C. 32904(c)-(e). EPCA granted EPA authority to
                establish fuel economy testing and calculation procedures. See
                Section IX for more information.
                 \1501\ 40 CFR 600.510-12(c)
                ---------------------------------------------------------------------------
                 As discussed further in Section IX.D Compliance Issues that Affect
                Both the CO2 and CAFE Programs, three pathways can be used
                to determine the value of A/C efficiency and off-cycle adjustments.
                First, manufacturers can use a predetermined list or ``menu'' of credit
                values established by EPA for specific off-cycle technologies.\1502\
                Second, manufacturers can use 5-cycle testing to demonstrate and
                justify off-cycle CO2 credits; \1503\ the additional tests
                allow emission benefits to be demonstrated over some elements of real-
                world driving not captured by the 2-cycle compliance tests, including
                high speeds, rapid accelerations, and cold temperatures. Third,
                manufacturers can seek EPA approval, through a notice and comment
                process, to use an alternative methodology other than the menu or 5-
                cycle methodology for determining the off-cycle technology improvement
                values.\1504\
                ---------------------------------------------------------------------------
                 \1502\ See 40 CFR 86.1869-12(b). The Technical Support Document
                (TSD) for the 2012 final rule for MYs 2017 and beyond provides
                technology examples and guidance with respect to the potential
                pathways to achieve the desired physical impact of a specific off-
                cycle technology from the menu and provides the foundation for the
                analysis justifying the credits provided by the menu. The
                expectation is that manufacturers will use the information in the
                TSD to design and implement off-cycle technologies that meet or
                exceed those expectations in order to achieve the real-world
                benefits of off-cycle technologies from the menu.
                 \1503\ See 40 CFR 86.1869-12(c). EPA proposed a correction for
                the 5-cycle pathway in a separate technical amendments rulemaking.
                See 83 FR 49344 (Oct. 1, 2019). EPA is not approving credits based
                on the 5-cycle pathway pending the finalization of the technical
                amendments rule.
                 \1504\ See 40 CFR 86.1869-12(d).
                ---------------------------------------------------------------------------
                 The agencies have been collecting data on the application of these
                technologies since implementing the programs.\1505\ Most manufacturers
                are generating A/C efficiency and off-cycle credits; in MY 2017, 15
                manufacturers generated A/C efficiency credits and 15 manufacturers
                generated off-cycle credits, through the level of deployment varies by
                manufacturer.\1506\
                ---------------------------------------------------------------------------
                 \1505\ See 77 FR at 62832, 62839 (Oct. 15, 2012). EPA introduced
                A/C and off-cycle technology credits for the CO2 program
                in the MYs 2012-2016 rule and revised the program in the MY 2017-
                2025 rule and NHTSA adopted equivalent provisions for MYs 2017 and
                later in the MY 2017-2025 rule.
                 \1506\ The 2018 EPA Automotive Trends Report, EPA-420-R-19-002,
                March 2019 at Chapter 5.B., Figures 5.10 and 5.11.
                ---------------------------------------------------------------------------
                a) A/C and Off-Cycle Effectiveness Modeling
                 The NPRM analysis used the off-cycle FCIVs and credits earned by
                each manufacturer in MY 2016 and carried these forward at the same
                levels for future years for the CO2 analysis and beginning
                in MY 2017 for the CAFE analysis. The 2016 values for off-cycle FCIVs
                for each manufacturer and fleet, denominated in grams CO2
                per mile,\1507\ are provided in Table VI-142.\1508\ Additional off-
                cycle FCIVs were added in future years if a manufacturer applied a
                technology that was explicitly simulated in the analysis and also was
                an off-cycle technology listed on the predefined menu.\1509\
                Technologies explicitly simulated in the analysis that are also on the
                off-cycle menu include start-stop systems that reduce fuel consumption
                during idle and active grille shutters that improve aerodynamic drag at
                highway speeds,
                [[Page 24578]]
                among others. Any off-cycle adjustments that accrued as the result of
                applying these technologies were calculated dynamically in each model
                year the technology was applied, with adjustments accumulating up to
                the 10 g/mi cap. As a practical matter, most of the adjustments for
                which manufacturers can claim off-cycle FCIVs exist outside of the CAFE
                model technology tree so the off-cycle menu cap was rarely reached for
                the NPRM analysis.
                ---------------------------------------------------------------------------
                 \1507\ For the purpose of estimating their contribution to CAFE
                compliance, the grams CO2/mile values in Table I-1 are
                converted to gallons/mile and applied to a manufacturer's 2-cycle
                CAFE performance. When calculating compliance with EPA's
                CO2 program, there is no conversion necessary (as
                standards are also denominated in grams/mile).
                 \1508\ 2016 GHG Manufacturer Performance Report. EPA-420-R-18-
                002. January 2018. https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100TGIA.pdf. Last Accessed Nov. 14, 2019. 2016
                Report Tables for the GHG Manufacturer Performance Report. January
                2018. https://www.epa.gov/sites/production/files/2018-01/ghg-report-2016-data-tables.xlsx. Last Accessed Nov. 14, 2019.
                 \1509\ For more details, see Section IX.D Compliance Issues that
                Affect Both the CO2 and CAFE Programs and Section IX.D.3
                Flexibilities for Off-Cycle Technologies.
                ---------------------------------------------------------------------------
                 The agencies sought comment on both the A/C and off-cycle data that
                was used for the NPRM analysis as well as the assumptions for applying
                those technologies.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.288
                BILLING CODE 4910-59-C
                 Universally, stakeholders believed the application of off-cycle
                adjustments in the analysis was too conservative. Stakeholders believed
                the A/C and off-cycle technologies would be rapidly deployed and
                manufacturers would reach the cap values within the rulemaking
                timeframe.
                ---------------------------------------------------------------------------
                 \1510\ See 83 FR 43159-60 (``. . . this analysis uses the off-
                cycle credits submitted by each manufacturer for MY 2017 compliance
                and carries these forward to future years with a few exceptions.'').
                ---------------------------------------------------------------------------
                 The Institute for Policy Integrity (IPI) questioned the position
                the agencies assumed in the NPRM analysis, and suggested the agencies
                ``assume that manufacturers will efficiently deploy all cost-saving
                offset opportunities, especially in the face of increasingly stringent
                standards.'' \1511\
                ---------------------------------------------------------------------------
                 \1511\ Comments from Institute from Policy Integrity, Attachment
                1, NPRM Docket No. NHTSA-2018-0067-12213, at 20-21.
                ---------------------------------------------------------------------------
                 ICCT stated ``far greater use of the off-cycle provisions will
                occur by 2025'' and emphasized that off-cycle technologies are ``highly
                cost-effective and being deployed in greater sales penetrations than
                many of the test-cycle efficiency technologies that the agencies are
                analyzing.'' \1512\ ICCT supported manufacturers maximizing the use of
                off-cycle technologies, and supported the analysis estimating
                ``fleetwide off-cycle credit use at over 10 g/mile by 2020,'' and
                further suggested fleetwide achievement of 15 g/mile by 2025.\1513\
                ---------------------------------------------------------------------------
                 \1512\ Comments from ICCT, Attachment 1, NPRM Docket No. NHTSA-
                2018-0067-11741, at I40--I41.
                 \1513\ Note there is a regulatory ``cap'' on menu technologies
                of 10 g/mi (see Section IX for further discussion of the cap),
                however a manufacturer can receive additional off-cycle credit/FCIV
                by using the pathways described above to petition for off-menu
                technologies. ICCT's comment suggests that manufacturers will reach
                the regulatory menu cap and apply additional technologies to get an
                additional 5 g/mi credit above the menu cap.
                ---------------------------------------------------------------------------
                 FCA, General Motors and the Auto Alliance all provided similar
                observations, stating ``[m]anufacturers have rapidly deployed
                technology in response to this all new regulatory
                [[Page 24579]]
                mechanism.'' Each of the commenters provided support for an argument of
                rapid off-cycle technology adoption, stating ``[i]n the MY2021-2026
                timeframe of the proposed rule, it is likely that manufacturers will
                hit the existing 10 g/mi cap.'' \1514\
                ---------------------------------------------------------------------------
                 \1514\ Comments from Automotive Alliance, Appendix 1, NPRM
                Docket No. NHTSA-2018-0067-12073, at 92; Comments from Fiat Chrysler
                Automobiles, Attachment1, NPRM Docket No. NHTSA-2018-0067-11943, at
                8; Comments from General Motors, Appendix 4--Comments to Technical
                Issues, NPRM Docket No. NHTSA-2018-0067-11858, at 1.
                ---------------------------------------------------------------------------
                 The DENSO Corporation further supported the increased use of off-
                cycle technologies, commenting that ``[a]vailable data on OEM off-cycle
                technology credit utilization within the past few years demonstrates
                that the use of off-cycle technologies is expected to grow--
                particularly technologies on the credit menus.'' \1515\
                ---------------------------------------------------------------------------
                 \1515\ Comments from DENSO Corporation, Attachment 1, NPRM
                Docket No. NHTSA-2018-0067-11880, at 6.
                ---------------------------------------------------------------------------
                 However, Toyota Motors North America asked for constraints on
                considerations of off-cycle technology in the analysis.\1516\ Toyota
                expressed concern for over-reliance on off-cycle technologies to
                provide flexibilities for compliance, as ``most of the technologies
                provide little tangible value proposition for customers.'' In
                additional comments, Toyota repeated the concern noting, ``most of
                these technologies lack consumer demand.'' Finally, Toyota specifically
                cautioned against overusing off-cycle technologies in the analysis,
                stating ``[t]he suggested pursuit of maximum credits overlooks the
                associated costs and market acceptance challenge for certain off-cycle
                technologies.'' Toyota listed costs versus risk of customer acceptance
                and agency approval as factors that ``introduce a high level of
                uncertainty for an auto manufacturer's planning and make investments in
                off-cycle technologies risky and less appealing.''
                ---------------------------------------------------------------------------
                 \1516\ Comments from Toyota Motors North America, Attachment 1,
                NHTSA Docket No. NHTSA-2018-0067-130798, at 9-10; Supplemental
                Comments from Toyota Motors North America, Attachment 1, NHTSA
                Docket No. NHTSA-2018-0067-12150, at 24; Supplemental Comments from
                Toyota Motors North America, Attachment 1, NHTSA Docket No. NHTSA-
                2018-0067-12376, at 4-5.
                ---------------------------------------------------------------------------
                 After carefully considering the comments, the agencies agree that
                A/C and off-cycle technologies are likely to be more broadly applied by
                manufacturers within the rulemaking timeframe. The final rule analysis
                has been updated to reflect an increased application of the
                technologies. Similar to the NPRM, the final rule analysis used the A/C
                and off-cycle FCIVs earned by each manufacturer in the baseline fleet
                (MY 2017 for the final rule analysis) as a starting point. However, the
                final rule analysis increased these values in subsequent model years.
                In addition to the dynamic application of off-cycle FCIVs, as in the
                NPRM, each manufacturer's fleet FCIVs were increased by extrapolating
                the manufacturers' historical rate of FCIV application through
                2017.\1517\ In line with most commenters, the agencies increased the
                FCIVs for each manufacturer such that the maximum value of 10 g/mi will
                be reached by MY 2023. For manufacturers who did not reach maximum
                values prior to 2023 through data extrapolation, a linear increase to
                the cap was assumed. The agencies believe this approach balances a
                greater application of FCIV technologies across the fleet, while
                avoiding uncertain over-reliance on flexibilities for the analysis.
                ---------------------------------------------------------------------------
                 \1517\ The 2018 EPA Automotive Trends Report, https://www.epa.gov/fuel-economy-trends/download-report-co2-and-fuel-economy-trends. Accessed Aug 23, 2019.
                ---------------------------------------------------------------------------
                 The agencies disagreed with the proposal to model the application
                of 15 g/mi of FCIVs universally in the rulemaking timeframe. Based on
                historical data and industry comments from both manufacturers and
                suppliers, the agencies expect there will be an increase in off-cycle
                technology application. However, there are two issues with assuming
                manufacturers will exceed the existing off-cycle caps. First, only a
                few manufacturers approached the cap limit in MY 2018, and the fleet
                average menu credit was 4.7 grams/mile, less than half the cap
                value.\1518\ Second, new off-cycle technologies may address the same
                inefficiencies as menu technologies, rather than work in conjunction.
                Accordingly, the agencies believe there is a reasonable basis for
                assuming manufacturers could, and would only achieve 10 g/mi on average
                by MY 2023, and used that assumption for the final rule analysis.
                ---------------------------------------------------------------------------
                 \1518\ The 2018 EPA Automotive Trends Report, Greenhouse Gas
                Emissions, Fuel Economy, and Technology since 1975, EPA-420-R-19-002
                (Mar. 2019).
                ---------------------------------------------------------------------------
                 Table VI-143 shows passenger car values for FCIVs and Table VI-144
                shows light truck values for FCIVs applied for the final rule analysis.
                BILLING CODE 4910-59-P
                [[Page 24580]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.289
                [[Page 24581]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.290
                [GRAPHIC] [TIFF OMITTED] TR30AP20.291
                [[Page 24582]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.292
                [[Page 24583]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.293
                BILLING CODE 4910-59-C
                [[Page 24584]]
                A/C Efficiency, A/C Leakage and Off-Cycle Costs
                 As discussed above, the only A/C efficiency and off-cycle
                technologies applied dynamically in the NPRM analysis were explicitly
                simulated technologies like stop-start systems and active aerodynamic
                technologies. The NPRM analysis fully accounted for both the
                effectiveness and cost of these technologies and therefore separate
                cost accounting was not needed. For example, when stop-start or active
                aerodynamics technology was added by the model to a vehicle, the
                corresponding off-cycle FCIVs were applied and the technology costs
                were captured the same as every other technology on the decision trees.
                 For the final rule analysis, A/C and off-cycle technologies are
                applied independently of the decision trees using the extrapolated
                values, so it is necessary to account for the costs of those
                technologies independently. Table VI-145 shows the costs used for A/C
                and off-cycle FCIVs the final rule analysis. The costs are shown in
                dollars per gram of CO2 per mile ($ per g/mile). The A/C
                costs and off-cycle technology costs are the same costs used in the EPA
                Proposed Determination and described in the EPA Proposed Determination
                TSD.\1519\
                ---------------------------------------------------------------------------
                 \1519\ EPA PD TSD. EPA-420-R-16-021. November 2016. At 2-423-2-
                245. https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=P100Q3L4.pdf. Last
                accessed Nov.14, 2019.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.294
                BILLING CODE 4910-59-C
                D. Impacts that Result From Simulating Manufacturer Compliance with
                Regulatory Alternatives
                1. Simulating Economic Impacts of Regulatory Alternatives
                a) What Economic Impacts Occur When Vehicle Manufacturers Comply With
                Different CAFE and CO2 Standards?
                1) The NPRM Framework for Analyzing Economic Impacts
                 In the proposed rule, the agencies noted the importance of
                identifying the mechanisms by which vehicle manufacturers' compliance
                with different CAFE and CO2 standards generated impacts on
                manufacturers, owners of new and used vehicles, and the remainder of
                the U.S. The agencies organized the analysis of alternative standards
                using a framework that clarified the economic impacts on vehicle
                producers, illustrated how costs were transmitted to buyers of new
                vehicles, highlighted the collateral economic effects on owners of used
                vehicles, and identified how these responses created various indirect
                costs and benefits. Throughout the analysis, the agencies stressed the
                distinction between the proposal's economic consequences for private
                businesses and households, and its ``external'' economic impacts--those
                ultimately borne by the rest of the U.S. economy.
                 To clarify the framework used in the proposal, the agencies used
                Table VI-146 below (which is based on Tables II-25 to II-28 from the
                NPRM) \1520\ to report costs and benefits and to trace how they pass
                through the economy. As the table shows, the economic impacts of
                standards initially fall on vehicle manufactures, but ultimately are
                borne by consumers who purchase and drive new models. Smaller, indirect
                economic effects of the proposal would be borne by owners of used cars
                and light trucks (vehicles produced during model years prior to those
                affected by the proposal, but still in use) as well as by the general
                public and government agencies. On balance, the agencies projected that
                most of the proposal's economic effects would fall on private
                businesses and households, with the remainder of the U.S. economy
                bearing much smaller impacts.
                ---------------------------------------------------------------------------
                 \1520\ See 83 FR at 43062-66.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24585]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.295
                BILLING CODE 4910-59-C
                 More specifically, the agencies' analysis showed that the proposal
                would initially have saved manufacturers the costs of adding the
                technologies that would otherwise have been necessary to enable their
                new cars and light trucks to comply with the baseline fuel economy and
                CO2 emissions regulations, with the estimated dollar value
                of those savings shown in line 1 of Table VI-146. The proposal also
                enabled some manufacturers to make lower civil penalty payments for
                failing to comply with the more demanding standards that were
                supplanted (line 2), although these savings would have been exactly
                offset by lower civil penalty revenue to the
                [[Page 24586]]
                Federal Government (line 16). The analysis assumed that manufacturers
                would have the ability, in a competitive market, to pass their savings
                in technology costs and any reduction in civil penalties paid on to
                buyers, by charging lower prices for new vehicles. Although lower
                prices reduced their revenues (line 3), on balance, their savings in
                compliance costs, reduced civil penalty payments, and lower sales
                revenue were assumed to leave manufacturers financially unaffected
                (shown by the zero entry in line 4 of the table).
                 Under the proposal, the analysis showed that buyers of new cars and
                light trucks benefited directly from those vehicles' lower purchase
                prices and financing costs (line 5). They also avoided the increased
                risk of crash-related injuries that would have resulted from reductions
                in the weight of some new models, as manufacturers attempted to improve
                fuel economy to comply with the baseline standards. The economic value
                of this reduction in risk represented an additional benefit from the
                proposal to reducing the stringency of the standards vis-[agrave]-vis
                the baseline (line 6).
                 At the same time, however, the lower fuel economy that some new
                cars and light trucks were expected to offer with less stringent
                standards in place would have imposed various additional costs on their
                buyers and users. Drivers experienced higher fuel costs as a
                consequence of new vehicles' increased fuel consumption (line 7), as
                well as the added time and inconvenience of having to make more
                frequent refueling stops required by reduced driving range (line 8).
                They also forfeited some mobility benefits as they drove newly-
                purchased cars and light trucks less in response to their higher fuel
                costs (line 9). On balance, the agencies' analysis of the proposal
                showed that buyers of new cars and light trucks produced during the
                model years it affected would experience significant economic benefits
                (line 10).
                 A novel feature of the agencies' evaluation of the proposal showed
                that lowering prices for new cars and light trucks, some owners of used
                vehicles retired them from service earlier than they otherwise would
                have done. In combination with increased sales of new models, this
                transferred some driving that would have occurred with used cars and
                light trucks to newer and safer models, thus reducing the total costs
                of fatalities and injuries sustained in motor vehicle crashes.\1521\ In
                the proposal, this reduction in injury risks provided benefits to
                owners and drivers of older cars and light trucks that had not been
                recognized or quantified in its analyses of previous CAFE and
                CO2 standards (line 11).
                ---------------------------------------------------------------------------
                 \1521\ This improvement in safety resulted from the fact that
                cars and light trucks have become progressively more protective in
                crashes over time (and also slightly less prone to certain types of
                crashes, such as rollovers). Thus, shifting some travel from older
                to newer models reduced injuries and damages sustained by drivers
                and passengers because they were traveling in inherently safer
                vehicles, rather than because of changes to driver risk profiles.
                ---------------------------------------------------------------------------
                 Table VI-146 also showed that the changes in fuel consumption and
                vehicle use resulting from the proposal would in turn generate both
                benefits and costs to the remainder of the U.S. economy. The analysis
                described these as ``external'' effects, in the sense that they were
                by-products of households' choices among new vehicle models, decisions
                about keeping older cars and light trucks in service, and allocations
                of driving across the fleet that were experienced broadly throughout
                the U.S. economy, rather than by the individuals making such decisions.
                The largest of these was additional refining and consumption of
                petroleum-based fuel and the associated increases in emissions of
                carbon dioxide and other gases, which were projected to increase the
                cost of economic damages inflicted on the U.S. economy by future
                changes in the global climate (line 13). Added fuel production and use
                under the proposal also led to higher emissions of localized air
                pollutants, and the resulting increase in the U.S. population's
                exposure and its adverse effects on health imposed additional external
                costs (line 14).
                 Increased consumption of petroleum-derived fuel also imposed higher
                external costs on the U.S. economy, in the form of potential losses in
                economic output and costs to businesses and households for adjusting to
                any sudden changes in energy prices (line 15 of the table). Reduced
                driving by buyers of new cars and light trucks in response to their
                higher operating costs also reduced the external costs from their
                contributions to traffic delays and noise, benefits that were expected
                to be experienced throughout the U.S. economy (line 17). Finally, some
                of the higher fuel costs to buyers of new cars and light trucks will
                consist of increased fuel taxes; this increase in revenue was projected
                to enable Federal and State government agencies to improve upkeep of
                roads and highways, fund increases in other services, or reduce other
                tax burdens (line 18).\1522\
                ---------------------------------------------------------------------------
                 \1522\ In some States, levies on gasoline include both general
                sales taxes as well as excise taxes, and not all proceeds are
                dedicated to transportation purposes.
                ---------------------------------------------------------------------------
                 The net economic effect (line 22) of the proposal consisted of the
                benefits and costs imposed directly on car and light truck
                manufacturers, accompanying indirect effects on buyers of new vehicles
                and owners of used ones, external costs driving decisions generated
                throughout the U.S. economy, and changes in revenue to government
                agencies. The agencies' organization was intended to convey the causal
                connections among these impacts, by highlighting how the proposed
                change in fuel economy standards faced by manufacturers would set in
                motion the sequence of behavioral responses that determined its
                economy-wide costs and benefits. This contrasted with the way benefits
                and costs of previous proposals to establish CAFE and CO2
                standards were analyzed and presented, which obscured their sequence
                and causal connections.
                 In those previous analyses, most economic effects other than
                manufacturers' costs to comply with proposed standards and anticipated
                changes in fuel consumption were grouped together and reported as ``co-
                benefits.'' This obscured how these various consequences arose from the
                proposed standards, providing no information about who would ultimately
                experience the costs of complying with the standards, or who would
                experience their direct and indirect benefits. In contrast, the recent
                analysis spelled out how each category of benefits and costs resulted
                from the proposed change in standards, identified the mechanisms that
                translated direct economic impacts into indirect costs and benefits,
                and distinguished between those arising from changes in fuel
                consumption, and safety consequences of changes in vehicle use. The
                proposal's framework also clarified who would bear each category of
                impacts, distinguishing between the proposal's economic impacts on
                private actors--vehicle manufacturers, new car and light truck buyers,
                and owners of used vehicles--and the external economic consequences for
                the general public and government agencies that stem indirectly from
                such private impacts.
                2) Final Rule Framework
                 While the agencies received several comments about which economic
                effects are included in the analysis, the agencies received no comments
                about the specific structure of the framework. Substantive comments
                about individual
                [[Page 24587]]
                effects are addressed over the next several sections.
                 The agencies have expanded the accounting framework for benefits
                and costs shown in Table VI-146 above to include two additional
                entries, as well as to distinguish financial impacts on government
                agencies from externalities borne broadly across the remainder of the
                U.S. economy. The revised accounting framework for costs and benefits
                is shown in Table VI-147, below. Line 6 of the revised table reports
                the change in consumer surplus experienced by buyers of new cars and
                light trucks when prices and sales of those vehicles adjust in response
                to changes in CAFE and CO2 standards. The gain in consumer
                surplus that occurs when production costs and prices for vehicles fall
                and sales increase in response represents a benefit to buyers, while
                any loss in consumer surplus that occurs when more stringent standards
                increase costs and prices and cause sales to decline appears as a loss
                to new car and light truck buyers.
                 Line 7 of Table VI-147 reports the estimated value of changes to
                attributes of new cars and light trucks other than fuel economy that
                their manufacturers make to comply with changes in CAFE and
                CO2 standards. In the case where standards are less
                stringent, manufacturers are able to employ many of the same resources
                they would have deployed to increase fuel economy for the alternative
                purpose of improving other attributes of vehicles that their potential
                buyers value more highly than the forgone improvements in fuel economy.
                This response provides an additional benefit to purchasers of new cars
                and light trucks that was not recognized in the agencies' analysis of
                the proposal, but is included in the analysis of this final rule. Of
                course, if CAFE and CO2 standards are made more stringent,
                manufacturers employ those technologies to increase fuel economy, thus
                sacrificing potential improvements in competing attributes--those that
                entail tradeoffs with higher fuel economy--and the value of
                improvements in those other attributes that is sacrificed or forgone
                represents an opportunity cost to those buyers. This implicit
                opportunity cost is analyzed in a sensitivity analysis and is not
                included in the primary analysis.
                 Finally, the agencies revised the framework for reporting costs and
                benefits of changes in CAFE and CO2 standards to identify
                government agencies separately from the entry previously labeled ``Rest
                of U.S Economy.'' This minor revision is intended to distinguish more
                clearly between changes in external costs imposed by externalities that
                result from fuel production and use, and the revenue effects on
                government agencies from changes in tax and civil penalty payments.
                While both effects ultimately result from manufacturers' compliance
                with revised standards and the resulting changes in fuel consumption,
                externalities represent real economic costs; in contrast, changes in
                tax revenues received by government agencies are financial transfers,
                whose offsetting effects on manufacturers and vehicle buyers are also
                recognized elsewhere in the accounting framework.
                BILLING CODE 4910-59-P
                [[Page 24588]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.296
                BILLING CODE 4910-59-C
                b) Economic Assumptions
                 The agencies' analysis of CAFE and CO2 standards for the
                model years covered by this final rule rely on a range of forecast
                information, estimates of economic, safety, and environmental
                variables, and input parameters. While the analysis accompanying the
                proposal largely resembled previous CAFE and CO2 analyses,
                the agencies updated many of the underlying inputs and assumptions--
                based on the most up-to-date data--and expanded the central analysis to
                account for changes in new vehicle sales and the retirement of older
                vehicles.
                 EDF, UCS, CARB and others commented that the agencies acted
                arbitrarily and capriciously by changing inputs and assumptions from
                previous analyses, and argued that the agencies failed to provide
                ``good reasons'' for the changes.\1523\ In the following sections, the
                agencies will respond directly to these comments. However, the agencies
                note that it would be uncommon to retain inputs and assumptions from
                prior analyses--which are typically informed by transitory empirical
                observations--on the basis of precedent. The agencies are ``neither
                required nor
                [[Page 24589]]
                supposed to regulate the present and the future within the inflexible
                limits of yesterday.'' \1524\
                ---------------------------------------------------------------------------
                 \1523\ See, e.g,. IPI, Appendix, NHTSA-2018-0067-12213, at 99-
                100.
                 \1524\ American Trucking Associations v. Atchison, 387 U.S. 397,
                416 (1967).
                ---------------------------------------------------------------------------
                 The agencies also received a number of comments focused on the
                agencies' attempt to incorporate the effects of changes in new vehicle
                prices on new vehicle sales, retirement rates of used vehicles, and the
                resulting ``turnover'' of the vehicle fleet. Some comments endorsed the
                agencies' more comprehensive analysis, although many of those same
                commenters later disagreed with aspects of the results. For example,
                RFF noted that ``Incorporating sales and scrappage effects represents a
                step in the right direction for modeling the effects of the
                regulation.'' \1525\ Similarly, NRDC stated that ``it is reasonable and
                appropriate to develop a mechanism for estimating future vehicle
                populations, and the NPRM documents appropriately present considerable
                discussion on the topic and the derivation of the utilized algorithm.''
                \1526\ One commenter explicitly recognized that the narrower analysis
                utilized in previous rules likely led to incorrectly estimating costs
                and benefits, and endorsed the broader approach used by the proposal.
                Specifically, American Fuel & Petrochemical Manufacturers stated that
                the absence of scrappage in prior rules ``likely led to a significant
                overestimation of the existing standard's benefits with respect to fuel
                and air pollutant emission reductions and an underestimation of safety
                risks and societal costs.'' FCA also expressed general support for the
                agency's expanded analysis.\1527\
                ---------------------------------------------------------------------------
                 \1525\ Resources for the Future, NHTSA-2018-0067-11789, at 2.
                 \1526\ Meszler Engineering Services & Baum and Associates, on
                behalf of Natural Resources Defense Council, NHTSA-2018-0067-11943-
                43, NHTSA-2018-0067-11723.
                 \1527\ FCA, NHTSA-2018-0067-12078.
                ---------------------------------------------------------------------------
                 In contrast, some commenters objected to the inclusion of `new'
                impacts, including the effect of fuel economy regulations on new
                vehicle prices, the resulting changes in their sales, and retirement
                rates for used cars. Workhorse Group, Inc. noted that the agencies
                ``made novel assumptions about the safety impacts of consumers delaying
                vehicle purchases due to the increased costs of fuel economy
                improvements that contradicts the analytical approach NHTSA has
                followed in all prior safety and CAFE rulemakings.'' \1528\ Honda
                agreed ``that significantly higher-priced new vehicles have the
                potential to depress the new vehicle market and thus increase the fleet
                of used vehicles, with concomitant increased safety risks associated
                with driving greater numbers of older vehicles in lieu of newer ones,''
                but found it ``premature and ill-advised'' to model the impact of fleet
                turnover.\1529\ CBD et. al. argued that the sales and scrappage effects
                were too uncertain to include in the analysis and cited EPA's 2016
                proposed determination as stating, ``a reasonable qualitative
                assessment is preferable to a quantitative estimate lacking sufficient
                basis, or (due to uncertainties like those here) having such an
                enormous range as to be without substantial value.'' \1530\
                ---------------------------------------------------------------------------
                 \1528\ Workhorse Group, Inc., NHTSA-2018-0067-12215.
                 \1529\ American Honda Motor Company, Inc., NHTSA-2018-0067-
                11818.
                 \1530\ Environmental group coalition, Appendix A, NHTSA-2018-
                0067-12000, at 174.
                ---------------------------------------------------------------------------
                 As was done repeatedly throughout the proposal, the agencies
                acknowledge that dynamically modeling fleet turnover is new for this
                rulemaking; however, the agencies disagree that the analysis relied on
                `novel' assumptions or contradicted previous analyses. The agencies
                have described the sales and scrappage responses similarly in prior
                rulemakings,\1531\ and have expressed an interest in quantitatively
                measuring them.\1532\ The agencies agree with commenters that--like
                many of the effects included in today's analysis--there remains a
                degree of uncertainty about the magnitude of the sales and scrappage
                responses. However, CBD v. NHTSA stressed that a variable should not be
                excluded from the analysis simply because it is uncertain when the
                effect is quantifiable, ``certainly not zero,'' and the analysis
                ``monetize[s] other uncertain benefits.'' \1533\ As discussed in the
                coming sections, the agencies are confident that (a) changes in new
                vehicle prices impact the volume of new vehicle sales and rate of
                retirement of older vehicle, (b) of the direction of those effects, and
                (c) their ability to reasonably estimate the impacts. As such, the
                agencies strongly believe that including the sales and scrappage
                responses improves the thoroughness of the analysis, is consistent with
                case law, and is necessary to comprehensively analyze the cost-benefits
                of the rule.
                ---------------------------------------------------------------------------
                 \1531\ See, e.g., 76 FR 75153.
                 \1532\ See, e.g., 77 FR 61971.
                 \1533\ 538 F.3d 1172, 1200-02 (2008).
                ---------------------------------------------------------------------------
                 The following subsections briefly describes the sources of the
                agencies' estimates of each of the economic, environmental, and safety
                estimates. In reviewing these variables and the agencies' estimates of
                their values for purposes of this final rule, NHTSA and EPA considered
                comments received in response to the proposed rule and, in response,
                made several changes to the economic assumptions used for the final
                analysis.
                1) Macroeconomic Assumptions That Affect the Agencies' Analysis
                 As the proposed rule noted, the more comprehensive economic impact
                analysis of CAFE and CO2 included in this rulemaking
                requires a more detailed and explicit explanation of the macroeconomic
                context in which regulatory alternatives are evaluated. The agencies
                continued to rely on projections of future fuel prices to evaluate
                manufacturers' use of fuel-saving technologies, the resulting changes
                in fuel consumption, and various other benefits. Furthermore, the
                agencies expanded the scope of their analysis to include projecting
                future sales of new cars and light trucks, as well as the retirement of
                used vehicles under each regulatory alternative. In addition to
                projections of future fuel prices, constructing these forecasts
                requires explicit projections of macroeconomic variables, including
                U.S. Gross Domestic Product (GDP), labor force participation (the
                number of persons employed or actively seeking employment), and
                bellwether interest rates, which are likely to vary according to
                roughly the same pattern as interest rates on new car loans.
                 The analysis presented in the proposal as well as the accompanying
                RIA and EIS employed forecasts of future fuel prices developed by the
                agencies using the U.S. Energy Information Administration's (EIA's)
                National Energy Model System (NEMS). An agency within the U.S.
                Department of Energy (DOE), EIA collects, analyzes, and disseminates
                independent and impartial energy information to promote sound
                policymaking, efficient markets, and public understanding of energy and
                its interaction with the economy and the environment. EIA uses NEMS to
                produce its Annual Energy Outlook (AEO), which presents forecasts of
                future fuel prices, among many other energy-related variables. AEO
                projections of energy prices and other variables are not intended as
                predictions of what will happen; rather, they are projections of the
                likely course of these variables that reflect their past relationships,
                specific assumptions about future developments in global energy
                markets, and the forecasting methodologies incorporated in NEMS. Each
                AEO includes a ``Reference'' case as well as a range of alternative
                scenarios that each incorporate
                [[Page 24590]]
                somewhat different assumptions from those underlying the Reference
                Case.
                 For the proposal, the agencies used the AEO2017 version of NEMS, as
                this was the most current version of the model that was available at
                the time. Using this version of NEMS, the agencies reevaluated the
                ``Reference,'' ``Low Oil Price,'' and ``High Oil Price'' cases
                described in AEO2017, by setting aside their assumption that mandates
                by California and other States to sell ``Zero Emission Vehicles''
                (ZEVs) would be enforced. The agencies used the resulting modified
                Reference case fuel prices as inputs to the proposal's central case
                results, and used the modified ``Low Oil Price'' and ``High Oil Price''
                case fuel prices, which were generated using NEMS, as inputs to several
                of the sensitivity analysis cases that were presented in the proposal.
                The sensitivity analysis also included a case that applied the
                Reference case fuel prices from the then recently issued AEO2018, which
                did not reflect the modification of EIA's forecasting model to set
                aside state mandates for ZEV sales.\1534\
                ---------------------------------------------------------------------------
                 \1534\ The results of these and other sensitivity analyses were
                reported in NHTSA and EPA, ``Notice of Proposed Rulemaking: The
                Safer Affordable Fuel-Efficient (SAFE) Vehicles Rule for Model Years
                2021-2026 Passenger Cars and Light Trucks,'' Federal Register Vol.
                83, No. 165, August 24, 2018, Tables Vii-90 to Vii-98, pp. 43353-69.
                ---------------------------------------------------------------------------
                 The analysis supporting the proposed rule simulated the economic
                impacts of car and light truck manufacturers' compliance with
                alternative CAFE and CO2 standards through model year 2032,
                and in doing so estimated the number of vehicles originally produced
                and sold in each model year that would remain in service during each
                year of their useful lives (assumed to extend for a maximum of 40
                years), as well as their usage, fuel consumption, and safety
                performance. This required the forecasts of macroeconomic variables
                that affect vehicle sales, use, and retirement rates, which include
                U.S. Gross Domestic Product (GDP), the size of the domestic labor
                force, and key interest rates, to extend well beyond calendar year
                2050. One of the few sources that provides forecasts of these variables
                spanning such a long time horizon was the 2017 OASDI Trustees Report
                from the U.S. Social Security Administration, and the analysis
                supporting the proposed rule relied on this source for forecasts of
                these key macroeconomic measures.\1535\
                ---------------------------------------------------------------------------
                 \1535\ Social Security Administration, The 2017 Annual Report of
                the Board of Trustees of the Federal Old-Age and Survivors Insurance
                and Federal Disability Insurance Trust Funds, available at https://www.ssa.gov/OACT/TR/2017/.
                ---------------------------------------------------------------------------
                (a) Comments on the Fuel Price Forecasts and Macroeconomic Assumptions
                Used in the NPRM Analysis
                 The agencies received relatively few comments on the projections of
                fuel prices and macroeconomic variables that were used in their
                analysis supporting the proposed rule, virtually all of them focused on
                the fuel price projections the agencies employed. While only one
                comment questioned the agencies' use of price projections that rely on
                EIA's methodology and assumptions, a few commenters called attention to
                the unreliability of price projections reported in earlier editions of
                AEO. Other comments noted the importance of updating projections used
                to analyze the proposal to reflect more recent developments in energy
                markets, without necessarily questioning the reliability of EIA's fuel
                price projections. Several comments emphasized the implications for the
                agencies' analysis of the wide variation in alternative fuel price
                projections reported in both EIA's 2017 and 2018 Annual Energy
                Outlooks, with most stressing the possibility that future prices might
                be above even those projected in their High Oil Price cases. Only a
                single comment identified a potential alternative source of fuel price
                projections, but noted that it was within the range of projections the
                agencies considered.
                 One commenter claimed that AEO's projections of fuel prices are
                ``inappropriate'' for the agencies to employ in analyzing the
                consequences of CAFE and CO2 standards; because EIA ``does
                not speculate on changes in international policy or geopolitics,''
                which contribute to the uncertainty surrounding future prices.\1536\
                However, this commenter did not identify an alternative source for fuel
                price projections that reflect such considerations; and because
                projections of fuel prices are a central element in the agencies'
                evaluation of alternative future standards, the observation that EIA's
                projections do not incorporate some sources of uncertainty is unhelpful
                by itself.
                ---------------------------------------------------------------------------
                 \1536\ NHTSA-2018-0067-11837, Alliance to Save Energy, p. 2
                (``EIA takes a transparently conservative approach in modeling
                future oil prices, and does not speculate on changes in
                international policy or geopolitics. As a result, their projections
                are an inappropriate measure of future fuel prices.'').
                ---------------------------------------------------------------------------
                 Some commenters asserted that by relying on the AEO2017 Reference
                Case projections of fuel prices in their central analysis of the
                proposed rule while considering the significantly higher fuel prices
                projection in the AEO High Oil Price scenario only in the accompanying
                sensitivity analyses, the agencies inadequately considered the possible
                effect of higher fuel prices on the estimated economic benefits from
                alternatives that would have relaxed the augural standards, including
                the preferred alternative.\1537\ Surprisingly, none of these comments
                acknowledged that the fuel price projections reported in the High Oil
                Price cases accompanying past editions of the Annual Energy Outlook
                have so far proven to be significantly above actual prices, or that EIA
                has consistently lowered its fuel price projections in more recent
                editions of the AEO. In any case, supplemental material included in the
                NPRM regulatory docket showed that the ranking of regulatory
                alternatives by their estimated net economic benefits remained
                unchanged from the central analysis in the sensitivity analysis that
                substituted the AEO2017 High Oil Price case projection of fuel prices.
                ---------------------------------------------------------------------------
                 \1537\ See e.g., Securing America's Future Energy (SAFE), NHTSA-
                2018-0067-11981, pp. 12 & 30 and Institute for Policy Integrity,
                NHTSA-2018-0067-12213, p. 31.
                ---------------------------------------------------------------------------
                 None of the commenters who argued that the agencies inadequately
                considered the possibility of higher fuel prices observed that the
                agencies' analogous use of lower fuel price projections from the
                AEO2017 Low Oil Price case only in their sensitivity analyses
                inadequately considered the possibility that future fuel prices might
                prove to be lower than projected in the AEO2017 Reference Case, and its
                potential effect on the proposal's estimated benefits. Nor did any of
                the commenters offer substantive guidance about how the agencies might
                revise their analysis to accord greater emphasis to fuel price
                projections above (or below) those from the AEO Reference Case.\1538\
                ---------------------------------------------------------------------------
                 \1538\ One commenter did refer to guidance to EPA contained in a
                National Research Council report on incorporating and conveying
                uncertainty about key inputs directly into that agency's estimates
                of benefits from reducing air pollution, rather than simply
                recognizing it in supplemental sensitivity analyses. This was
                presumably intended as potential guidance to the agencies about how
                they might do so in their evaluations of fuel economy and
                CO2 standards, although that was not stated explicitly.
                See American Fuel & Petrochemical Manufacturers, NHTSA-2018-0067-
                12078, p. 19, citing National Research Council (2002), Estimating
                the Public Health Benefits of Proposed Air Pollution Regulations,
                2002, available at https://www.nap.edu/catalog/10511/estimating-the-public-health-benefits-of-proposed-air-pollution-regulations.
                ---------------------------------------------------------------------------
                 Other comments stressed the fact that EIA's current projections of
                future fuel prices are significantly lower than those the agencies
                relied on when they established CAFE standards through
                [[Page 24591]]
                model year 2021 and introduced the augural standards for subsequent
                model years in the rulemaking they conducted in 2012, citing this as
                support for the agencies' reconsideration of the augural standards in
                the current rulemaking.\1539\
                ---------------------------------------------------------------------------
                 \1539\ For example, Fiat Chrysler Automobiles (FCA) pointed out
                that the AEO 2017 Reference Case forecast of gasoline prices through
                2025 is approximately 36% lower than that in the AEO 2012 Reference
                Case, which the agencies relied on in the analysis supporting that
                earlier rulemaking; see NHTSA-2018-0067-11943, p. 33.
                ---------------------------------------------------------------------------
                 One comment compared the range of fuel price projections spanned by
                the High and Low Oil Price cases from AEO2017 and AEO2018 to the range
                of future prices spanned by another widely-recognized and relied-upon
                projection, concluding that the alternative scenarios included in
                AEO2017 incorporated an even wider range of uncertainty about future
                prices, and noted that the net economic benefits of the preferred
                alternative were positive over this entire range of alternative future
                fuel prices. This same commenter noted that by combining high and low
                fuel price projections with alternative assumptions about other key
                economic variables (such as GDP growth) and parameter assumptions
                (principally payback period), the agencies' sensitivity analyses
                captured potentially important interactions between uncertainty
                regarding fuel prices and other key economic inputs.\1540\
                ---------------------------------------------------------------------------
                 \1540\ See Alliance of Automobile Manufacturers, NHTSA-2018-
                0067-1207, p. 108.
                ---------------------------------------------------------------------------
                (b) Macroeconomic Assumptions Used To Analyze Economic Consequences of
                the Final Rule
                 After considering these comments, the agencies have concluded that
                there is no convincing reason to rely on sources other than EIA's NEMS
                model to project future energy prices, or to rely on alternatives to
                the Reference Case scenario in the current edition of AEO as their
                basis for using NEMS. The agencies agree that the resulting projections
                will be uncertain, but note that EIA regularly publishes retrospective
                analyses comparing past Reference case projections to subsequent market
                price outcomes, thus enabling an assessment of this uncertainty.
                Although EIA does not identify its Reference case as a ``most likely''
                outcome, in the agencies' judgment that case's design--which assumes
                future trends are consistent with historical and current market
                behavior--makes it a reasonable and appropriate basis for projecting
                fuel prices to use in the agencies' central analysis of alternative
                CAFE and CO2 standards.
                 The agencies also conclude that the wide range of uncertainty about
                future petroleum prices encompassed in EIA's ``Low Oil Price'' and
                ``High Oil Price'' cases means that including them in the accompanying
                sensitivity analyses provides a meaningful basis for assessing the
                potential economic consequences of future energy prices that prove to
                be considerably lower or higher than those reflected in the Reference
                case. Although these alternative cases do not incorporate unbridled
                speculation regarding hypothetical changes in ``international policy or
                geopolitics,'' the agencies believe that this restraint means that
                relying on them produces a more, rather than less, meaningful test of
                the effect of the inherent uncertainty surrounding projections of fuel
                prices.
                 For today's final rule, the agencies have therefore used the
                AEO2019 version of NEMS to develop projections of future prices for
                transportation fuels, as this was the most current version available
                when this analysis was conducted. Using this version of NEMS, the
                agencies modified EIA's AEO2019 Reference case by (1) setting aside
                presumed enforcement by California and other States of any mandates to
                sell ``Zero Emission Vehicles'' (ZEVs), (2) setting aside post-2020
                increases in the stringency of CAFE and CO2 standards, and
                (3) modifying inputs regarding battery costs, in order to bring those
                costs down to levels more consistent with battery cost estimates
                applied in the CAFE model analysis.\1541\ All other NEMS inputs used to
                develop the AEO2019 Reference case were left unchanged in this
                analysis.
                ---------------------------------------------------------------------------
                 \1541\ These inputs are all contained in the ``trnldvx.xlsx''
                NEMS input file. The input file utilized for today's analysis is
                available in regulatory docket NHTSA-2018-0067, https://www.regulations.gov/docket?D=NHTSA-2018-0067 (see Supporting
                Documents), as is the corresponding output file from which reference
                case fuel and electricity prices were obtained to be used as inputs
                to the CAFE model. The version of NEMS utilized for today's analysis
                is available at https://www.eia.gov/outlooks/aeo/info_nems_archive.php.
                ---------------------------------------------------------------------------
                 Setting aside enforcement of state mandates to sell ZEVs makes the
                supporting analysis consistent with the agencies' recent One National
                Program Action,\1542\ under which EPA withdrew aspects of a Clean Air
                Act Preemption waiver previously granted to California, and NHTSA
                concluded that EPCA expressly and implied preempted State ZEV mandates.
                Setting aside the post-2020 increase in the stringency of CAFE and
                CO2 standards ensures that the fuel prices used in the
                agencies' analysis are at least as high as those that would prevail
                under the least stringent regulatory alternative considered, since that
                alternative produces the highest level of fuel consumption and thus the
                highest fuel prices.
                ---------------------------------------------------------------------------
                 \1542\ 84 FR 51310.
                ---------------------------------------------------------------------------
                 Figure VI-55 and Figure VI-56 below show the resulting modified
                projections of BEV prices and sales, and compare them to the
                projections reported in EIA's AEO2019 Reference case. As they
                illustrate, the combination of these modifications led NEMS to project
                significantly lower BEV prices and correspondingly higher BEV sales
                volumes. Figure VI-57 and Figure VI-58 show the modified projections of
                gasoline and electricity prices, and again compare these to the
                projections reported in EIA's AEO2019 Reference case. As those figures
                indicate, the agencies' modifications to NEMS did not significantly
                affect its projections of future prices for transportation fuels.
                BILLING CODE 4910-59-P
                [[Page 24592]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.297
                BILLING CODE 4910-59-C
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.298
                BILLING CODE 4910-59-C
                BILLING CODE 4910-59-P
                [[Page 24593]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.299
                BILLING CODE 4910-59-C
                 The agencies used the resulting Reference case fuel prices as
                inputs to the rule's central analysis. The agencies also used the as-
                published (by EIA) ``Low Oil Price'' and ``High Oil Price'' case fuel
                prices as inputs to several of the cases included in the sensitivity
                analysis presented in the accompanying RIA.
                 For the projections of macroeconomic variables used in the analysis
                supporting this rule, the agencies elected to rely on different sources
                from those that informed their analysis of the proposed rule.
                Specifically, the agencies rely on projections of future growth in U.S.
                GDP reported in AEO2019 to support their central analyses of the final
                rule's impacts on new car and light truck sales and the retirement of
                used vehicles. These incorporate underlying projections generated using
                the IHS Markit Global Insight long-term macroeconomic model, as
                modified via this model's interaction with NEMS' representation of
                global energy markets and their future outcomes. The alternative
                projections of future growth in GDP used in the agencies' accompanying
                sensitivity analyses are drawn from the AEO2019 High Economic Growth
                and Low Economic Growth cases. These reflect alternative future trends
                in U.S. labor force and productivity growth, and are also consistent
                with the energy market outcomes projected by NEMS under the resulting
                future performance of the U.S. economy.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.300
                [[Page 24594]]
                 For estimates of the number of U.S. households during future years,
                which influence the projections of new car and light truck sales used
                in the analysis, the agencies rely on projections of new household
                formation developed the Harvard University Joint Center for Housing
                Studies.\1543\ These are consistent with the most recent projections of
                future growth in the nation's population prepared by the U.S. Bureau of
                the Census.\1544\
                ---------------------------------------------------------------------------
                 \1543\ See Harvard University Joint Center for Housing Studies,
                Updated Household Growth Projections: 2018-2028 and 2028-2038,
                December 18, 2018, available at https://www.jchs.harvard.edu/sites/default/files/Harvard_JCHS_McCue_Household_Projections_Rev010319.pdf.
                 \1544\ Ibid., pp. 2-5.
                ---------------------------------------------------------------------------
                (2) Approach To Estimating Sales Response Under Different Standards
                 Prior to the NPRM, all previous CAFE and CO2 rulemaking
                analyses used static fleet forecasts that were based on a combination
                of manufacturer compliance data, public data sources, and proprietary
                forecasts (or product plans submitted by manufacturers). When
                simulating compliance with regulatory alternatives, those analyses
                projected identical sales across the alternatives, for each
                manufacturer down to the make/model level--where the exact same number
                of each model variant was assumed to be sold in a given model year
                under both the least stringent alternative (typically the baseline) and
                the most stringent alternative considered (intended to represent
                ``maximum technology'' scenarios in some cases). To the extent that an
                alternative matched the assumptions made in the production of the
                proprietary forecast, using a static fleet based upon those assumptions
                may have been warranted. However, a sales forecast is unlikely to be
                representative of a broad set of regulatory alternatives with
                significant variation in the cost of new vehicles. A number of
                commenters on previous regulatory actions encouraged consideration of
                the potential impact of fuel efficiency standards on new vehicle prices
                and sales, and the changes to compliance strategies that those shifts
                could necessitate.\1545\ In particular, the continued growth of the
                utility vehicle segment creates compliance challenges within some
                manufacturers' fleets as sales volumes shift from one region of the
                footprint curve to another, or as mass is added to increase the ride
                height of a vehicle on a sedan platform to create a crossover utility
                vehicle, which exists on the same place of the footprint curve as the
                sedan upon which it might be based.
                ---------------------------------------------------------------------------
                 \1545\ See, e.g., Alliance of Automobile Manufacturers, Comment,
                EPA-HQ-OAR-2015-0827-4089, at 115-16.
                ---------------------------------------------------------------------------
                 However, some NPRM commenters referenced the agencies' previous
                omission of this effect as justification to continue ignoring this
                issue in the current rulemaking. EDF commented,\1546\ ``use of a sales
                response model constitutes an unexplained reversal in the agency's
                position on the feasibility of doing so.'' To say that the agencies
                never used a model is a misrepresentation. Assuming that sales never
                change in any model year, even at the individual nameplate level,
                regardless of the stringency of fuel economy regulations or the
                technology costs required to comply with those regulations, is, itself,
                a model. It is a model that implicitly asserts that, while fuel economy
                regulation impacts vehicle prices, such regulations have no impact on
                the quantity or mix of new vehicle sold, regardless of stringency. This
                is an implicit argument that new vehicle demand is perfectly
                inelastic--and that no change in vehicle prices can impact the number
                of cars consumers will buy. Logically, however, there must exist a
                level of stringency that would have a negative impact on new sales.
                Picking an extreme example to prove the point, if the agencies set
                standards at an extraordinarily stringent level that forced all
                vehicles into battery electric propulsion systems next year, sales
                would obviously be impacted. The increase in new vehicle price or
                changes to other relevant attributes like range, refueling time, or
                operating cost would surely affect the decisions of some buyers. But,
                by arguing that the agencies should continue to model new vehicle sales
                as if they are entirely unaffected by standards, commenters are
                effectively asking the agencies to assume that the alternatives
                considered in this rule are insufficiently stringent to affect the
                market. By endorsing the approach from the 2012 final rule, which
                assumed no impact on the new vehicle market from standards as stringent
                as 7 percent increase, year-over-year, beginning in 2017, commenters
                are suggesting that even those standards would have no impact on new
                vehicle sales. Manufacturers have asserted in their comments that fuel
                economy regulations change both the cost of producing new vehicles and
                consumer demand for them. In the recent peer review of the NPRM release
                of the CAFE model, all reviewers encouraged the inclusion of a sales
                response to fuel economy regulations (albeit not necessarily the
                version of the response model that appeared in the NPRM).\1547\ Based
                on earlier comments and the agencies' own analysis, the agencies were
                persuaded to include a sales response mechanism in the NPRM, and do so
                again in this final rule.
                ---------------------------------------------------------------------------
                 \1546\ EDF, Appendix B, NHTSA-2018-0067-12108, at 37-38.
                 \1547\ CAFE Model Peer Review, DOT HS 812 590, Revised (July
                2019), available at https://www.regulations.gov/contentStreamer?documentId=NHTSA-2018-0067-0055&attachmentNumber=2&contentType=pdf.
                ---------------------------------------------------------------------------
                 While several commenters (CARB, NCAT, CBD, Aluminum Association)
                discouraged the agencies from attempting to account for the effect of
                regulations on new vehicle sales, other commenters stated that the NPRM
                analysis was improved by explicitly considering this effect (RFF,
                Toyota, the Alliance of Automobile Manufacturers). CBD cited EPA's 2016
                proposed determination, stating ``[a] reasonable qualitative assessment
                is preferable to a quantitative estimate lacking sufficient basis, or
                (due to uncertainties like those here) having such an enormous range as
                to be without substantial value.'' \1548\ However, RFF supported the
                inclusion of the effect (with caveats about the specific
                implementation, for which they suggested alternative approaches),
                stating ``[i]ncorporating sales and scrappage effects represents a step
                in the right direction for modeling the effects of the
                regulation.\1549\ It is reasonable to conclude that regulations as
                transformative as fuel economy standards will impact the market for new
                vehicles, and excluding the effect (as CBD and others suggested) is
                equivalent to stating that it does not exist.
                ---------------------------------------------------------------------------
                 \1548\ Environmental group coalition, Appendix A, NHTSA-2018-
                0067-12000, at 174.
                 \1549\ RFF, Comments, NHTSA-2018-0067-11789, at 3.
                ---------------------------------------------------------------------------
                 The NPRM version of the sales response relied on differences in the
                average price of new vehicles to produce sales differences between
                regulatory alternatives. Some commenters (ACEEE, IPI, CBD, UCS,
                Aluminum Association, and Alliance to Save Energy) argued that new
                vehicle prices do not increase with the addition of technology required
                to comply with fuel economy regulations. Some argued that manufacturers
                will choose not to ``pass through'' the full incremental cost of fuel
                saving technologies to consumers, instead absorbing those costs into
                their profit margin.\1550\ The question of cost pass-through is one
                that academic and industry researchers have considered for decades--and
                two of the
                [[Page 24595]]
                agencies' recent peer reviewers addressed this issue in their comments.
                ---------------------------------------------------------------------------
                 \1550\ E.g. IPI, Appendix, NHTSA-2018-0067-12213, 28-29; CBD et
                al., Attachment 1, NHTSA-2018-0067-12123, at 23-24.
                ---------------------------------------------------------------------------
                 Dr. John D. Graham, one of the peer reviewers, argued that the
                assumption of complete cost pass-through is defensible, and more likely
                in the long-run than the short-run.\1551\ The reviewer also suggested
                that changes to the CAFE (and subsequent CO2) program that
                base a manufacturer's standard on the mix of vehicle footprints in each
                fleet more equitably spreads the impact of the standards across the
                industry, and that industry shifts toward increasingly competitive
                market models (rather than the oligopolistic models that existed
                earlier in the last century) both act to increase the likelihood that
                manufacturers will pass regulatory costs through to consumers. In
                particular, this reviewer stated: \1552\
                ---------------------------------------------------------------------------
                 \1551\ CAFE Model Peer Review, DOT HS 812 590, Revised (July
                2019), pp. B31-B33, available at https://www.regulations.gov/contentStreamer?documentId=NHTSA-2018-0067-0055&attachmentNumber=2&contentType=pdf.
                 \1552\ Gron Anne, Swenson, Deborah L, Cost Pass-Through in the
                US Automobile Market, Review of Economics and Statistics, Vol. 82(2)
                (May 2000), at 3.
                 In a classic study, Gron and Swenson (2000) examined list prices
                of automobiles at the model level in the U.S. from 1984 to 1994
                coupled with data on production, vehicle characteristics, foreign
                versus domestic firm ownership, wages of employees, exchange rates,
                imported parts content, tariffs and other variables. Although their
                work rejects the hypothesis of 100% pass through of cost to consumer
                price, they find higher rates of pass through than previous studies,
                and much of the incomplete pass through occurs when cost increases
                impact only a few models or firms. Confirming earlier studies, they
                show that U.S. auto manufacturers engage in more aggressive pass-
                through pricing than Asian and European manufacturers (greater than
                100% in some specifications), possibly due to the eagerness of
                importers to enlarge market share in lieu of recovering regulatory
                costs, at least in the short run (see Dinopolous and Kreinin, 1988;
                \1553\ Froot, 1989 \1554\[hairsp]). This study helps explain why
                pass-through pricing is a more viable hypothesis in the long run
                than in the short run.
                ---------------------------------------------------------------------------
                 \1553\ Dinopoulos, Elias, Kreinin, Mordechai, Effects of U.S.-
                Japan Auto VER on European Prices and on U.S. Welfare, The Review of
                Economics and Statistics, Vol. 70(3) (1988), at 484-91.
                 \1554\ Froot, Kenneth A, Klemperer, Paul D, Exchange Rate Pass-
                Through When Market Share Matters, American Economic Review, Vol.
                79(4) (1989), at 637-54.
                ---------------------------------------------------------------------------
                 The original design of the CAFE program is a contrasting case
                where pass-through pricing was difficult for some automakers. All
                auto makers, regardless of their product mix, were subject to the
                same fleet-wide average CAFE standard, such as 27.5 miles per gallon
                for cars in 1990. In practice, those standards impacted only three
                high-volume companies (General Motors, Ford and Chrysler) because
                the Big Three produced a higher proportion of large and performance-
                oriented vehicles than did Japanese companies. As a result,
                manufacturers such as Toyota and Honda consistently surpassed the
                federal fleet-wide standard for cars without any regulatory cost
                (i.e., partly due to their smaller product mix). In the 1975-2007
                period, the Big Three were not able to pass on all of their
                compliance costs to consumers and thus experienced some declines in
                profitability due to CAFE (Kleit, 1990; \1555\ Kleit, 2004; \1556\
                Jacobsen, 2013\1557\[hairsp]).
                ---------------------------------------------------------------------------
                 \1555\ Kleit, Andrew N., The Effect of Annual Changes in
                Automobile Fuel Economy Standards, Journal of Regulatory Economics,
                Vol. 2. (1990,), at 151-72.
                 \1556\ Kleit, Andrew N, Impact of Long-Range Increases in the
                Fuel Economy (CAFE) Standard, Economic Inquiry, Vol. 42(2) (2004),
                at 279-94.
                 \1557\ Jacobsen, Mark R., Evaluating U.S. Fuel Economy Standards
                in a Model with Producer and Household Heterogeneity, American
                Economic Journal: Economic Policy, Vol. 5(2) (2013), at 148-87.
                ---------------------------------------------------------------------------
                 When the CAFE program was reformed for light trucks in 2008 (and
                for cars in 2011) on the basis of vehicle size (the so-called
                ``footprint'' adjustments to CAFE stringency), the, the technology
                costs of CAFE standards were spread more evenly among automakers,
                although the overall societal efficiency of the regulation
                diminished due to the removal of downsizing as a compliance
                option.\1558\ Given that the size-based fuel economy programs are
                not concentrating the costs of compliance on one or two automakers,
                it is reasonable to predict a fairly high degree of pass-through
                pricing for the 2021-2025 fuel economy standards. In related
                literature on manufacturer pricing responses to a national carbon
                tax, Bento and Jacobsen (2007) \1559\ and Bento (2013) \1560\ report
                high rates of pass-through pricing (on the order of 85%). Carbon
                taxes are more efficient than footprint-based CAFE standards, but
                both instruments are likely to impact a wide range of companies in
                the auto sector and result in a high degree of pass-through pricing
                by impacted companies.
                ---------------------------------------------------------------------------
                 \1558\ See Ito, Koichiro, Sallee, James M., The Economics of
                Attribute-Based Regulation: Theory and Evidence from Fuel-Economy
                Standards, Review of Economics and Statistics, in press (2018).
                 \1559\ Bento, Antonio M., Jacobsen, Mark R, Environmental Policy
                and the `double-dividend' hypothesis, Journal of Environmental
                Economics and Management, Vol. 53(1) (January 2007) at 17-31.
                 \1560\ Bento, Antonio M. Equity Impacts of Environmental Policy,
                Annual Review of Resource Economics, Vol. 5 (May 2013), at 181-96.
                ---------------------------------------------------------------------------
                 Also, it should be noted that the U.S. automotive industry is
                much more competitive today than it was from 1970 to 2000. The
                market share of General Motors, once the dominant, majority producer
                in the U.S. market, has declined dramatically, and a variety of
                Japanese and Korean companies have captured substantial market
                share. Moreover, the rise of startups (e.g., Tesla and other
                electric vehicle start-ups) and ride-sharing services (e.g., Uber)
                are adding a new competitive dimension in the U.S. industry. As a
                result, some of the most recent auto regulatory studies have given
                more emphasis to analytic results based on competitive models than
                oligopolistic models (see, e.g., Davis and Knittel (2016)
                \1561\[hairsp]).
                ---------------------------------------------------------------------------
                 \1561\ Davis, Lucas, Knittel, Christopher R., Are Fuel Economy
                Standards Regressive? Working Paper 22925, National Bureau of
                Economic Research, Cambridge, MA (2016).
                 Another peer reviewer, Dr. James Sallee, suggested that costs would
                pass through to new vehicle buyers to different degrees, depending upon
                the stringency of the standards.\1562\ The reviewer argued that more
                stringent standards, which result in larger increases to the cost of
                production, are likely to induce greater degrees of pass-through than
                less stringent standards, which automakers may, as some commenters have
                suggested, be able to absorb in the form of lost profit. If the degree
                of cost pass-through should vary by the stringency of the alternative,
                the agencies are underestimating the difference in price between the
                most and least stringent alternatives--which would favor alternatives
                with higher stringency.
                ---------------------------------------------------------------------------
                 \1562\ CAFE Model Peer Review, DOT HS 812 590, Revised (July
                2019), pp. B54-B75, available at https://www.regulations.gov/contentStreamer?documentId=NHTSA-2018-0067-0055&attachmentNumber=2&contentType=pdf.
                ---------------------------------------------------------------------------
                 Other commenters argued that manufacturers are able to compensate
                fully for the costs of fuel economy standards by increasing the prices
                of luxury vehicles--which would increase the average new vehicle price,
                but leave large sections of the market unaffected by the increased cost
                of producing fleets that comply with the standards. While it seems
                likely that manufacturers employ pricing strategies that push
                regulatory costs (as well as increases in costs like pension
                obligations and health care costs for employees) into the prices of
                models and segments with less elastic demand, the extent to which any
                OEM is able to succeed at this is unknown by the agencies. At some
                point, however, price increases on even luxury models will merely price
                more and more purchasers out of the market, and make competition with
                other manufacturers and market segments that much more difficult. And
                the more that avoided price increases for lower ends of the vehicle
                market are subsidized by luxury vehicles, the more either prices for
                luxury models would need to be increased, or (if moderately increasing
                prices) more of those luxury models would need to be sold. It is worth
                noting that luxury vehicles tend to be more powerful and content-rich,
                and often have fuel economy levels below (or CO2 levels
                above) their targets on the curves--so that selling more of them to
                compensate for lost profit elsewhere
                [[Page 24596]]
                further erodes the compliance levels of the fleets in which they
                reside.
                 While manufacturers could conceivably push some small cost
                increases into the prices of their vehicle segments that have less
                elastic demand to cover accordingly small increases in stringency,
                larger stringency increases would exhaust the ability of such segments
                to absorb additional costs. In addition, the agencies do not attempt to
                adjust the mix of vehicle models based on their own price elasticity of
                demand; doing so would require a pricing model that takes the
                compliance cost for each manufacturer (which the agencies' model
                estimates dynamically) and apportions that cost to the prices of
                individual nameplates and trim levels. The agencies have experimented
                with pricing models (when integrating vehicle choice models, pricing
                models are a necessity), but each manufacturer almost certainly has a
                unique pricing strategy that is unknown to the agencies, and involves
                both strategic decisions about competitive position within a segment
                and the volumes needed fully to amortize fixed costs associated with
                production. To the extent that the agencies assume all regulatory costs
                are passed through and affect the average regulatory cost of each
                vehicle instead of being priced in a fashion to minimize the impact on
                aggregate sales, the agencies note that--more stringent alternatives
                are provided an artificial analytical advantage because manufacturers
                are better positioned to incorporate smaller price adjustments into
                their current strategic pricing models. The agencies opted to take the
                conservative approach instead of speculating on manufacturer's private
                business models.
                 Finally, some commenters have argued that, even if regulations do
                increase the cost of producing vehicles and those costs are passed on
                to new vehicle buyers, it does not matter because sales have increased
                in recent years under both rising standards and rising prices. EDF,
                CARB, Aluminum Association, SAFE, CBD, and CA et al. and Oakland et
                al., all make some version of this argument in their comments.\1563\
                The commenters are confusing correlation with causation and failing to
                consider the counterfactual case. Higher prices of new vehicles
                certainly did not cause sales to increase since 2012. Sales increased
                over that period, in large part, as a result of economic expansion
                following the great recession.\1564\ The statistical model used in the
                NPRM attempted to isolate the effect of average price on new vehicle
                sales, independent of the overall health of the US economy which plays
                an obviously important role. That model showed a negative relationship
                between sales and price (albeit a modest one), and positive
                relationships with GDP and employment. Even under the most stringent
                alternative in the NPRM, sales increased over time. However, in other
                alternatives, where the same macroeconomic conditions prevailed but
                average new vehicle prices were lower, sales increased relative to the
                baseline. That is the counterfactual case that is relevant for
                regulatory analysis--it attempts to answer the question, ``would sales
                have been even higher if average prices had been lower?''
                ---------------------------------------------------------------------------
                 \1563\ See, e.g. EDF, Appendix B, NHTSA-2018-0067-12108, at 37;
                CARB, Detailed Comments, NHTSA-2018-0067-11873, at 198-204; Aluminum
                Association, Comments, NHTSA-2018-0067-11952, at 19-21; SAFE,
                Comments, NHTSA-2018-0067-11981 at 36; CBD et al., Attachment 1,
                NHTSA-2018-0067-12123, at 20. States and Cities, Detailed Comments,
                NHTSA-2018-0067-11735, at 87-89.
                 \1564\ Table VI-148 below shows a large and statistically
                significant effect of GDP on sales.
                ---------------------------------------------------------------------------
                 As discussed below, identifying the independent contribution of
                price to new vehicle sales is econometrically challenging. In the NPRM,
                the agencies stated that the simultaneous nature of price and sales--
                where transaction prices are higher in periods of higher demand,
                because the market will bear them, and lower in periods of lower
                demand, because the market will not, for an otherwise identical
                vehicle--creates a form of reverse causality. As commenters suggested,
                in recent years sales have increased along with average transaction
                price increases--and transaction price increases will occur when
                regulation forces manufacturers to add content, and their corresponding
                costs, to the vehicles they sell. Thus, it is understandable that some
                commenters could interpret the recent increase in new vehicle sales
                following the recession as evidence that standards (and maybe prices)
                have no impact on new sales. However, that view confuses correlation
                for causation (or lack thereof, in this case).
                 In response to these comments, the agencies have modified their
                approach to modeling the sales impacts of regulatory alternatives. In
                order to isolate the impact of the standards, the agencies have broken
                the sales response module into two discrete components. The first
                captures the effects of broader economic forces such as GDP growth. The
                second measures how changes in vehicle prices influence sales. As
                elaborated in more detail in the following passages, the agencies
                considered alternative approaches and specific changes suggested by
                commenters, but concluded that the comments either lacked enough
                information to implement a change, failed to remedy identified alleged
                weaknesses of the NPRM model, or created new limitations for which
                there were no practical solutions. Furthermore, the two-pronged
                approach addresses many of the concerns raised by commenters better
                than any specific modeling alteration. First, the structural changes to
                the model address many of the econometric concerns raised by
                commenters. Second, by modeling sales in the first step as a function
                of macroeconomic conditions, and then applying an independent own-price
                elasticity to estimate the change in sales across alternatives, the
                agencies are able to more clearly distinguish between demand-side and
                supply-side impacts on prices, the issue that appears to have tripped
                up some of the commenters.
                Comments on the Econometric Model Used in the NPRM
                 Any model of sales response must satisfy two requirements: It must
                be appropriate for use in the CAFE model, and it must be based in both
                sound economic theory and appropriate empirical analysis. The first of
                these requirements implies that forecasts of any variable used in the
                estimation of the econometric model must also be available as a
                forecast throughout the duration of the years covered by the
                simulations (this analysis explicitly simulates compliance through MY
                2050). Some values the model calculates endogenously, making them
                available in future years for sales estimation, but others must be
                known in advance of the simulation. As the CAFE model simulates
                compliance, it accumulates technology costs across the industry and
                over time. By starting with the last known average transaction price
                (associated with MY 2016, in this analysis) and adding accumulated
                regulatory costs to that value, the model is able to represent an
                estimated average selling price in each future model year, assuming
                that manufacturers are able to pass their compliance costs on to buyers
                of new vehicles. Other variables used in the estimation can be entered
                into the model as inputs prior to the start of the compliance
                simulation.
                 The NPRM analysis was based on an econometric model that attempted
                to estimate the price elasticity of aggregate demand for new light-duty
                vehicles based on exogenous factors, intended to represent (1)
                macroeconomic forces that influence demand for new vehicles, and (2)
                average new vehicle price, intended
                [[Page 24597]]
                to represent the impact of regulation. A number of commenters voiced
                opposition to the approach. Some disagreed with the theoretical framing
                of the issue--arguing that the model of sales response should have
                acknowledged the relevance of other vehicle attributes, included
                consumer valuation of fuel savings for new vehicles, based the response
                on something other than price, and considered the effect at a lower
                level of aggregation, rather than average price across the industry.
                 In the NPRM, the agencies relied upon an autoregressive distributed
                lag (ARDL) statistical model to estimate the impact of price
                differences between regulatory alternatives and to produce a time
                series of total new vehicle sales in each year of the analysis. The
                statistical model estimated new vehicle sales per year based on two
                lagged variables of new sales (new sales in the previous period, and
                the period before that), GDP and lagged GDP, and labor force
                participation and lagged labor force participation. The model used
                quarterly data and seasonally adjusted annual rates to increase the
                number of observations over the sample period for which reliable sales
                data existed (1978-2015). The ARDL model used in the NPRM was chosen to
                address sales impacts at a high level of aggregation, namely the total
                new vehicle market (across all vehicle brands and body styles), and to
                resolve the econometric issues associated with the time series data
                related to total new vehicle sales.
                 Stock et al. commented at length on the econometric specification
                of the NPRM sales response model, identifying limitations and
                suggesting alternative approaches.\1565\ In particular, they argued
                that the length of the response to price shocks should dissipate faster
                than the NPRM model allows--an artifact of using quarterly data and
                seasonally adjusted annual rates to estimate the effect and
                implementing it on an annual basis in the CAFE model. The agencies
                agree that this was a flaw in the implementation of the NPRM model.
                While this approach produced the correct units (i.e., annual sales) the
                response to changes in price should have dissipated at a quarterly
                rate, rather than an annual rate. As a result, a single price shock,
                which appears in one year and disappears the next, was projected to
                have a longer impact on sales in future years than was appropriate
                given the specification. The sales response in the final rule corrects
                for this objective error and takes a more conservative approach to
                price shocks.
                ---------------------------------------------------------------------------
                 \1565\ EPA-HQ-OAR-2018-0283 and NHTSA-2018-0067.
                ---------------------------------------------------------------------------
                 Stock et al. commented that ``it is important to estimate the
                dynamic effect on sales of a price increase, that is, the causal effect
                on current and future demand of a price increase'' because ``it allows
                the response to an intervention--here, a one-time price increase or
                sequence of such increases--to evolve over time.'' \1566\ The comment
                suggests that the agencies should include future responses in sales to
                a one-time price increase that exists for a single period and then
                disappears. In our analytical framework, this implies that a price
                difference between any alternative and the baseline that causes a
                difference in sales in that year should also produce a difference in
                sales in the following year (and possibly subsequent years), though of
                smaller magnitude, even if the price difference only exists for a
                single period. The Stock et al. comment illustrates a quickly
                diminishing response to a single price shock. The final rule assumes
                (more conservatively) that each price shock lasts only for a single
                year, and produces no future ``ripple'' effects in the new vehicle
                market in subsequent years. Furthermore, the regulatory alternatives
                considered in this analysis do not produce single period price shocks
                (in the form of price differences between alternatives), but rather
                persistent price differences between alternatives that result from
                continued differences in stringency. The persistent nature of the price
                differences resulting from fuel economy and CO2 regulations
                further reduce the importance of capturing these multi-period effects
                caused by single-period price shocks.
                ---------------------------------------------------------------------------
                 \1566\ Ibid.
                ---------------------------------------------------------------------------
                 Stock et al. also objected to the use of an ARDL model to estimate
                the impact of price on new vehicle sales. In order for the estimation
                of causality to be valid in a time series model, the current price
                movements must be uncorrelated with unobserved demand shocks in the
                past, present, and future; so-called strict exogeneity. The commenters
                argue that the NPRM fails this test because actions taken in the market
                (by both buyers and sellers) can influence the response to price
                changes in the next period. They suggest the use of a vector
                autoregression (VAR) model to address the relationship between past
                demand disturbances and current prices to address the temporal
                exogeneity issues they identify. However, an important caveat is that
                this approach still does not resolve the largest econometric
                challenge--that of contemporaneous endogeneity between price and sales
                (in the same period). To address that challenge, one needs to employ
                instrumental variable methods.
                 The agencies attempted several modifications to the statistical
                model developed for the NPRM based on the Stock et al. comment. The
                agencies reviewed the initial approach and attempted several
                specifications that would explicitly address the temporal endogeneity
                bias identified in the comment. In particular, the agencies addressed
                data limitations that were raised by Stock et al. (and also by EDF),
                who encouraged us to reconsider the quarterly specification and to use
                quality-adjusted price data for new vehicles in order to ensure a more
                consistent definition of the average vehicle over the time series, as
                the ``average vehicle'' has consistently improved in a myriad of ways
                over successive model years. The quarterly price series was
                statistically interpolated in the NPRM to increase the number of
                observations,\1567\ but represented a less-than-ideal solution. The
                interpolating process may have impacted the underlying quarterly data
                generating process, resulting in unreliable, or potentially biased,
                regression results. This issue was remedied by sourcing both vehicle
                sales and price data from IHS Markit, which provides these data at the
                same base frequency (quarterly) and obviates the need for any
                interpolation. In addition, the macroeconomic data used in the model
                specification were also sourced from IHS, which provides consistency
                between historical and forecast data (i.e., forecasts of sales, price,
                personal income, etc., were all based on a consistent set of input
                assumptions and modeling framework during testing).
                ---------------------------------------------------------------------------
                 \1567\ Interpolation is the practice of adding unobserved data
                points based on observed trends to provide more observations to a
                limited data set.
                ---------------------------------------------------------------------------
                 Historical quarterly series for new light vehicle average price and
                total sales are presented in Figure VI-59 below. Due to the lack of
                data availability for business investment in light vehicles, the
                historical series for average vehicle price begins in 1987. Average
                prices were transformed into quality adjusted real terms using the CPI
                for new motor vehicles, and both series were seasonally adjusted.\1568\
                Quality adjusted prices have risen overtime, while total sales have
                remained relatively flat in recent years with the major exception being
                the significant economic downturn of 2008-2009. The difference in these
                trends suggests that the number of vehicles purchased per
                [[Page 24598]]
                household does not necessarily change, or grow, over time, as income
                grows, but rather households adjust the ``amount'' of new vehicle they
                are willing to purchase (i.e., switching from sedan to an SUV).\1569\
                Moreover, while disposable income has steadily increased during this
                period, sales have not seen the same type of upward trend, and instead
                only returned to its pre-recession average of around 17 million annual
                sales.
                ---------------------------------------------------------------------------
                 \1568\ Seasonal adjustment was made using X.12 in EViews.
                 \1569\ Aggregate light duty vehicle sales data does not allow
                for observing the distribution of vehicles being sold, which will
                have an effect on the average price.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.301
                 Even as real disposable income has risen since 2000, and outside of
                the great recession, new vehicle sales have remained relatively steady.
                This, in turn, suggests there are other economic, or behavioral,
                factors beyond disposable income influencing the decision to purchase a
                new vehicle. Given the significant cost to purchase a new vehicle, and
                the long multiyear timeframe over which they are typically financed,
                households' forward-looking view on the health of the economy likely
                plays a role in their willingness to purchase a new vehicle. Put
                differently, households may delay their purchasing decisions if their
                view outlook on the economy sours, regardless of income level. These
                observations are consistent with the framework of the NPRM model, and
                Figure VI-60 presents the consumer sentiment index and total new sales,
                with both series exhibiting similar trends over this period. Some
                commenters advocated that consumer sentiment (also known as consumer
                confidence) should be included in the sales forecast. For example, the
                Aluminum Association indicated that prior sales models have shown
                consumer behavior to be ``highly sensitive to macroeconomic conditions,
                consumer confidence and employment levels.'' While consumer sentiment
                was not included in the NPRM model, it was included in specifications
                that the agencies tested and considered and is a component of the
                forecasting model used in the final rule.\1570\
                ---------------------------------------------------------------------------
                 \1570\ Commenters mentioned consumer confidence as a predictor
                of consumer behavior. For instance, the Aluminum Association
                indicated that prior sales models have shown consumer behavior to be
                ``highly sensitive to macroeconomic conditions, consumer confidence
                and employment levels.'' Comments, NHTSA-2018-0067-11952, at 14.
                ---------------------------------------------------------------------------
                [[Page 24599]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.302
                 All macroeconomic data were sourced from IHS including real
                disposable income, number of US households, and the University of
                Michigan's consumer sentiment index. The summary statistics for all
                series are presented below in Table VI-148.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.303
                 Each series was transformed into natural logarithms and tested for
                stationarity using the modified Dicky-Fuller test.\1571\ Results
                presented in Table VI-149 indicate each variable containing contained a
                unit-root, while being differenced stationary (i.e., integrated of
                order one).
                ---------------------------------------------------------------------------
                 \1571\ Using nonstationary variables would generate unreliable
                estimates of their influence, as prior values of those variables are
                correlated with their future values, and this violates the
                assumption that values variables take on are independent over time.
                ---------------------------------------------------------------------------
                [[Page 24600]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.304
                 Two separate variables lists were then tested for the existence of
                one or more cointegrating relationships, with results from the Johansen
                test presented in Table VI-150.\1572\ In each set of variables, both
                total LDV sales and disposable income were converted to household units
                as a means to control for the growth in US households and the possible
                decision making process of buying/consuming a new unit of LDV. The
                results show that 4 out of the 5 lag length selections for both
                variable sets conclude there being one cointegrating relationship (rank
                I(1)) among them.
                ---------------------------------------------------------------------------
                 \1572\ The number of lag lengths were also tested formally, with
                general consensus between 2 and 6 lags as being optimal. Test
                results are available upon request, however, the final lag length
                selection was determined on the full set of VAR and VECM output that
                includes satisfying time series conditions such as no presence of
                autocorrelation and plausible interpretability of the estimated
                output.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.305
                 Taken together, these tests confirm the need to address the time
                series properties of each variable in any modeling framework. This will
                become especially important when discussing the correct modeling
                approach, as The pre-modeling tests provide evidence against running a
                simple OLS regression or VAR in first differences, because doing so
                would have the potential outcome of excluding important long-run
                information.
                 Furthermore, the endogeneity between vehicle sales and price is
                another element that needs to be considered for model specification.
                The IHS historical series for average price of a new light duty vehicle
                is defined as a
                [[Page 24601]]
                function of business and private residential spending on light vehicles
                divided by total new light vehicle sales; from this identity, the
                average price represents the nominal price per new unit of light duty
                vehicle sold. This definition supports the existence of an endogenous
                relationship between vehicle price and sales that needs to be accounted
                for when developing an econometric estimation of the influence of new
                vehicle price on sales. This is consistent with economic theory,
                whereby vehicle sales and price are simultaneously determined in the
                market, and therefore should be included together when specifying a
                forecasting equation.\1573\ This restriction holds even if nominal
                vehicle price is transformed into a quality adjusted real dollar
                series, as some commenters (EDF, Stock et al) proposed.\1574\
                ---------------------------------------------------------------------------
                 \1573\ Endogeneity results in correlation between an independent
                variable in a regression and the error term leading to biased
                coefficient estimates.
                 \1574\ For reference on how the BLS measures quality adjustments
                in vehicles: https://www.bls.gov/cpi/factsheets/new-vehicles.htm.
                ---------------------------------------------------------------------------
                Models
                 Faced with the simultaneity problem associated with price and
                sales, several specifications were reviewed to determine the best
                method for addressing this issue. An Instrumental Variable (IV) method
                was deemed the most direct approach, with the advantage of preserving
                the initial model's autoregressive distributed lag structure. In order
                to obtain consistent estimates of the price elasticity of demand, a
                suitable instrument that is correlated with average LDV price but
                uncorrelated with the error term is needed in the first stage. A
                suitable instrument must also make economic sense and have a plausible
                causal relationship. In theory, instruments that satisfy all three
                conditions (exogeneity, causality, and non-weak correlation) should
                exist. In practice, however, it is often prohibitively difficult to
                find a viable instrument. Both Stock et al. and CARB suggested
                instrumenting to resolve the endogeneity issue in the NPRM model, but
                neither suggested specific candidates for instrumental variables.
                 For the purposes of modeling vehicle sales, candidate IVs would
                reflect the price of inputs to production that are broad enough, so
                that the underlying behavior of the variable is not deterministic of
                LDV sales. Examples of candidate variables include producer price
                indices (PPIs) of auto or other related manufacturing, cost of capital
                required for production, labor market data, energy costs, technology
                changes, and exogenous shocks to price, production, labor, or policy
                changes.
                 The lack of data availability and quality concerns reduced the
                primary list of candidate IVs to relatable PPIs such as for
                manufacturing and automobile primary products. Even the most
                ``promising'' candidate IVs, however, proved to be poor instruments,
                with counterintuitive signs, lack of statistical significance, and poor
                overall first stage F-statistics (even by relatively lenient weak
                instrument test standards).
                 The lack of reasonable results from the IV approach led to testing
                vector autoregressive (VAR) and vector error correction (VECM) models.
                Relaxing the strict exogeneity assumption needed under an ARDL
                framework is the main advantage of modeling price, sales, and
                macroeconomic variables as a system of equations where the feedback
                from previous period shocks affect both price and sales.\1575\ In
                addition, a VAR or VECM can also adequately handle the time series and
                nonstationary properties discussed above. For both the VAR and VECM, a
                parsimonious specification was preferred with either a three or four
                variable system using the variables discussed above.
                ---------------------------------------------------------------------------
                 \1575\ Strict exogeneity requires there to be past,
                contemporaneous, and future exogeneity between the variables of
                interest.
                ---------------------------------------------------------------------------
                 We first estimated a simple VAR using a Wold causal ordering of
                real disposable income per household, average price of new LDV, and new
                total sales of LDVs per household.\1576\ The alternative specification
                included the consumer sentiment variable in the ordering the consumer
                sentiment variable after income and before price. This ordering assumes
                that households' disposable income (and consumer sentiment) do not
                respond to shocks to auto prices and sales within the same quarter. It
                also assumes that prices are contemporaneously exogenous of sales
                (demand), since the MSRPs are set in advance. Lastly, sales are able to
                respond to unexpected changes in price in the same quarter. The
                alternative ordering of placing sales before average price was deemed
                unrealistic as it would presume sales responding independently to an
                unexpected change in prices.
                ---------------------------------------------------------------------------
                 \1576\ The Wold causal ordering creates a lower triangular
                matrix for our shocks, so by construction these shocks are
                orthogonal to each other to allow for causal inference. This
                recursive or Wold ordering technique should be predetermined and
                based on economic theory as the causal interpretation of the impulse
                responses are dependent on the correct/plausible ordering of
                variables.
                ---------------------------------------------------------------------------
                 In the first specification, all variables were transformed to first
                differences to ensure stationarity, while ignoring any possible long-
                run information (for the moment). A combination of post-estimation
                tests for autocorrelation and stability conditions were considered
                along with impulse response functions to gauge the model performance.
                The preferred model was estimated with five lags, and the impulse
                response functions (IRF) of a 1 percent shock to price on sales for the
                two specifications are presented in Figure VI-61.
                [[Page 24602]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.306
                 Both figures show a similar trend of the response in sales
                oscillating from negative to positive before ultimately returning to
                zero 12 quarters out. The three variable VAR sees a positive response
                in the first few periods, while the four variable VAR manages to dip
                below zero briefly after 4 periods out. This behavior, which by
                definition is short-run due to the differencing of the variables, could
                be representing auto dealerships' attempts to pull sales back to its
                equilibrium level after the price shock pushes sales negative, implying
                some level of over compensation during this process. Nonetheless,
                despite the model showing there is some evidence of an immediate and
                negative price elasticity, the overly simplified VAR model is missing
                key long run information (as identified in the cointegration tests),
                creating some reservations about the results. It is also worth noting
                that the lagged positive response in sales from an unexpected price
                shock is persistent regardless of the lag length selection, and in many
                cases even more pronounced.
                 A number of preliminary conclusions can be drawn from the IRF
                results shown in Figure VI-61. First, at least at this level of
                aggregation, any short-run and immediate effect of a price increase on
                total LDV sales is relatively small in nature. This does not suggest,
                however, that the price elasticity of demand is zero. Instead, what may
                be the case is that when faced with an unexpected change in price,
                consumers will choose to purchase a less expensive car with fewer
                features as opposed to no car at all. In other words, the level of
                aggregation being used, total car sales, removes important variation
                between the type of vehicle being sold and consumer purchasing
                decisions from the data; what is left is a clouded version of the true
                relationship between price and sales. Second, this type of VAR ignores
                and throws out any long run information that may exist, which would
                create omitted variable bias if such a cointegrated relationship
                exists.
                 Based on the conclusions from the Johansen cointegration test, the
                next step involved estimating the system as a VECM. As with the VAR
                models, the VECM employs either a three or four variable system with
                five lag lengths and an unconstrained constant in the model (no trend
                in either the first differenced or cointegrating equations). In each
                model, the cointegrating vector is normalized around sales (i.e., the
                sales' coefficient is set to 1), and the model results indicate strong
                evidence of a cointegrating relationship between the variables.
                 Aside from general agreement on a cointegrating relationship, the
                VECM performance was weak in nearly every specification attempted, with
                implausible magnitudes for the long-run coefficient estimates and
                insignificant short-run dynamics. Moreover, the adjustment coefficient
                for the sales equation is particularly weak and insignificant.\1577\
                The limitations of the VECM could be rooted in the system being
                normalized around sales, which lacks significant variation,
                correlation, or possibly true causation with the other variables.
                ---------------------------------------------------------------------------
                 \1577\ The lack of a statistically significant adjustment
                variable could be an indication of weak exogeneity. In this case
                that would not be plausible given the clear endogeneity between
                price and sales, and is more likely an indication of poor data and
                the absence of reliable modelling approaches.
                ---------------------------------------------------------------------------
                 As with the VAR analysis, a similar focus is placed on the IRFs
                presented in Figure VI-62. Here a one percent shock in price on LDV
                sales shows a similar response between the two specifications, with an
                increase during the first several periods before returning to a
                negative and permanent long-run effect. This response is erroneous in
                two ways: First, the sharp positive response during the first 8 to 10
                quarters defies economic logic as an increase in the price of a normal
                good should not induce an increase in sales. Second, the permanent and
                negative effect is equally as confounding because it rules out the
                ability for dealerships or auto manufacturers to adjust prices or
                supply.\1578\
                ---------------------------------------------------------------------------
                 \1578\ Note that error bounds cannot be generated for VECM IRFs
                using most statistical packages, so determining statistical
                significance is difficult. Given the change from positive to
                negative and the low magnitude of the response, it is quite possible
                that this effect is indistinguishable from zero.
                ---------------------------------------------------------------------------
                 The updated econometric models of light duty vehicle sales
                (described above) thus did not provide clear, significant or robust
                insight into the magnitude of the price elasticity of demand. While the
                VAR model specification points to an immediate short-run negative price
                elasticity of demand (i.e., sales fall in the face of an immediate
                price shock), this relationship is relatively small. In addition, the
                fact that this specification excludes the identified cointegration
                between the variables suggests that it is not robust or unbiased. In
                short, the VECM and IV approaches were unable to provide reasonable and
                meaningful results.
                 These results strongly suggest that the relationship between sales
                and price is
                [[Page 24603]]
                not adequately estimated with the macro-level data used in this
                analysis. Recent peer reviewers of the CAFE model had similar concerns.
                In particular, these data are insufficient to explain the individual
                consumer (micro-) level decision making process of purchasing a new
                LDV. Aggregating the sales response to the national level reduces the
                useful variation in the decision making process to levels unsuitable
                for estimation. Commenters generally agreed with this conclusion.
                 Even assuming a theoretically and econometrically correct model was
                possible, this relationship is impossible to evaluate at the current
                data aggregation level. Future research may focus on constructing an
                aggregate price elasticity of demand from consumer level data utilizing
                discrete choice modeling or something similar. However, constructing
                such models and integrating them into the simulations of the final rule
                are beyond the scope of this analysis.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.307
                 Many commenters suggested that the NPRM model was unable to find a
                statistically significant influence of fuel economy on sales because
                the model was too highly aggregated, as the agencies found with the
                econometric experimentation to estimate a price response. EDF, CARB,
                and CA et al. and Oakland et al. expressed concern that using industry
                averages eliminated the variation needed to detect consumer valuation
                of fuel economy in new vehicle purchases. The agencies noted a similar
                concern in the NPRM, citing the level of aggregation as the most likely
                reason that the average fuel economy of a new vehicle was not a
                statistically significant explanatory variable in the ARDL model. The
                approach for the final rule includes an average value of improved fuel
                economy in the sales response, as commenters suggested it should.
                (a) How Do Car and Light Truck Buyers Value Improved Fuel Economy?
                 Many commenters (CARB, CA et al. and Oakland et al., NRDC, EDF,
                CBD, North Carolina Department of Environmental Quality, IPI, EPA
                Science Advisory Board, Stock et al.) stated that the agencies should
                explicitly consider fuel savings, and the value that consumers ascribe
                to it, in addition to changes in price when estimating the response of
                new vehicle sales to different regulatory alternatives. NRDC stated,
                ``The decision between new vehicle purchase alternatives must consider
                both differential costs and differential benefits. The CAFE model sales
                algorithm considers only differential costs and is, therefore,
                flawed.'' \1579\ The agencies agree that the degree to which new
                vehicle buyers value improvements in fuel economy is an important
                consideration when estimating the response of new vehicle sales to
                potential standards. The effect of vehicle prices on sales is difficult
                to detect at the aggregate level because price movements are correlated
                with the current strength of the economy, which can appear as a
                positive price elasticity when modeling sales, and there are various
                technical econometric difficulties in identifying the effect of price
                on sales (simultaneity, cointegration, etc., addressed above). The
                sales response model in the final rule accounts for fuel savings
                realized by buyers of new vehicles.
                ---------------------------------------------------------------------------
                 \1579\ NRDC, Attachment 3, NHTSA-2018-0067-11723, at 4.
                ---------------------------------------------------------------------------
                 Some commenters and EPA's Science Advisory Board noted that the
                sales response equation omitted any value of fuel savings to new
                vehicle buyers, while other elements of the analysis--notably the
                technology application algorithm--assumed that buyers would demand fuel
                economy technologies that ``pay back'' within the first 2.5 years of
                ownership (as a result of avoided fuel costs), and manufacturers would
                supply fuel economy at those levels even in the absence of standards.
                This observation was made in comments by CARB, CBD, and IPI--the last
                of which stated that 2.5 year payback assumption ``clashes directly
                with the contradictory assumption that the agencies rely on in the
                model's sales module, where they implicitly assume that customers
                entirely disregard fuel efficiency in their purchasing decisions.''
                \1580\ The agencies agree that this represented an internal
                inconsistency. The sales model used to analyze the final rule includes
                the estimated value of fuel savings to vehicle buyers, and is
                consistent with other assumptions throughout the analysis about the
                ``pay back'' period.
                ---------------------------------------------------------------------------
                 \1580\ IPI, Appendix, NHTSA-2018-0067-12213, at 16.
                ---------------------------------------------------------------------------
                 How potential buyers value improvements in the fuel economy of new
                cars and light trucks is an important issue in assessing the benefits
                and costs of government regulation. If buyers fully value the savings
                in fuel costs that result from higher fuel
                [[Page 24604]]
                economy, manufacturers will presumably supply any improvements that
                buyers demand, and vehicle prices will fully reflect future fuel cost
                savings consumers would realize from owning--and potentially re-
                selling--more fuel-efficient models. If consumers internalize fuel
                savings this case, more stringent fuel economy standards will impose
                net costs on vehicle owners and can only result in social benefits
                through correcting externalities, because consumers would already fully
                incorporate private savings into their purchase decisions, as discussed
                further below. If instead consumers systematically undervalue some
                market failure such as an information asymmetry leads to an
                underinvestment in fuel-saving technology, the cost savings generated
                by improvements in fuel economy when choosing among competing models,
                more stringent fuel economy standards will also lead manufacturers to
                adopt improvements in fuel economy that buyers might not choose despite
                the cost savings they offer and improve consumer welfare.
                 The potential for car buyers voluntarily to forego improvements in
                fuel economy that offer savings exceeding their initial costs is one
                example of what is often termed the ``energy-efficiency gap.'' This
                appearance of such a gap, between the level of energy efficiency that
                would minimize consumers' overall expenses and what they actually
                purchase, is typically based on engineering calculations that compare
                the initial cost for providing higher energy efficiency to the
                discounted present value of the resulting savings in future energy
                costs.
                 There has long been an active debate about why such a gap might
                arise and whether it actually exists. Economic theory predicts that
                individuals will purchase more energy-efficient products only if the
                savings in future energy costs they offer promise to offset their
                higher initial costs. However, the additional up-front cost of a more
                energy-efficient product includes more than just the cost of the
                technology necessary to improve its efficiency; because consumers have
                a scarcity of resources, it also includes the opportunity cost of any
                other desirable features that consumers give up when they choose the
                more efficient alternative. In the context of vehicles, whether the
                expected fuel savings outweigh the opportunity cost of purchasing a
                model offering higher fuel economy will depend, among other things, on
                how much its buyer expects to drive, his or her expectations about
                future fuel prices, the discount rate he or she uses to value future
                expenses, the expected effect on resale value, and whether more
                efficient models offer equivalent attributes such as performance,
                carrying capacity, reliability, quality, or other characteristics.
                 Published literature has offered little consensus about consumers'
                willingness-to-pay for greater fuel economy, and whether it implies
                over- under- or full-valuation of the expected discounted fuel savings
                from purchasing a model with higher fuel economy. Most studies have
                relied on car buyers' purchasing behavior to estimate their
                willingness-to-pay for future fuel savings; a typical approach has been
                to use ``discrete choice'' models that relate individual buyers'
                choices among competing vehicles to their purchase prices, fuel
                economy, and other attributes (such as performance, carrying capacity,
                and reliability), and to infer buyers' valuation of higher fuel economy
                from the relative importance of purchase prices and fuel economy.\1581\
                Empirical estimates using this approach span a wide range, extending
                from substantial undervaluation of fuel savings to significant
                overvaluation, thus making it difficult to draw solid conclusions about
                the influence of fuel economy on vehicle buyers' choices.\1582\ Because
                a vehicle's price is often correlated with its other attributes (both
                measured and unobserved), analysts have often used instrumental
                variables or other approaches to address endogeneity and other
                resulting concerns.\1583\
                ---------------------------------------------------------------------------
                 \1581\ In a typical vehicle choice model, the ratio of estimated
                coefficients on fuel economy--or more commonly, fuel cost per mile
                driven--and purchase price is used to infer the dollar value buyers
                attach to slightly higher fuel economy.
                 \1582\ See Helfand & Wolverton (2011) and Green (2010) for
                detailed reviews of these cross-sectional studies.
                 \1583\ See, e.g., Barry, et al. (1995).
                ---------------------------------------------------------------------------
                 Despite these efforts, more recent research has criticized these
                cross-sectional studies; some have questioned the effectiveness of the
                instruments they use,\1584\ while others have observed that
                coefficients estimated using non-linear statistical methods can be
                sensitive to the optimization algorithm and starting values.\1585\
                Collinearity (i.e., high correlations) among vehicle attributes--most
                notably among fuel economy, performance or power, and vehicle size--and
                between vehicles' measured and unobserved features also raises
                questions about the reliability and interpretation of coefficients that
                may conflate the value of fuel economy with other attributes (Sallee,
                et al., 2016; Busse, et al., 2013; Allcott & Wozny, 2014; Allcott &
                Greenstone, 2012; Helfand & Wolverton, 2011).
                ---------------------------------------------------------------------------
                 \1584\ See Allcott & Greenstone (2012).
                 \1585\ See Knittel & Metaxoglou (2014).
                ---------------------------------------------------------------------------
                 In an effort to overcome shortcomings of past analyses, three
                studies published fairly recently rely on panel data from sales of
                individual vehicle models to improve their reliability in identifying
                the association between vehicles' prices and their fuel economy
                (Sallee, et al. 2016; Allcott & Wozny, 2014; Busse, et al., 2013).
                Although they differ in certain details, each of these analyses relates
                changes over time in individual models' selling prices to fluctuations
                in fuel prices, differences in their fuel economy, and increases in
                their age and accumulated use, which affects their expected remaining
                life, and thus their market value. Because a vehicle's future fuel
                costs are a function of both its fuel economy and expected gasoline
                prices, changes in fuel prices have different effects on the market
                values of vehicles with different fuel economy; comparing these effects
                over time and among vehicle models reveals the fraction of changes in
                fuel costs that is reflected in changes in their selling prices
                (Allcott & Wozny, 2014). Using very large samples of sales enables
                these studies to define vehicle models at an extremely disaggregated
                level, which enables their authors to isolate differences in their fuel
                economy from the many other attributes, including those that are
                difficult to observe or measure, that affect their sale prices.\1586\
                ---------------------------------------------------------------------------
                 \1586\ These studies rely on individual vehicle transaction data
                from dealer sales and wholesale auctions, which includes actual sale
                prices and allows their authors to define vehicle models at a highly
                disaggregated level. For instance, Allcott & Wozny (2014)
                differentiate vehicles by manufacturer, model or nameplate, trim
                level, body type, fuel economy, engine displacement, number of
                cylinders, and ``generation'' (a group of successive model years
                during which a model's design remains largely unchanged). All three
                studies include transactions only through mid-2008 to limit the
                effect of the recession on vehicle prices. To ensure that the
                vehicle choice set consists of true substitutes, Allcott & Wozny
                (2014) define the choice set as all gasoline-fueled light-duty cars,
                trucks, SUVs, and minivans that are less than 25 years old (i.e.,
                they exclude vehicles where the substitution elasticity is expected
                to be small). Sallee et al. (2016) exclude diesels, hybrids, and
                used vehicles with less than 10,000 or more than 100,000 miles.
                ---------------------------------------------------------------------------
                 These studies point to a somewhat narrower range of estimates than
                suggested by previous cross-sectional studies; more importantly, they
                consistently suggest that buyers value a large proportion--and perhaps
                even all--of the future savings that models with higher fuel economy
                offer.\1587\
                [[Page 24605]]
                Because they rely on estimates of fuel costs over vehicles' expected
                remaining lifetimes, these studies' estimates of how buyers value fuel
                economy are sensitive to the strategies they use to isolate differences
                among individual models' fuel economy, as well as to their assumptions
                about buyers' discount rates and gasoline price expectations, among
                others. Since Anderson et al. (2013) found evidence that consumers
                expect future gasoline prices to resemble current prices, the agencies
                use this assumption to compare the findings of the three studies and
                examine how their findings vary with the discount rates buyers apply to
                future fuel savings.\1588\
                ---------------------------------------------------------------------------
                 \1587\ Killian & Sims (2006) and Sawhill (2008) rely on similar
                longitudinal approaches to examine consumer valuation of fuel
                economy except that they use average values or list prices instead
                of actual transaction prices. Since these studies remain
                unpublished, their empirical results are subject to change, and they
                are excluded from this discussion.
                 \1588\ Each of the studies makes slightly different assumptions
                about appropriate discount rates. Sallee et al. (2016) use five
                percent in their base specification, while Allcott & Wozny (2014)
                rely on six percent. As some authors note, a five to six percent
                discount rate is consistent with current interest rates on car
                loans, but they also acknowledge that borrowing rates could be
                higher in some cases, which could be used to justify higher discount
                rates. Rather than assuming a specific discount rate, Busse et al.
                (2013) directly estimate implicit discount rates at which future
                fuel costs would be fully internalized; they find discount rates of
                six to 21 percent for used cars and one to 13 percent for new cars
                at assumed demand elasticities ranging from -2 to -3. Their
                estimates can be translated into the percent of fuel costs
                internalized by consumers, assuming a particular discount rate. To
                make these results more directly comparable to the other two
                studies, we assume a range of discount rates and uses the authors'
                spreadsheet tool to translate their results into the percent of fuel
                costs internalized into the purchase price at each rate. Because
                Busse et al. (2013) estimate the effects of future fuel costs on
                vehicle prices separately by fuel economy quartile, these results
                depend on which quartiles of the fuel economy distribution are
                compared; our summary shows results using the full range of quartile
                comparisons.
                ---------------------------------------------------------------------------
                 As Table VI-148 indicates, Allcott & Wozny (2014) found that
                consumers incorporate 55% percent of future fuel costs into vehicle
                purchase decisions at a six percent discount rate, when their
                expectations for future gasoline prices are assumed to reflect
                prevailing prices at the time of their purchases. With the same
                expectation about future fuel prices, the authors report that consumers
                would fully value fuel costs only if they apply discount rates of 24
                percent or higher. However, these authors' estimates are closer to full
                valuation when using gasoline price forecasts that mirror oil futures
                markets, because the petroleum market expected prices to fall during
                this period (this outlook reduces the discounted value of a vehicle's
                expected remaining lifetime fuel costs). With this expectation, Allcott
                & Wozny (2014) find that buyers value 76 percent of future cost savings
                (discounted at six percent) from choosing a model that offers higher
                fuel economy, and that a discount rate of 15 percent would imply that
                they fully value future cost savings. Sallee et al. (2016) begin with
                the perspective that buyers fully internalize future fuel costs into
                vehicles' purchase prices and cannot reliably reject that hypothesis;
                their base specification suggests that changes in vehicle prices
                incorporate slightly more than 100 percent of changes in future fuel
                costs. For discount rates of five to six percent, the Busse et al.
                (2013) results imply that vehicle prices reflect 60 to 100 percent of
                future fuel costs. As Table VI-151 suggests, higher private discount
                rates move all of the estimates closer to full valuation or to over-
                valuation, while lower discount rates imply less complete valuation in
                all three studies.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.308
                 The studies also explore the sensitivity of the results to other
                parameters that could influence their results. Busse et al. (2013) and
                Allcott & Wozny (2014) find that relying on data that suggest lower
                annual vehicle use or survival probabilities, which imply that vehicles
                will not last as long, moves their estimates closer to full valuation,
                an unsurprising result because both reduce the changes in expected
                future fuel costs caused by fuel
                [[Page 24606]]
                price fluctuations. Allcott & Wozny's (2014) base results rely on an
                instrumental variables estimator that groups miles-per-gallon (MPG)
                into two quantiles to mitigate potential attenuation bias due to
                measurement error in fuel economy, but they find that greater
                disaggregation of the MPG groups implies greater undervaluation (for
                example, it reduces the 55 percent estimated reported in Table VI-148
                to 49 percent). Busse et al. (2013) allow gasoline prices to vary
                across local markets in their main specification; using national
                average gasoline prices, an approach more directly comparable to the
                other studies, results in estimates that are closer to or above full
                valuation. Sallee et al. (2016) find modest undervaluation by vehicle
                fleet operators or manufacturers making large-scale purchases, compared
                to retail dealer sales (i.e., 70 to 86 percent).
                 Since they rely predominantly on changes in vehicles' prices
                between repeat sales, most of the valuation estimates reported in these
                studies apply most directly to buyers of used vehicles. Only Busse et
                al. (2013) examine new vehicle sales; they find that consumers value
                between 75 to 133 percent of future fuel costs for new vehicles, a
                higher range than they estimate for used vehicles. Allcott & Wozny
                (2014) examine how their estimates vary by vehicle age and find that
                fluctuations in purchase prices of younger vehicles imply that buyers
                whose fuel price expectations mirror the petroleum futures market value
                a higher fraction of future fuel costs: 93 percent for one- to three-
                year-old vehicles, compared to their estimate of 76 percent for all
                used vehicles assuming the same price expectation.\1589\
                ---------------------------------------------------------------------------
                 \1589\ Allcott & Wozny (2014) and Sallee, et al. (2016) also
                find that future fuel costs for older vehicles are substantially
                undervalued (26-30%). The pattern of Allcott and Wozny's results for
                different vehicle ages is similar when they use retail transaction
                prices (adjusted for customer cash rebates and trade-in values)
                instead of wholesale auction prices, although the degree of
                valuation falls substantially in all age cohorts with the smaller,
                retail price based sample.
                ---------------------------------------------------------------------------
                 Accounting for differences in their data and estimation procedures,
                the three studies described here suggest that car buyers who use
                discount rates of five to six percent value at least half--and perhaps
                all--of the savings in future fuel costs they expect from choosing
                models that offer higher fuel economy. Perhaps more important in
                assessing the case for regulating fuel economy, one study (Busse et
                al., 2013) suggests that buyers of new cars and light trucks value
                three-quarters or more of the savings in future fuel costs they
                anticipate from purchasing higher-mpg models, although this result is
                based on more limited information.
                 In contrast, previous regulatory analyses of fuel economy standards
                implicitly assumed that buyers undervalue even more of the benefits
                they would experience from purchasing models with higher fuel economy,
                so that, without increases in fuel economy standards, little
                improvement would occur, and the entire value of fuel savings from
                raising CAFE standards represented private benefits to car and light
                truck buyers themselves. For instance, in the EPA analysis of the 2017-
                2025 model year CO2 standards, fuel savings alone added up
                to $475 billion (at three percent discount rate) over the lifetime of
                the vehicles, far outweighing the compliance costs: $150 billion). The
                assertion that buyers were unwilling to take voluntary advantage of
                this opportunity implies that collectively, they must have valued less
                than a third ($150 billion/$475 billion = 32 percent) of the fuel
                savings that would have resulted from those standards. In fact, those
                earlier analyses assumed that new car and light truck buyers attach
                relatively little value to higher fuel economy, since their baseline
                scenarios assumed that fuel economy levels would not increase in the
                absence of progressively tighter standards, despite increasing fuel
                prices. The evidence reviewed here makes that perspective extremely
                difficult to justify and would call into question any analysis that
                claims to show large private net benefits for vehicle buyers
                attributable to increases in fuel economy standards.
                 What analysts assume about consumers' vehicle purchasing behavior,
                particularly about potential buyers' perspectives on the value of
                increased fuel economy, clearly matters a great deal in the context of
                benefit-cost analysis for fuel economy regulation. In light of this
                recent evidence on this question, warrants a more nuanced approach that
                is more nuanced than merely assuming that buyers drastically undervalue
                benefits from higher fuel economy, (and that, as a consequence, these
                benefits are unlikely to be realized without stringent fuel economy
                standards,) seems warranted. One possible approach would be to use a
                baseline scenario where fuel economy levels of new cars and light
                trucks reflected full (or nearly so) valuation of fuel savings by
                potential buyers in order to reveal whether setting fuel economy
                standards above market-determined levels could produce net social
                benefits. Another might be to assume that, unlike in the agencies'
                previous analyses, where buyers were assumed to greatly to undervalue
                higher fuel economy under the baseline but to value it fully under the
                proposed standards, buyers value improved fuel economy identically
                under both the baseline scenario and with stricter CAFE standards in
                place.
                 The agencies requested comment on the consumer valuation of fuel
                economy and its use in the NPRM analysis. CBD and the North Carolina
                Department of Environmental Quality took issue with the agencies'
                characterization of the literature on the value of fuel economy, citing
                EPA's previous determination that the estimates in the literature
                represented too large a range, and the degree of uncertainty made
                including a value of fuel economy challenging. This final rule analysis
                accounts for the value of fuel economy in several places, though it
                uses a more conservative value than is suggested by the literature
                summarized above. Manufacturers have consistently told the agencies
                that new vehicle buyers will pay for about 2 or 3 years' worth of fuel
                savings before the price increase associated with providing those
                improvements begins to impact affect sales. The agencies have assumed
                the same valuation, 2.5 years, in all components of the analysis that
                reflect consumer decisions regarding vehicle purchases and
                retirements.\1590\ This analysis explicitly assumes that: (1) Consumers
                are willing to pay for fuel economy improvements that pay back within
                the first 2.5 years of vehicle ownership (at average usage rates); (2)
                manufacturers know this and will provide these improvements even in the
                absence of regulatory pressure; (3) potential buyers weigh these
                savings against increases in new vehicle prices when deciding to retire
                a vehicle; and (4) the amount of technology for which buyers will pay
                rises (or falls) with rising (or falling) fuel prices.\1591\ Excluding
                the value of fuel economy entirely from these calculations does not
                remove it from the analysis; it merely imposes an implausibly low value
                on the desired payback period of new
                [[Page 24607]]
                vehicle buyers and manufacturers--regardless of fuel prices or
                technology costs. And while the agencies acknowledge the uncertainty
                around the estimates in the literature, zero is far removed from the
                lower bounds of any study.
                ---------------------------------------------------------------------------
                 \1590\ When accounting for social benefits and costs associated
                with an alternative, the full lifetime value of fuel savings is
                included.
                 \1591\ NADA, the Alliance of Automobile Manufacturers, and
                American Fuel and Petrochemical Manufacturers argued that CAFE/
                CO2 standards have already reached the point where the
                price increases necessary to recoup manufacturers' increased costs
                for providing further increases in fuel economy outweigh the value
                of fuel savings, and requiring further increases in fuel economy
                will reduce new vehicle sales. The sales response in the final rule
                recognizes and incorporates the effect of fuel prices and fuel
                economy on new vehicle purchases. See NADA, NHTSA-2018-0067-12064,
                at 11; Auto Alliance, Full Comment Set, NHTSA-2018-0067-12073 at
                163-64; AMFP, Comments, NHTSA-2018-0067-12078-29,at 3.
                ---------------------------------------------------------------------------
                 CARB asserted that the various market failures suggested by the
                agencies in past rules (lack of information about fuel savings from
                higher MPG, inability to calculate cost savings from higher MPG, loss
                aversion, first-mover disadvantage), together with advertising that
                only emphasizes fuel economy during periods of high fuel prices, leads
                buyers to undervalue fuel economy.\1592\ In contrast, CARB (and
                others--such as SCAQMD, Alliance to Save Energy, Save EPA, AAA,
                Environmental group coalition, Consumers Union, EDF, and IPI) argues
                elsewhere that new vehicle buyers do value fuel economy highly, and
                nearly fully once fuel prices return to ``normal'' levels.\1593\ The
                agencies' payback period assumption, and the matching adjustment it
                makes to changes in new car prices to account for accompanying changes
                in fuel economy, recognizes that on average potential car buyers value
                a significant share of lifetime cost savings resulting from higher fuel
                economy. The agencies considered longer payback periods along the lines
                suggested by Consumer Federation of America (CFA),\1594\ but chose 2.5
                years as a conservative approach. Our assumption is consistent with
                survey evidence cited by the commenters, but at odds with their
                assertions that this program is necessary to save buyers from their own
                limited ability to make decisions in their best interest.
                ---------------------------------------------------------------------------
                 \1592\ See CARB, Detailed Comments, NHTSA-2018-0067-11873 at
                212-16.
                 \1593\ E.g. id. at 190-91. See also, id. at 188-89. See also,
                SCAQMD, Supplemental comments, NHTSA-2018-0067-11813, at 4-5;
                Alliance to Save Energy, Comment, NHTSA-2018-0067-11837, at 2; Save
                EPA, Comments, NHTSA-2018-0067-11930, at 6; AAA, Comments, NHTSA-
                2018-0067-11979, at 2-3; Environmental group coalition, Appendix A,
                NHTSA-2018-0067-12000, at 54-56; Consumers Union, Attachment A,
                NHTSA-2018-0067-12068, 27-29; EDF, Appendix B, NHTSA-2018-0067-
                12108, at 84-86; and IPI, Appendix, NHTSA-2018-0067-12213, at 40-47.
                 \1594\ CFA, Comments, NHTSA-2018-0067-12005, at 12.
                ---------------------------------------------------------------------------
                 More recently, the agencies have justified stricter CAFE and
                CO2 emissions standards by asserting that buyers do not take
                advantage of opportunities to improve their own well-being, by
                purchasing models whose higher fuel economy would more than repay their
                higher initial purchase prices via future savings in fuel costs. This
                newer rationale is fundamentally different from asserting that some
                externality--whereby buyers' choices cause economic harm to others--
                exists to justify regulating fuel economy or CO2 emissions,
                or adopting more demanding regulations. EPA and NHTSA have previously
                labeled this behavior an example of the ``energy paradox,'' whereby
                consumers voluntarily forego investments that conserve energy even when
                those initial outlays appear likely to repay themselves--in the form of
                savings in energy costs--over the relatively near term.\1595\
                ---------------------------------------------------------------------------
                 \1595\ See, e.g., EPA Regulatory Impact Analysis: Final
                Rulemaking for 2017-2025 Light-Duty Vehicle Greenhouse Gas Emission
                Standards and Corporate Average Fuel Economy Standards, available at
                https://nepis.epa.gov/Exe/ZyPDF.cgi/P100EZI1.PDF?Dockey=P100EZI1.PDF.
                ---------------------------------------------------------------------------
                 However, recent research cast doubt on whether such an energy
                paradox exists in the case of fuel economy--that is, on whether buyers
                of new vehicles inadequately consider the value of future savings in
                fuel costs they would experience from purchasing models that feature
                higher fuel economy--and about how extensive it might be. Several
                recent studies have estimated the fraction of appropriately discounted
                lifetime fuel savings offered by models featuring higher fuel economy
                that car shoppers appear to value or willing to pay for. These
                estimates are typically drawn from one of three sources--(1) buyers'
                choices among competing models with different purchase prices, fuel
                economy levels, and other features; (2) statistically ``decomposing''
                vehicle prices into the values buyers attach to their individual
                features, one of which is fuel economy; or (3) analyzing how selling
                prices for vehicles with different fuel economy levels respond to
                variation in fuel prices and the changes it causes in their lifetime
                fuel costs.
                 The estimates these studies report may partly reflect variation
                among buyers' preferences for different vehicle features (such as fuel
                economy, but also size or utility), the financial constraints they
                face, how much they drive, or their expectations about future fuel
                prices, so they should be interpreted cautiously. However, the most
                careful recent studies suggest that on average buyers appear to
                undervalue the savings from higher fuel economy at most modestly, and
                perhaps not at all, after accounting for the influence of vehicles'
                other attributes on prices and purchasing decisions.\1596\ This
                research suggests that the energy paradox, sometimes described as
                buyers' ``myopia'' in assessing the value of future fuel savings, is a
                much weaker rationale for regulating fuel economy than the agencies had
                previously asserted.
                ---------------------------------------------------------------------------
                 \1596\ For a review of these recent studies, see Table VI-120--
                Percent of Future Fuels Costs Internalized in Used Vehicle Purchase
                Price using Current Gasoline Prices to Reflect Expectations (for
                Base Case Assumptions).
                ---------------------------------------------------------------------------
                 IPI commented that the agencies' obligation to consider market
                failures in setting standards derives not just from Executive Order
                12,866 but also from the agencies' respective statutes, and argued that
                the agencies had defined market failures too narrowly in their
                proposal.\1597\ Specifically, IPI stated that NHTSA's task under EPCA
                is ``not so restricted to only protecting consumers from gas price
                spikes,'' and argued that NHTSA must also consider ``externalities
                relating to energy security, national security, positional goods,
                global climate change, and air and water pollution associated with fuel
                production and consumption; asymmetric information, attention costs,
                and other information failures; internalities, including myopia; and
                various supply-side market failures, including first-mover
                disadvantage.'' \1598\
                ---------------------------------------------------------------------------
                 \1597\ IPI, Appendix, NHTSA-2018-0067-12213, at 9-10.
                 \1598\ Id.
                ---------------------------------------------------------------------------
                 For EPA's task under the CAA, IPI stated that, although while EPA
                must ``protect the planet from unchecked climate change, [it] must not
                ignore other related market failures that cause harm to public health
                and welfare, including the issues and market failures [as described for
                NHTSA above].'' \1599\ IPI argued that the proposal was arbitrary and
                capricious for not ``consider[ing] important aspects of the problem set
                before the agencies by Congress,'' and also for not considering the
                market failures discussed in the 2012 final rule.\1600\ CBD, et al.,
                asserted similarly that the agencies' respective statutes require their
                actions to be more technology-forcing than what markets would otherwise
                achieve, in effect asserting that innovations in technology confer
                external benefits that vehicle manufacturers or buyers do not fully
                consider.\1601\
                ---------------------------------------------------------------------------
                 \1599\ Id.
                 \1600\ Id.
                 \1601\ CBD, et al., NHTSA-2018-0067-12057, at 2 and 9.
                ---------------------------------------------------------------------------
                 With regard to the specific market failures CAFE and CO2
                standards could potentially address, Global Automakers suggested that
                climate effects are indeed an externality that more stringent standards
                can address,\1602\ while CFA stated that regulating fuel economy and
                CO2 emissions can address an extensive catalog of market
                failures, including externalities, marketing, availability of
                [[Page 24608]]
                fuel-efficient models, transaction cost friction, information
                asymmetry, behavioral issues, and access to capital, among
                others.\1603\ CFA asserted that advances in economic theory had heavily
                criticized the neoclassical model, and that ``a great deal of empirical
                evidence supports [that the] standards are seen as an important and, in
                many ways, preferred policy approach.'' \1604\ On this basis, CFA
                stated that attribute-based standards that ``are set at a moderately
                aggressive level'' and are ``consistent with the rate of improvement
                that the auto industry achieved in the first decade of the fuel economy
                standard setting program,'' among other things, would address the
                market failure.\1605\
                ---------------------------------------------------------------------------
                 \1602\ Global Automakers, Attachment A, NHTSA-2018-0067-12032,
                at A-22.
                 \1603\ CFA, Comments, NHTSA-2018-0067-12005, at 61-64.
                 \1604\ Id. at 63.
                 \1605\ Id. at 64.
                ---------------------------------------------------------------------------
                 IPI argued that regulation of fuel economy (presumably also
                CO2 emissions) is necessary because ``many vehicle
                attributes, like horsepower and size, are positional goods--that is,
                they confer status on buyers of cars and light truck models that
                feature them prominently, so regulation of fuel economy can help
                correct the positional externality.'' \1606\ IPI also noted the
                externality of health effects associated with refueling. IPI cited
                Alcott and Sunstein (2015) to argue, like CFA, that fuel economy
                standards can correct market failures like informational failure,
                myopia, supply-side failures, positional externalities, etc., and by
                doing so, can provide net private welfare gains--that is, improve the
                utility of vehicle buyers themselves, not just that of other households
                or businesses.\1607\
                ---------------------------------------------------------------------------
                 \1606\ IPI, Appendix, NHTSA-2018-0067-12213, at 33.
                 \1607\ Id. at 34. Note, however, that the reference cited does
                not address the question of whether fuel economy standards can be
                effective in correcting those market failures. Instead, it explores
                the circumstances under which fuel economy standards can improve
                welfare when vehicle buyers undervalue savings in fuel costs from
                purchasing more fuel-efficient models. See generally, Allcott, Hunt,
                and Cass R. Sunstein, ``Regulating Internalities,'' Working Paper
                20087, National Bureau of Economic Research, May 2015, available at
                https://www.nber.org/papers/w21187.pdf.
                ---------------------------------------------------------------------------
                 EDF and CARB both asserted that an energy paradox exists in the
                case of fuel economy, with EDF arguing (like CFA) that information
                asymmetry--that is, unequal access of vehicle manufacturers and
                potential buyers to information about the cost savings likely to result
                from owning higher-mpg models--coupled with limited availability of
                fuel-efficient models, leads consumers to purchase vehicles with lower
                fuel economy than they otherwise would.\1608\ CARB simply stated that
                the NPRM analysis did not account for the energy paradox.\1609\
                ---------------------------------------------------------------------------
                 \1608\ EDF, Appendix B, NHTSA-2018-0067-12108, at 88-89.
                 \1609\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 188-
                89.
                ---------------------------------------------------------------------------
                 The agencies agree with these commenters that the market failures
                CAFE and CO2 standards can help address are likely to exist,
                but note that little of the behavior in the broad catalog identified by
                commenters actually represents market failures, and instead simply
                reflects consumers' preferences for features other than fuel economy.
                Even in the few cases of potential market failures that commenters
                identify related to the hypothetical energy paradox, the agencies
                question whether more stringent CAFE and CO2 standards are
                necessary to address the phenomena, or are even likely to be effective
                in doing so. In the agencies' view, neither the logical arguments nor
                the limited empirical evidence that commenters presented convincingly
                demonstrate the capacity of more stringent CAFE and CO2
                standards to resolve, or even mitigate, most of the various phenomena
                they describe as market failures.
                 For example, the idea that regulating fuel economy and
                CO2 emissions can mitigate the consequences of inadequate
                access to information by placing decisions that depend on access to
                complete information in the hands of regulators rather than buyers has
                superficial appeal. Yet commenters do not establish that such a drastic
                step is necessary to overcome any inadequacy of information, or that
                requiring manufacturers to supply higher fuel economy will be more
                effective than less intrusive approaches such as expanding the range of
                information available to buyers. As OMB Circular A-4 notes, ``Because
                information, like other goods, is costly to produce and disseminate,
                your evaluation will need to do more than demonstrate the possible
                existence of incomplete or asymmetric information.'' \1610\
                ---------------------------------------------------------------------------
                 \1610\ Circular A-4, at 5.
                ---------------------------------------------------------------------------
                 In the few cases where commenters cited empirical evidence to
                support their arguments that stricter fuel economy and CO2
                regulations are an appropriate response to market failures, that
                evidence is limited and unpersuasive. As one illustration, the frequent
                assertion that buyers' widespread aversion to the prospect of financial
                losses makes them hesitant to purchase higher-mpg models appears to be
                traceable to findings from classroom experiments on small numbers of
                university students, rather than to large-scale empirical evidence
                drawn from buyers' observed behavior.\1611\ Commenters' repeated
                emphasis on loss aversion as a critical source of buyers' unwillingness
                to choose levels of fuel economy that appear to be in their own
                financial interest also ignores recent research questioning whether
                loss aversion is a plausible motivation for such systematic or
                universal behavior by consumers.\1612\
                ---------------------------------------------------------------------------
                 \1611\ CFA, Comments, NHTSA-2018-0067-12005, at 16 et seq;
                Consumers Union, Attachment 4, NHTSA-2018-0067-12068, at 12;
                Attachment 3, NHTSA-2018-0067-11741, at 5-6, CARB at 214, and States
                at 87 each assert that loss aversion is an important source of car
                buyers' hesitance to purchase higher-mpg models, variously citing
                Greene, David L., John German, and Mark A. Delucchi, ``Fuel Economy:
                The Case for Market Failure,'' Reducing Climate Impacts in the
                Transportation Sector, Springerin James S. Cannon and Daniel
                Sperling, eds., Springer, 2009, at pp. 181-205; (2009); Greene,
                David L. (2010). How consumers value fuel economy: A literature
                review (No. EPA-420-R-10-008); Greene, David L., ``Uncertainty, Loss
                Aversion and Markets for Energy Efficiency,'' Energy Economics, vol.
                33, at pp. 608-616, (2011) and Greene, David L., ``Consumers'
                Willingness to Pay for Fuel Economy: Implications for Sales of New
                Vehicles and Scrappage of Used Vehicles,'' attachment to comments by
                CARB, Oct. 10, 2018. However, none of these sources presents
                empirical evidence on how the frequency of actual common loss
                aversion actually is among real world vehicle buyers, instead simply
                asserting (or implicitly assuming) that loss aversion it is likely
                to be widespread. Further, their (identical) estimates of the degree
                of loss aversion are difficult to trace, and appear to be drawn from
                classroom exercises administered to limited numbers of university
                students, not from empirical research involving real world vehicle
                buyers. One source cited for their repeated assertion that losses of
                a given dollar amount are valued twice as highly as gains of the
                same amount is Gal, David, ``A psychological law of inertia and the
                illusion of loss aversion,'' Judgment and Decision Making, Vol. 1,
                No. 1, at pp. 23-32 (July 2006,), pp. 23-32, but this reference does
                not report such a value. Another source repeatedly cited by Greene
                and co-authors, Benartzi, Shlomo, and Richard H. Thaler, ``Myopic
                Loss Aversion and the Equity Premium Puzzle,'' Quarterly Journal of
                Economics, Vol. 110, No. 1, at pp. 73-92 (February 1995), pp. 73-92,
                does report this value (at p. 74), although only in passing, and
                cites other references as its original source. The original sources
                of the claim that losses are values twice as highly as equivalent
                gains appear to be Kahneman, Daniel, Jack L. Knetsch, and Richard H.
                Thaler, ``Experimental Tests of the Endowment Effect and the Coase
                Theorem,'' Journal of Political Economy, Vol. 98, No. 6, pp. 1325-
                48. (Dec., 1990) (pp. 1325-1348, specifically Section II), pp. 1329-
                1336; and Tversky, Amos, and Daniel Kahneman, ``Loss Aversion in
                Riskless Choice: A Reference-Dependent Model,'' Quarterly Journal of
                Economics, Vol. 106, No. 4, at pp. 1039-61 (Nov., 1991) (pp. 1039-
                1061, specifically pp. 1053-1054). Neither of these references,
                however, makes any claim about the generality of the estimate or its
                applicability to non-experimental settings for consumer behavior.
                 \1612\ See Gal, David, ``A psychological law of inertia and the
                illusion of loss aversion,'' Judgment and Decision Making, Vol. 1,
                No. 1, pp. 23-32 (July 2006,) pp. 23-32,; Erev, I., E. Ert, and E.
                Yechiam, ``Loss aversion, diminishing sensitivity, and the effect of
                experience on repeated decisions.'', Journal of Behavioral Decision
                Making, Vol. 21 (2008), pp. 575-97; (2008); Ert, E., and I. Erev,
                ``On the descriptive value of loss aversion in decisions under risk:
                Six clarifications,'' Judgment and Decision Making, Vol. 8 (2013),
                at pp. 214-35; (2013); Gal, David and Rucker, Derek, ``The Loss of
                Loss Aversion: Will It Loom Larger Than Its Gain?'' Journal of
                Consumer Psychology, Vol. 28 No. 3, (July 2018), at pp. 497-516
                (July 2018) available at (https://onlinelibrary.wiley.com/doi/abs/10.1002/jcpy.1047); and Gal, David, ``Why the Most Important Idea in
                Behavioral Decision-Making Is a Fallacy,'' Scientific American,
                Observations, (July 31, 2018), available at (https://blogs.scientificamerican.com/observations/why-the-most-important-idea-in-behavioral-decision-making-is-a-fallacy/).
                ---------------------------------------------------------------------------
                [[Page 24609]]
                 Another example is commenters' repeated citation of the study of
                households' difficulties in analyzing the financial value of purchasing
                vehicles with higher fuel economy conducted by Turrentine and Kurani,
                which relies on interviews with a limited number of subjects (57
                California households) to conclude that consumers are systematically
                unable to perform the calculations necessary to estimate the value of
                fuel savings.\1613\ These same commenters consistently ignore the
                wealth of detailed, publicly-available information on the fuel economy
                of new vehicle models, and shoppers' ready access to user-friendly
                tools to estimate the savings they are likely to realize from
                purchasing higher-mpg models. These tools include the label that
                prominently displays how much a vehicles' fuel economy will save, or
                conversely, cost a purchaser in fuel costs over 5 years of use in color
                and large type (see Figure VI-63), which is legally required to be
                prominently displayed on all new cars vehicles offered for sale.\1614\
                Separately, new car dealers are also required to prominently display
                the Federal Fuel Economy Guide for each model year of new vehicles
                offered for sale, which provides fuel economy information for all
                vehicles from that model year.\1615\
                ---------------------------------------------------------------------------
                 \1613\ ICCT at p. 4 and Consumers Union at p. 12 (among others),
                citing Turrentine, T.S., & Kurani, K.S., ``Car buyers and fuel
                economy?,'' Energy policy, Vol. 35 No. 2 (2007), at 1213-1223,
                available at https://www.sciencedirect.com/science/article/pii/S0301421506001200, as evidence that most or all new-car shoppers are
                incapable of calculating the savings they would realize from
                purchasing a higher-mpg model, and further misinterpret the study as
                evidence that buyers invariably underestimate the value of increased
                fuel economy. Yet this widely relied-upon analysis included only 57
                households, all located in California. As an illustration, citing
                Turrentine and Kurani, ICCT asserts ``There is substantial
                circumstantial evidence that most consumers in the U.S. place a low
                value on fuel economy.'' See ICCT at 4 (emphasis added). Similarly,
                Consumers Union simply asserts that ``Households do not track
                gasoline prices over time and cannot accurately estimate future gas
                prices or cost savings.'' See Consumers Union at 12, again citing
                Turrentine and Kurani as authority).
                 \1614\ See 15 U.S.C. 1531, et seq., and 49 CFR 575.401.
                 \1615\ 40 CFR 600.405-08 and 600.407-08.
                 [GRAPHIC] [TIFF OMITTED] TR30AP20.309
                
                 Similarly, no commenters offered empirical evidence to support
                their repeated assertions that buyers or the public actually view
                features such as styling, size, or performance as ``positional goods''
                to which other potential buyers might aspire, or considered the
                possibility that high fuel economy or advanced technology (such as
                hybrid or electric propulsion) might themselves represent such
                positional attributes.\1616\ Nor do commenters
                [[Page 24610]]
                provide any empirical evidence that the various aspects of behavior
                they allege lead buyers to underinvest in fuel economy--ranging from
                unwillingness to spend time or effort estimating likely fuel savings,
                to inattentiveness to the economic and social importance of improved
                fuel economy, inability to obtain information about the savings it
                offers them, and incorrect ``framing'' of the choice among models with
                different levels of fuel economy--are widespread, empirically
                significant, or systematically likely to lead buyers to under- rather
                than over-invest in fuel economy.
                ---------------------------------------------------------------------------
                 \1616\ For evidence that prestige appears to be a motivation for
                purchasing advanced-technology vehicles, see Hidrue, Michael K., et
                al., ``Willingness to pay for electric vehicles and their
                attributes,'' Resource and Energy Economics, Vol. 33, Issue 3
                (September 2011), at pp. 686-705; Chua, Wan Ying, Lee, Alvin and
                Sadeque, Saalem 2010, ``Why do people buy hybrid cars?,''
                Proceedings of Social Marketing Forum, University of Western
                Australia, Perth, Western Australia, Edith Cowan University,
                Churchlands, W.A., at pp. 1-13; Liu, Yizao, ``Household demand and
                willingness to pay for hybrid vehicles,'' Energy Economics, Volume
                44, 2014, at pp. 191-197; Hur, Won-Moo, Jeong Woo, and Yeonshim Kim,
                ``The Role of Consumer Values and Socio-Demographics in Green
                Product Satisfaction: The Case of Hybrid Cars,'' Psychological
                Reports, Volume 117, issue 2, October 2015, at pp. 406-427. A useful
                summary of many studies appears in Table 1 (p. 196) of Makoto
                Tanaka, Takanori Ida, Kayo Murakami, Lee Friedman, ``Consumers'
                willingness to pay for alternative fuel vehicles: A comparative
                discrete choice analysis between the US and Japan,'' Transportation
                Research Part A: Policy and Practice, Volume 70, 2014, at pp. 194-
                209 (Table 1 at p. 196). Some of these studies find that buyers are
                apparently willing to pay significant price premiums for the
                prestige or status value of hybrids or battery-electric vehicles--
                which their authors speculate may derive from their ``greenness''--
                because their purchases cannot be explained on the basis of economic
                or financial considerations. Others find that average or typical
                shoppers' willingness to pay advanced-technology vehicles is below
                the price premiums they command, suggesting that their purchasers
                must derive some status or prestige value from owning and driving
                them.
                ---------------------------------------------------------------------------
                 The most frequent argument that an energy paradox or energy
                efficiency ``gap'' exists in the case of fuel economy is the
                observation that many U.S. vehicle buyers seem unwilling to pay higher
                prices for models whose increased fuel economy would appear to repay
                their additional investment within a relatively brief ownership period.
                However, this argument is unpersuasive for at least three reasons: Most
                obviously, it does not acknowledge the possibility that engineering
                studies systematically underestimate costs to produce vehicles with
                higher fuel economy, and thus the prices that buyers would be asked to
                pay for models with improved fuel economy. Nor does it account for
                potential sacrifices in other vehicle attributes that manufacturers may
                make in order to achieve higher fuel economy without increasing
                vehicles' purchase prices beyond consumers' willingness to pay.
                Finally, claims that consumers are acting irrationally by refusing to
                purchase higher-mpg models usually reach this conclusion by comparing
                rates at which they implicitly discount future fuel costs--and thus
                evaluate savings from purchasing more fuel-efficient models--to
                interest rates in financial markets that incorporate time horizons or
                risk profiles that may be very different from those of consumers.
                 Even putting these concerns aside, comparing future fuel savings to
                the costs of purchasing more expensive models that offer higher fuel
                economy demonstrates only that buyers are not behaving as analysts
                expect them to and believe they should behave. These comparisons do not
                demonstrate that consumers are necessarily acting irrationally, and
                cannot diagnose the nature of information shortcomings buyers face,
                reasons that they might interpret such information incorrectly, or
                identify behavioral inconsistencies they may exhibit. In short,
                conjectures about why buyers might undervalue potential savings from
                investing in higher-efficiency vehicle models do not represent evidence
                that they actually do so, and as discussed above, recent research seems
                to show that such behavior is not widespread, if it exists at all.
                 Past joint rulemaking efforts by NHTSA and EPA have repeatedly
                sought to identify a plausible explanation for car buyers' perceived
                undervaluation of improved fuel economy. The agencies have occasionally
                relied on explanations such as consumers' insufficient appreciation of
                the importance of fuel economy, the difficulty of obtaining adequate
                information about the fuel economy of competing models or of converting
                competing models' fuel economy ratings to future fuel costs and
                savings, or consumers' misunderstanding or mistrust of such information
                when it is provided to them. At other times, the agencies have pointed
                to consumers' ``myopia'' about the future--asserting that for some
                reason, they appear to underestimate future fuel costs and savings--or
                argued that shoppers are insufficiently attentive to fuel costs when
                comparing competing models, that the value of improved fuel economy is
                obscured (``shrouded'') by vehicles' other, more visible attributes, or
                that uncertainty about the savings in fuel costs owners will actually
                realize causes them to undervalue those savings when comparing the
                upfront costs of models with different fuel economy.
                 Despite the frequency with which the agencies have cited these
                hypotheses, clear support for any of them remains elusive. Consumers
                have long had ready access to detailed information about individual
                models' fuel economy, which appears prominently on the labels displayed
                by new cars,\1617\ and is published online and in printed outlets that
                shoppers use routinely rely widely on to compare models.\1618\ In
                addition, the fuel economy actually experienced by previous buyers of
                individual models is increasingly reported in readily accessible on-
                line databases.\1619\
                ---------------------------------------------------------------------------
                 \1617\ Fuel economy labels have been displayed on the window
                sticker of all new light duty cars and trucks since the mid-1970s,
                as required by the Energy Policy and Conservation Act. See https://www.epa.gov/fueleconomy/history-fuel-economy-labeling. Among the
                information currently required to be posted on the fuel economy
                label is both an estimated annual fuel cost for the vehicle, as well
                as an estimate of how that cost compares to the fuel cost over five
                years for an average new vehicle, so it is unclear what information
                consumers lack that prevents them from making an informed decision
                in this regard.
                 \1618\ See, e.g., http://www.fueleconomy.gov, where consumers
                can find and compare the fuel economy (and greenhouse gas
                CO2 and smog emissions) of different vehicle models
                across model years, as well as upload information about their own
                real-world fuel economy and compare it to other drivers.
                 \1619\ See id.
                ---------------------------------------------------------------------------
                 Similarly, consumers appear to be well aware of the prices they pay
                for gasoline and how those vary among retail outlets, and are reminded
                clearly and frequently of the financial consequences of their fuel
                economy choices each time they purchase fuel. Increasingly, consumers
                also have ready online access to comparisons of fuel prices at
                competing locations near their homes or along routes they travel.\1620\
                There is also considerable evidence that drivers' forecasts of future
                fuel prices are more accurate than those issued by government agencies
                or private forecasting services.\1621\ Evidence exists
                [[Page 24611]]
                that car buyers and owners anticipate extreme volatility in fuel
                prices, recognize that there is considerable uncertainty about future
                fuel prices and potential savings from driving a higher-mpg model, and
                respond cautiously to these uncertainties when evaluating competing
                vehicle models,\1622\ none of which suggests a market failure as much
                as it suggests that consumers balance multiple, often competing
                objectives, and make choices based on the outcome of such balancing.
                ---------------------------------------------------------------------------
                 \1620\ See, e.g., Gas Buddy, available at www.gasbuddy.com.
                 \1621\ Anderson et al. report evidence that consumers believe
                fuel prices are likely to remain constant in inflation-adjusted
                terms.; see Anderson, Soren T., Ryan Kellogg, and James M. Sallee,
                ``What do consumers believe about future gasoline prices?'' Journal
                of Environmental Economics and Management, vol. 66 no. 3 (2013), at
                pp. 383-403. (2013). Other evidence generally supporting this view
                is reported by Allcott, Hunt, ``Consumers' Perceptions and
                Misperceptions of Energy Costs,'' American Economic Review: Papers &
                Proceedings, Vol. 101 No. 3 (2011), at pp. 98-104, (2011), although
                Allcott finds that some fraction of consumers consistently believes
                that gasoline prices will rise in the future. In related research,
                Anderson et al. demonstrate that consumers' expectations that
                gasoline prices will return to their current levels, even after
                sudden and significant variation, is generally accurate; see
                Anderson, Soren T., Ryan Kellogg, James M. Sallee, and Richard T.
                Curtin, ``Forecasting Gasoline Prices Using Consumer Surveys.''
                American Economic Review: Papers & Proceedings, Vol. 101 No. 3
                (2011), at pp. 110-14. (2011). In contrast to many consumers'
                expectation that fuel prices may vary over the future but will
                generally return to current levels, the U.S. Energy Information
                Administration predicted that gasoline prices would rise
                significantly over the future at the time the two previous rules
                establishing CAF[Eacute]E standards for model years 2012-16 and
                2017-21 were adopted, in 2010 and 2012; see Energy Information
                Administration (EIA), Annual Energy Outlook 2010), Table A12, p.
                131, available at https://www.eia.gov/outlooks/archive/aeo10/pdf/0383(2010).pdf, Table A12, p. 131; and Annual Energy Outlook 2012,
                Appendix A, Table A12, at p. 155, available at https://www.eia.gov/outlooks/archive/aeo12/pdf/appa.pdf, Table A12, p. 155. As of those
                same dates, forecasts of future petroleum prices issued by other
                government agencies and most private forecasting services (with the
                notable exception of HIS-Global Insight, which projected little or
                no increase in future prices) agreed closely with EIA's forecasts
                that prices would increase significantly over both the near- and
                longer-term futures; see EIA, Annual Energy Outlook 2010, Table 10,
                at p. 86; and Annual Energy Outlook 2012, Table 23, available at
                https://www.eia.gov/outlooks/archive/aeo12/table_23.php. Expressed
                in constant-dollar terms, U.S. gasoline prices in 2019 are
                essentially unchanged from those in 2010, although prices have
                varied significantly above and below that level during the
                intervening period. See https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=pet&s=emm_epm0_pte_nus_dpg&f=m.
                 \1622\ For such evidence, see Allcott, Hunt, ``Consumers'
                Perceptions and Misperceptions of Energy Costs,'' American Economic
                Review: Papers & Proceedings, Vol. 101 No. 3 (2011), at pp. 98-104;
                (2011); Greene, David L., (2010). ``How consumers value fuel
                economy: A literature review'' No. EPA-420-R-10-008 (2010) (No. EPA-
                420-R-10-008); Brownstone, David, David Bunch, and Kenneth Train,
                ``Joint Mixed Logit Models of Stated and Revealed Preferences for
                Alternative-Fuel Vehicles,'' Transportation Research Part B, Vol. 34
                (2000), at pp. 315-338, (2000), among many other sources.
                ---------------------------------------------------------------------------
                 In past rulemakings, the agencies have also hypothesized that
                consumers may ``satisfice''--that is, select some minimum acceptable
                level of fuel economy, and then evaluate models that achieve that
                minimum on the basis of their other attributes. This explanation for
                buyers' reluctance to purchase more fuel-efficient vehicles ignores the
                possibility that they do account fully for the value of higher fuel
                economy in their decision-making, but simply value differences in
                vehicles' other attributes more highly than they do fuel economy, which
                would not reveal irrational or myopic behavior.
                 A related argument has been that calculating future savings
                attributable to fuel economy is complicated, so car shoppers resort to
                simplified decision rules to choose among models with different fuel
                economies, and relying on these rules-of-thumb causes them to choose
                models with lower fuel economy.\1623\ However, it is unclear why
                buyers' reliance on simplified procedures or approximations for
                estimating the value of fuel savings would necessarily lead them to
                systematically choose models with lower fuel economies rather than
                leading some to underinvest in fuel economy while others overinvest.
                ---------------------------------------------------------------------------
                 \1623\ See, e.g., 77 FR at 63115 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 The agencies have also frequently described consumers as ``loss
                averse,'' making them reluctant to pay the upfront and certain higher
                prices for models offering better fuel economy when the future savings
                they expect to realize are more distant and less certain.\1624\ The
                agencies' past assumption that loss aversion is universal (and equally
                strong) among new-car shoppers appears to be a simplification that is
                largely unsupported by empirical evidence, and in any case has been
                challenged both as a widespread feature of consumer behavior and more
                specifically as an explanation for vehicle shoppers' reluctance to
                purchase more costly models that offer higher fuel economy.\1625\
                Further, the extremely wide variety of competing models among which car
                buyers can choose enables many of those searching for a model with
                better fuel economy at a comparable price to do so simply by choosing a
                version with fewer other features, which might partly offset the effect
                of their aversion to the prospect of losses from paying a higher
                purchase price. Lastly, the agencies note that both increased fuel
                costs and increased upfront car prices will appear as ``losses,'' so it
                is not obvious why potential buyers would react to the prospects of
                these different forms of losses in different ways.
                ---------------------------------------------------------------------------
                 \1624\ Id. at 63114-15; see also 74 FR at 25511, 25653 (May 7,
                2010).
                 \1625\ See supra notes 1611 and 1612.
                ---------------------------------------------------------------------------
                 OMB Circular A-4 does acknowledge that ``[e]ven when adequate
                information is available, people can make mistakes by processing it
                poorly.'' It goes on to say that people may rely on ``mental rules-of-
                thumb'' that produce errors, or cognitive ``availability'' may lead to
                consumers overstating the likelihood of an event. However, Circular A-4
                also cautions that ``the mere possibility of poor information
                processing is not enough to justify regulation,'' and that potential
                problems with information processing ``should be carefully
                documented.'' Some of the above examples of potential market failures
                may fall into this category, but lack evidentiary support. As with
                claims of asymmetric information, it is very difficult to distinguish
                between information processing errors and behavior consistent with
                consumer preferences for time and other vehicle attributes that differ
                from what government agency analysts believe they should be.
                 Similarly, the agencies have occasionally noted (and seemingly been
                critical of) some consumers' apparent preferences for vehicle
                attributes that convey social status, such as size or styling, and
                suggested that they may give inadequate attention to fuel economy
                because it does not provide similar status. The agencies have also
                suggested that consumers may be reluctant to purchase more fuel-
                efficient models because they associate higher fuel economy with
                inexpensive, less well-designed vehicles. These might be plausible
                explanations, were they not contradicted by concurrent arguments that
                potential buyers are inattentive to or uninformed about fuel economy,
                or have difficulty isolating it from vehicles' other attributes.
                Moreover, the market currently offers a wide range of highly fuel
                efficient (and advanced technology) vehicles at many different price
                points, including in the luxury and performance segments, which belies
                the assumption that fuel economy is inconsistent with positional
                attributes. In any case, consumers' hesitance to choose models offering
                higher fuel economy because they are reluctant to sacrifice
                improvements in other vehicle attributes on which they place higher
                values cannot reasonably be characterized as a market failure.
                 Although past rulemakings have raised the possibility that car
                buyers' apparent tendency to underinvest in fuel economy could
                plausibly be explained by their use of discount rates exceeding those
                the agencies employ to assess the present value of fuel savings, the
                agencies have generally dismissed that possibility. In combination with
                factors such as their valuation of vehicles' attributes other than fuel
                economy, differences in driving habits that affect fuel economy and in
                how much they expect to drive newly- purchased cars, and variation in
                their expectations about future fuel prices, differing attitudes about
                the importance of future costs relative to more immediate ones could
                readily explain buyers' apparent reluctance to purchase models offering
                fuel economy levels that the agencies interpret as privately
                ``optimal.''
                 As with consumption of any good or service, the agencies believe
                consumers'
                [[Page 24612]]
                choice in vehicles represents what economists call ``constrained
                optimization.'' That is, consumers select a bundle of vehicle
                features--within their budget constraint--that optimizes the value to
                them. The agencies also believe, as is the case in every constrained
                consumer choice, that each of these attributes provide what economists
                call diminishing marginal returns (or value) to consumers. For
                instance, the agencies believe that consumers value vehicle size,
                comfort, performance, trim-level, appearance, etc. As such, fuel-saving
                technologies that increase the cost of the car are just one of many
                vehicle attributes that consumers balance against each other. And
                instead of using their entire budget on a single vehicle attribute,
                consumers tend to sacrifice some degree of many or all attributes in a
                degree that varies according to their preferences so that they can
                consume some degree of most or all attributes they value. This means
                that many consumers may not maximize fuel-saving technologies in their
                vehicle selection, but instead may choose some other bundle of
                attributes. The agencies' use of a 30 month pay-back period in this
                analysis--as opposed to fuel-savings over the life of the vehicle--is
                consistent with the constrained optimization consumers perform when
                selecting a vehicle. It is a reasonable representation of consumers'
                valuation of fuel-saving technologies, given the diminishing marginal
                returns of additional fuel economy. If the agencies had used the entire
                undiscounted fuel-savings over the entire life of the vehicle, the
                agencies would be effectively modeling a scenario where consumers
                maximize fuel economy to the detriment of all other vehicle
                attributes--an assumption that is evidently wrong. As such, it is not
                necessary that purchasers do not value lifetime fuel savings--and, in
                all likelihood, purchasers would prefer vehicles with better fuel
                efficiency and all of their preferred attributes--but rather consumers
                are forced to choose between fuel economy and other vehicle attributes
                while weighing how much each attribute contributes to the total cost of
                the vehicle.
                 Finally, the agencies have also previously speculated that vehicle
                producers may be reluctant to offer models featuring the higher levels
                of fuel economy that buyers are willing to pay for, and that buyers'
                apparent underinvestment in fuel economy reflects this lack of choice.
                The agencies have speculated that such behavior by manufacturers could
                arise from their collective underestimation of the value that buyers
                attach to fuel economy, or failing this, from limitations on
                competition among them to supply improved fuel economy, whether
                voluntarily or as a consequence of the industry's structure.\1626\ The
                agencies have also raised the seemingly contradictory argument that
                producers have more complete knowledge about fuel economy than
                potential buyers (``asymmetric information'') causing them to provide
                lower levels than buyers demand, and speculated that deliberate
                decisions by manufacturers may limit the range of fuel economy they
                offer in particular market segments.\1627\
                ---------------------------------------------------------------------------
                 \1626\ See 75 FR at 25653-64 (May 7, 2010); and 77 FR at 63115
                (Oct. 15, 2012).
                 \1627\ See, e.g. 75 FR 25510-13; 76 FR 57315-19; 77 FR 62914.
                ---------------------------------------------------------------------------
                 The overarching theme of these arguments seems to be that vehicle
                manufacturers cannot identify--or can, but voluntarily forego--
                opportunities to increase sales and profits at the expense of their
                rivals by offering models that feature higher fuel economy. The
                agencies have sometimes ascribed this behavior to the risk that
                producers might incur large investments to produce the more fuel-
                efficient models that would enable them to seize these opportunities,
                but subsequently lose sales and profits to competitors who simply
                followed suit after their rivals were successful. This explanation is
                at odds with the customary view that innovative producers can be
                rewarded--substantially, even if only temporarily--with commensurate
                profits that justify taking such risks, when they correctly assess
                consumer demand for innovative features or products.
                 In any case, behavior on the part of individual businesses that
                leaves obvious opportunities to increase profits unexploited by an
                entire industry seems extremely implausible, particularly in light of
                the fact that auto manufacturers are profit-seeking businesses whose
                ownership shares are publicly traded and subject to regular market
                valuation. This notion also seems to ignore the range of choices
                already available in the current automobile market, where
                extraordinarily efficient models are available in nearly every vehicle
                class or market segment, including plug-in hybrid and fully electric
                versions of a rapidly increasing number of models. Automobile
                manufacturers can, and in fact are, competing on the basis of fuel
                economy.
                 The central analysis presented in this final regulatory impact
                analysis does not account for the possibility that imposing stricter
                standards may require manufacturers to make sacrifices in other vehicle
                features that compete with fuel economy, and that some buyers may value
                more highly. If this proved to be the case, more stringent alternatives
                could impose offsetting losses on buyers well beyond the increases in
                vehicle prices that are necessary for manufacturers to recover their
                outlays for adding new technology (or changing design features) to
                improve fuel economy. By doing so, it could significantly reduce the
                estimates of total and net benefits the agencies report. To further
                illustrate this issue, the agencies have conducted a sensitivity
                analysis that incorporates a conservative estimate of consumers'
                valuation of other vehicle attributes, as further discussed in Chapter
                VII of the FRIA accompanying today's notice.\1628\ The agencies also
                recognize that buyers may have time preferences that cause them to
                discount the future at higher rates than the agencies are directed to
                consider in their regulatory evaluations.
                ---------------------------------------------------------------------------
                 \1628\ This sensitivity analysis assumes that consumer's value
                of other vehicle attributes is at least as great as a portion of the
                fuel savings that consumers supposedly ``leave on the table.'' In
                this analysis, the private net benefits of the final rule are a
                positive $15 billion using a 7% discount rate--which is consistent
                with the theory that providing consumers with greater choices will
                enhance their private welfare. The net external benefits are
                identical to the primary analysis, or $34 billion, so the
                sensitivity results show the final rule improves net social benefits
                by $49 billion.
                ---------------------------------------------------------------------------
                 If either case is true--that the analysis is incomplete regarding
                consumer valuation of other vehicle attributes or discount rates used
                in regulatory analysis inaccurately represent consumers' time
                preferences--no market failure would exist to support the hypothesis of
                a fuel efficiency gap. In either case, the agencies' central analysis
                would overstate both the net private and social benefits from adopting
                more stringent fuel economy and CO2 emissions standards. For
                instance, Table VII-93 (Combined LDV Societal Net Benefits for MYs
                1975-2029, CAFE Program, 7% Discount Rate) shows that the CAFE final
                rule would generate $16.1 billion in total social net benefits using a
                7% discount rate, but without the large net private loss of $26.1
                billion, the net social benefits would equal the external net benefits,
                or $42.2 billion. Because government action cannot improve net social
                benefits in the absence of a market failure, if no market failure
                exists to motivate the $26.1 billion in private losses to consumers,
                the net benefits of these final standards would be $42.2 billion.
                 In sum, the agencies do not take a position in this rule on whether
                a fuel
                [[Page 24613]]
                efficiency gap exists or constitutes a failure of private markets.
                Accordingly, the final regulatory impact analysis is not constrained in
                any manner that ensures the private net benefits of more stringent
                standards will necessarily be either positive or negative. In fact,
                however, the analysis supporting this final rule does present a
                situation where adopting more stringent CAFE and CO2
                emission standards aligns consumers' decisions with a simplified
                representation of their own economic interests, and by doing so
                improves their well-being from what they would experience under less
                stringent standards. In other words, our final modelling results
                reflect the case where some fuel efficiency gap persists (albeit of
                smaller magnitude than the agencies found in previous analyses),
                despite our expressed reservations about its likelihood.
                (b) Representing Sales Responses in CAFE/CO2 Analysis
                 The approach used in the NPRM relied on a single model to produce
                the total number of new vehicle sales in each calendar year for a given
                regulatory scenario. Many commenters expressed reservations about the
                predictive capabilities of the model (CARB, North Carolina Department
                of Environmental Quality, EDF, Aluminum Association). As the Aluminum
                Association commented, ``[D]eveloping a model to predict consumer
                reaction to changes in prices is complicated and highly sensitive to
                macroeconomic conditions, consumer confidence and employment levels.''
                \1629\ As discussed above, the agencies agree that development of such
                a model is complicated, and the agencies have elected to simplify the
                approach for the final rule. For the purposes of regulatory evaluation,
                the relevant sales metric is the difference between alternatives rather
                than the absolute number of sales in any of the alternatives. As such
                and in response to these comments and others previously addressed, the
                agencies divided the sales response model for the final rule into two
                parts: A nominal forecast that provides the level of sales in the
                baseline (based primarily upon macroeconomic inputs), and a price
                elasticity that creates sales differences relative to that baseline in
                each year. The nominal forecast does not include price, and is merely a
                (continuous) function of several macroeconomic variables that are
                provided to the model as inputs. While the statistical model used in
                the NPRM attempted to account for the influence of these other factors
                in estimating the price elasticity, the forecast in this analysis
                separates the two completely (as described further below). The price
                elasticity is also specified as an input, but this analysis assumes a
                unit elastic response of 1.0--meaning that a one percent increase in
                the average price of a new vehicle produces a one percent decrease in
                total sales.\1630\
                ---------------------------------------------------------------------------
                 \1629\ NHTSA-2018-0067-11952-4.
                 \1630\ The ``price increase'' in this case represents the new
                vehicle price net of a portion of fuel savings, described further in
                this section.
                ---------------------------------------------------------------------------
                 The revised sales model features three broad changes: (1) It uses
                the change in average vehicle price net of fuel costs instead of
                vehicle prices on their own, (2) it uses macroeconomic factors to
                project baseline sales without considering vehicle prices, and (3) it
                assesses the change in sales across the various regulatory alternatives
                considered using an own-price elasticity from the literature. These
                changes were made in response to comments that consumers are willing to
                pay for some level of fuel economy and vehicle prices and sales are
                simultaneously and jointly determined (e.g. endogenous). This section
                discusses these three broad changes, as well as other more technical
                and minor changes.
                 The first component of the new sales response model is the nominal
                forecast, which is a function (with a small set of inputs) that
                determines the size of the new vehicle market in each calendar year in
                the analysis for the baseline. It leverages some of the same structure
                of the statistical model used in the NPRM, though the dependent
                variable and some of the explanatory variables have changed. It is of
                some relevance that this statistical model is intended only as a means
                to project a baseline sales series. Some commenters raised econometric
                objections about the NPRM specification's ability to isolate the causal
                effect of new vehicle prices on new vehicle sales. The agencies note
                that the nominal forecast model does not include prices and is not
                intended for statistical inference.
                 The forecast is derived from a statistical model that accounts for
                a similar set of exogenous factors related to new light-duty vehicle
                sales. In particular, the model accounts for the number of households
                in the U.S., recent number of new vehicles sold, GDP, and consumer
                confidence. The structure of the forecast model is similar to the NPRM
                model, which also used a ARDL specification, but even the variables
                that are common between the two models have different structural forms
                in the final rule version. In particular, the dependent variable has
                been transformed to reflect the fact that, as some commenters
                suggested, households are an important component of demand for new
                vehicles. As such, the dependent variable is defined as new vehicles
                sold per household.\1631\ While this variable still exhibits the cyclic
                behavior that new vehicle sales exhibit over time, the trend shows the
                number of new vehicles sold per household declining since the 1970's,
                as shown in Figure VI-64, where the dotted line is the trend over time.
                As this time series is non-stationary,\1632\ a lagged variable (the
                value in the previous year) is included on the right-hand side of the
                regression equation. In addition, the model includes a lagged variable
                that represents the three-year running sum of new vehicle sales,
                divided by the number of households in the previous year. This variable
                represents the saturation effect, where the existing number of
                households can only buy so many new vehicles before a significant
                number of households already have one (and do not need to buy another).
                As vehicle durability and cost has increased over time, and average
                length of initial ownership has increased similarly, this variable acts
                to put downward pressure on sales after successive years of high sales
                (particularly during extrapolation).
                ---------------------------------------------------------------------------
                 \1631\ Number of U.S. households is taken from Federal Reserve
                Economic data, https://fred.stlouisfed.org/series/TTLHH.
                 \1632\ Stationary refers to whether a time series statistical
                properties are constant over time. Since car sales are increasing
                over time, the time series non-stationary.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24614]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.310
                BILLING CODE 4910-59-C
                 Similar to the NPRM model, the forecast model includes real U.S.
                GDP,\1633\ but in natural logarithm form (as some commenters suggested
                was more appropriate).\1634\ The final variable is consumer sentiment,
                as measured by the University of Michigan survey of consumers.\1635\ As
                both of these series are non-stationary (determined by applying
                augmented Dickey-Fuller unit root tests to the time series), lagged
                versions of the variables are included to ensure stationarity in the
                residuals. The functional form appears below in Equation 2.
                ---------------------------------------------------------------------------
                 \1633\ Federal Reserve Economic Data, available at https://fred.stlouisfed.org/series/GDPC1#0.
                 \1634\ EPA-HQ-OAR-2018-0283-6220-1.
                 \1635\ http://www.sca.isr.umich.edu/tables.html.
                ---------------------------------------------------------------------------
                 Equation 2--Statistical Model Used to Generate Nominal Forecast
                 The model fit is described in Table VI-152. The included lag term
                of the dependent variable and both GDP variables are statistically
                significant at nearly zero, while both the lagged three year sum term
                and consumer sentiment are both marginally significant. Being a time
                series model, the agencies also computed the Durbin-Watson test
                statistic for autocorrelation (1.77) and the Breusch-Godfrey test for
                serial correlation (0.65) at order 1. The signs of the coefficients are
                all correct, in the sense that they are consistent with our
                expectations.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.311
                [[Page 24615]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.312
                BILLING CODE 4910-59-C
                 Because the dependent variable is the number of new vehicles sold
                per household, it is necessary to multiply by the number of households
                to produce an estimate of new vehicle sales. This model is used to
                produce a forecast of new vehicle sales out to 2050, so it is necessary
                to have projections of each variable used in Equation 2 through
                calendar year 2050. In an effort to be consistent with other inputs to
                the analysis, the projection of U.S. GDP is taken from the 2019 AEO.
                The forecast of households in this analysis comes from the Harvard
                Joint Center for Housing Studies 2018 Household projections.\1636\ The
                consumer confidence forecast is taken directly from the University of
                Michigan index for 2017 and 2018, and from the Global Insight forecast
                of consumer confidence for all subsequent years.
                ---------------------------------------------------------------------------
                 \1636\ https://www.jchs.harvard.edu/research-areas/working-papers/updated-household-growth-projections-2018-2028-and-2028-2038.
                ---------------------------------------------------------------------------
                 While the analysis could have relied on a forecast of new vehicle
                sales taken from a published source (the 2019 AEO, for example), using
                a function is an attractive option because it allows the CAFE Model
                dynamically to adjust the forecast in response to input changes. If a
                sensitivity case requires a forecast that is consistent with a set of
                specific, possibly unlikely, assumptions, a forecast of new vehicle
                sales that is consistent with those assumptions may not exist in the
                public domain, for example low GDP growth sensitivity cases. As
                implemented in this rulemaking, using a functional form allows the user
                to vary some of the assumptions to the analysis without creating
                inconsistencies with other elements of the analysis. However, it is
                incumbent upon the analyst to ensure that any set of assumptions that
                deviate from the central analysis are logically consistent.
                 This function, and the set of assumptions contained in the central
                analysis, produces a projection that is comparable in magnitude to the
                forecast in the 2019 AEO reference case, though there are differences.
                The two forecasts, and the percentage difference relative to the AEO
                2019, appear in Table VI-153, as does a recent forecast published by
                the Center for Automotive Research.\1637\ The reader will notice that
                even 2017 shows a discrepancy of nearly 7 percent between the final
                rule forecast and the Annual Energy Outlook, one of the larger
                differences between annual forecasts. However, the final rule analysis
                is based upon the certified production volumes of MY2017, which exceed
                17 million units. So, while the difference may seem significant, the
                final rule volumes in 2017 represent the ground truth for model year
                production.\1638\ The CAR forecast, while shorter in length, is
                consistently higher than both the AEO and final rule forecasts--though
                likely also includes class 2b (and possibly class 3) pickup trucks in
                its light vehicle forecast. Finding a public forecast that explicitly
                excludes light-duty vehicles exempt from these regulations is
                challenging. However, all three forecasts exhibit similar trends--
                decreases in sales starting in 2019 that last for a few years before
                ticking up again slowly. As commenters observed, all forecasts are
                almost guaranteed to have some errors, and projections out to 2050
                should be taken as potential future projections limited by our
                knowledge at the time, rather than an ironclad prediction of the
                future.
                ---------------------------------------------------------------------------
                 \1637\ https://www.cargroup.org/u-s-light-vehicle-sales-expected-to-take-a-dip-in-2019/, last accessed 11.21.2019.
                 \1638\ See CAFE Public Information Center, https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Home.htm.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24616]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.313
                BILLING CODE 4910-59-C
                 Although the forecast produces the total number of new vehicle
                sales in the baseline, an elasticity is imposed on price differences to
                produce sales changes between alternatives. The NPRM version of the
                model considered only differences in average new vehicle prices between
                alternatives, and the agencies received a number of comments (from CBD,
                IPI, EDF, CARB, CA et al., and Oakland et al., as well as recent peer
                reviewers) encouraging the agencies to account for some component of
                fuel savings associated with those price changes. In their comment,
                California et al. and Oakland et al. stated the model failed ``to
                consider
                [[Page 24617]]
                how consumers will respond to the reduced cost of operating the vehicle
                from better gas mileage and therefore inaccurately predicts a decline
                in vehicle sales under the existing standards.'' \1639\ The agencies
                agree that price is not the only consideration, and that the value of
                fuel savings to new vehicle buyers is also relevant to the purchase
                decision.
                ---------------------------------------------------------------------------
                 \1639\ States and Cities, Attachment 1, NHTSA-2018-0067-11735,
                at 86.
                ---------------------------------------------------------------------------
                 In previous rules, while the agencies produced analyses that
                qualitatively considered sales and employment impacts, the agencies
                acknowledged that fuel economy and CO2 standards were likely
                to increase vehicle prices, while simultaneously reducing operating
                costs, and that estimating how consumers would choose to balance those
                two factors in the new vehicle market was challenging.\1640\
                Furthermore, the agencies recognized that there is a broad consensus in
                the economic literature that the price elasticity of demand for
                automobiles is approximately -1.0.\1641\ The agencies feel that a unit
                elasticity of -1.0 is still a reasonable estimate.\1642\
                ---------------------------------------------------------------------------
                 \1640\ Final Regulatory Impact Analysis, Corporate Average Fuel
                Economy for MY 2017-MY 2025 Passenger Cars and Light Trucks, August
                2012, at 821.
                 \1641\ See, e.g., Kleit, A.N., ``The Effect of Annual Changes in
                Automobile Fuel Economy Standards,'' Journal of Regulatory
                Economics, Vol. 2 (1990), at pp 151-72; Bordley, R., ``An
                Overlapping Choice Set Model of Automotive Price Elasticities,''
                Transportation Research B, Vol. 28B no. 6 (1994), at pp 401-408; and
                McCarthy, P.S. ``Market Price and Income Elasticities of New Vehicle
                Demands,'' The Review of Economics and Statistics, Vol. LXXVII no. 3
                (1996), at pp. 543-547.
                 \1642\ For example, a recent review of 12 studies examining
                vehicle price elasticities conducted by the Center of Automotive
                Research (``CAR'') found an ``average short-run elasticity of -
                1.09'' and focusing ``only those models which also employ time
                series methods, the average short-run own-price elasticity is higher
                yet, at -1.25.'' CAR's own analysis found a -.79 short-run
                elasticity. Appendix II of the CAR report shows that the long-run
                elasticities ranged from -.46 and -1.2 with an average of -.72. In
                sum, a -1.0 elasticity is well-aligned with the totality of
                research. McAlinden Ph.D., Sean P., Chen, Yen, Schultz, Michael,
                Andrea, David J., The Potential Effects of the 2017-2025 EPA/NHTSA
                GHG/Fuel Economy Mandates of the US Economy, Center for Automotive
                Research, Ann Arbor, MI (Sept. 2016), available at https://www.cargroup.org/wp-content/uploads/2017/02/The-Potential-Effects-of-the-2017_2025-EPANHTSA-GHGFuel-Economy-Mandates-on-the-US-Economy.pdf.
                ---------------------------------------------------------------------------
                 Because the elasticity assumes no perceived change in the quality
                of the product, and the vehicles produced under different regulatory
                scenarios have inherently different operating costs, the price metric
                must account for this difference. As commenters suggested is
                appropriate, the price to which the unit elasticity is applied in this
                analysis represents the residual price change between scenarios after
                accounting for 2.5 years' worth of fuel savings to the new vehicle
                buyer. This approach is consistent with the 2012 FRIA analysis of sales
                impacts, that which considered several payback periods over which the
                value of fuel savings was subtracted from the change in average new
                vehicle price.
                 Similar to the NPRM, the price elasticity is applied to the
                percentage change in average price (in each year). However, the average
                price to which the elasticity is applied is calculated differently in
                the final rule in response to comments. As discussed below the price
                change does not represent an increase/decrease over the last observed
                year, but rather the percentage change relative to the baseline. In the
                baseline, the average price is defined as the observed new vehicle
                price in 2017 plus the average regulatory cost associated with the
                alternative. In the case of CO2 standards, the regulatory
                cost is equivalent to the retail equivalent price of technology
                improvements. In the case of CAFE standards, the regulatory cost
                includes both technology costs and civil penalties paid for non-
                compliance in a model year. So the change in sales for alternative a in
                year y is:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.314
                 [Delta]RegCost is the difference in average regulatory cost between
                alternative a and the baseline scenario in year y to make a vehicle
                compliant with the standards, $34,449 is the average transaction price
                of a new vehicle in 2016, NominalSales is the forecasted sales (in the
                baseline) in year y, [Delta]FuelCosts is the change in average fuel
                costs over 2.5 years relative to the baseline in year y and
                PriceElasticity is -1.0:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.315
                 Where 35,000 miles is assumed to be equivalent to 2.5 years of
                vehicle usage.\1643\ The agencies assume that consumers behave as if
                the fuel price faced at the time of purchase is the fuel price that
                they will face over the first 2.5 years of ownership and usage.
                Essentially, they behave as if fuel prices follow a random walk,
                where the best prediction of (near) future prices is the price
                today. Scrappage rates in the first few years of ownership are close
                to zero, so buyers can reasonably expect to travel the full annual
                mileage in each of the first three years of ownership. Total sales
                in each alternative (that is not the baseline) will equal
                NominalSalesy + [Delta]Salesa,y for
                alternative a in year y.
                ---------------------------------------------------------------------------
                 \1643\ Based on odometer data, 35,000 miles is a good
                representation of typical new vehicle usage in the first 2.5 years
                of ownership and use--though the distribution of usage is large.
                 This implementation produces a range of differences in total sales,
                both between alternatives and over time. Table VI-154 shows the range
                of differences in the final rule at the industry level for
                CO2, and Table VI-155 shows the sales changes under CAFE.
                While cost decreases between the baseline and alternatives differ by
                program, one can see that removing the value of fuel savings from the
                price limits the sales increases in the alternatives to under 300,000
                units in a single year under the preferred alternative, and about one
                percent of total sales between 2017 and 2050.
                BILLING CODE 4910-59-P
                [[Page 24618]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.316
                [GRAPHIC] [TIFF OMITTED] TR30AP20.317
                BILLING CODE 4910-59-C
                 Table VI-154 and Table VI-155 show sales under the baseline
                (augural standards), and differences under the proposal (0 percent
                increase in stringency) and final rule (1.5 percent increase in
                stringency) of MYs 2017-2050.
                c) Dynamic Fleet Share (DFS)
                 The first module described above (the forecast function and applied
                elasticity)
                [[Page 24619]]
                determine the total industry sales in each model year from 2018 (in
                this analysis, 2017 is based on certified compliance data) to 2050. A
                second module, the dynamic fleet share, acts to distribute the total
                industry sales across two different body-types: ``cars'' and ``light
                trucks.'' While there are specific definitions of ``passenger cars''
                and ``light trucks'' that determine a vehicle's regulatory class, the
                distinction used in this phase of the analysis is more simplistic. All
                body-styles that are obviously cars--sedans, coupes, convertibles,
                hatchbacks, and station wagons--are defined as ``cars'' for the purpose
                of determining fleet share. Everything else--SUVs, smaller SUVs
                (crossovers), vans, and pickup trucks--are defined as ``light
                trucks''--even though they may not be treated as such for compliance
                purposes. In the case of SUVs, in particular, many models may have
                sales volumes that reside in both the passenger car and light fleets
                for regulatory purposes, but the dynamic fleet share does not make this
                distinction. The fleet share model was applied at the same level in the
                NPRM--namely, at the level of body-style rather than regulatory class.
                EDF expressed concern that any simulated increase in the light truck
                share represented consumers shifting from sedans to either 4WD drive
                crossovers, SUVs or pickup trucks.\1644\ However, this was not the
                case. All crossovers are considered light trucks for the purposes of
                fleet share, even though they may be 2WD crossovers treated as
                passenger cars for compliance purposes. So, while the number may
                increase overall for a given scenario, the proportion of crossovers
                sold as 4WD, rather than 2WD, does not.
                ---------------------------------------------------------------------------
                 \1644\ EDF, Appendix B, NHTSA-2018-0067-12108, at 40-41.
                ---------------------------------------------------------------------------
                 EDF was also concerned that the sales implementation in the NPRM,
                which relied on the absolute average price to determine differences
                between alternatives, was unduly influenced by fleet share--as
                differences in the share of light-trucks had the potential to skew
                differences in average price because light-trucks are generally more
                expensive than sedans and hatchbacks. The final rule implementation,
                which starts from an observed average transaction price and evolves the
                average price in the alternatives based on average regulatory cost, is
                less vulnerable to this potential distortion. Even if the fleet share
                model (described in greater detail below) increases the share of light
                trucks (for example), the inherent price difference between passenger
                cars and light trucks does not pass through to the average price--only
                the relative difference in compliance costs associated with the vehicle
                types. Despite the fact that light trucks have generally higher
                transaction prices than passenger cars, there is no guarantee that
                regulatory costs will be higher for light-trucks than for cars (which
                depend upon the mix of footprints, their distance from the relevant
                curve, and the technology cost needed to bring each fleet into
                compliance). Thus, the average price differences used in the sales
                calculations are relatively unaffected by the fleet share model.
                 As in the NPRM, the dynamic fleet share represents two difference
                equations that independently estimate the share of passenger cars and
                light trucks, respectively, given average new market attributes (fuel
                economy, horsepower, and curb weight) for each group and current fuel
                prices, as well as the prior year's market share and prior year's
                attributes. The two independently estimated shares are then normalized
                to ensure that they sum to one. As with the Sales Response model, the
                DFS utilizes values from one and two years preceding the analysis year
                when estimating the share of the fleet during the model year being
                evaluated. For the horsepower, curb weight, and fuel economy values
                occurring in the model years before the start of analysis, the DFS
                model uses the observed values from prior model years. After the first
                model year is evaluated, the DFS model relies on values calculated
                during analysis by the CAFE model. The DFS model begins by calculating
                the natural log of the new shares during each model year, independently
                for each vehicle class, as specified by the following equation:
                [[Page 24620]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.318
                HPVC,MY	1: The average horsepower of all vehicle models belonging to
                vehicle class VC, in the year immediately preceding model year MY,
                ---------------------------------------------------------------------------
                 \1645\ As discussed elsewhere in this final rule, model year and
                calendar year are assumed to be equivalent in the simulation--as
                they always have been in all prior rulemaking analyses.
                ---------------------------------------------------------------------------
                HPVC,MY	2: The average horsepower of all vehicle models belonging to
                vehicle class VC, in the year preceding model year MY by two years,
                CWVC,MY	1: The average curb weight of all vehicle models belonging
                to vehicle class VC, in the year immediately preceding model year
                MY,
                CWVC,MY	2: The average curb weight of all vehicle models belonging
                to vehicle class VC, in the year preceding model year MY by two
                years,
                FEVC,MY	1: The average on-road fuel economy rating of all vehicle
                models (excluding credits, adjustments, and petroleum equivalency
                factors) belonging to vehicle class VC, in the year immediately
                preceding model year MY,
                FEVC,MY	2: The average on-road fuel economy rating of all vehicle
                models (excluding credits, adjustments, and petroleum equivalency
                factors) belonging to vehicle class VC, in the year preceding model
                year MY by two years,
                0.423453: a dummy coefficient, and
                1n(ShareVC,MY): The natural log of the calculated share of the total
                industry fleet classified as vehicle class VC, in model year MY.
                 In the equation above, the beta coefficients, [beta]C through
                [beta]Dummy, are provided in the following table. The beta coefficients
                differ depending on the vehicle class for which the fleet share is
                being calculated.
                [[Page 24621]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.319
                 Once the initial car and light truck fleet shares are calculated
                (as a natural log), obtaining the final shares for a specific vehicle
                class is simply a matter of taking the exponent of the initial value,
                and normalizing the result at one (or 100%). This calculation is
                demonstrated by the following:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.320
                 These shares are applied to the total industry sales derived in the
                first stage of the sales response. This produces total industry volumes
                of car and light truck body styles. Individual model sales are then
                determined from there based on the following sequence: (1) individual
                manufacturer shares of each body style (either car or light truck)
                times the total industry sales of that body style, then (2) each
                vehicle within a manufacturer's volume of that body-style is given the
                same percentage of sales as appear in the 2017 fleet. This implicitly
                assumes that consumer preferences for particular styles of vehicles are
                determined in the aggregate (at the industry level), but that
                manufacturers' sales shares of those body styles are consistent with
                MY2017 sales. Within a given body style, a manufacturer's sales shares
                of individual models are also assumed to be constant over time. The
                agencies assume that manufacturers are currently pricing individual
                vehicle models within market segments in a way that maximizes their
                profit. Without more information about each OEM's true cost of
                production and operation, fixed and variables costs, and both desired
                and achievable profit margins on individual vehicle models, the
                agencies have no reason to assume that strategic shifts within a
                manufacturer's portfolio will occur in response to standards.
                 The Global Automakers noted in their comments that the market share
                of SUVs continues to grow, while conventional passenger car body-styles
                continue to lose market share.\1646\ The agencies are aware of this,
                and include the DFS model in an attempt to address these market
                realities. In the 2012 final rule, the agencies projected fleet shares
                based on the continuation of the baseline standards (MY2012-2016) and a
                fuel price forecast that was much higher than the realized prices since
                that time. As a result, that analysis showed passenger car body-styles
                comprising
                [[Page 24622]]
                about 70 percent of the new vehicle market by 2025. The reality, as
                Global Automakers note, has been quite different.
                ---------------------------------------------------------------------------
                 \1646\ Global Automakers, Attachment A, NHTSA-2018-0067-12032,
                at 13.
                ---------------------------------------------------------------------------
                 The coefficients of the DFS model show passenger car styles gaining
                share with higher fuel prices and losing them when prices are lower.
                Similarly, as fuel economy increases in light truck models, which offer
                consumers other desirable attributes beyond fuel economy (ride height
                or interior volume, for example) their relative share increases. NRDC,
                in particular, found this counterintuitive.\1647\ However, this
                approach does not suggest that consumers dislike fuel economy in
                passenger cars, but merely recognizes the fact that fuel economy has
                diminishing returns. As the fuel economy of light trucks increases, the
                tradeoff between passenger car and light truck purchases increasingly
                involves a consideration of other attributes. Similarly, the
                coefficients show a relatively stronger preference for power
                improvements in cars than light trucks because that is an attribute
                where trucks have outperformed cars, like cars have outperformed trucks
                for fuel economy.
                ---------------------------------------------------------------------------
                 \1647\ NRDC, Attachment 3, NHTSA-2018-0067-11723, at 5.
                ---------------------------------------------------------------------------
                 Rather than estimate new functions to determine relative market
                shares of cars and light trucks, the agencies applied existing
                functions from the transportation module of the National Energy
                Modeling System (NEMS) that was used to produce the 2017 Annual Energy
                Outlook. The functions above appear in the ``tran.f'' input file to
                that version of NEMS, and were embedded (in their entirety) in the CAFE
                model in the NPRM (and this final rule). NEMS uses the functions to
                estimate the percent of total light vehicles less 8,500 GVW that are
                cars/trucks. While NRDC asserted that the agencies must demonstrate the
                propriety of the fleet share model before relying on its
                estimates,\1648\ they ignore the fact that, by using the AEO to develop
                a static fleet in prior rulemakings, the agencies have always relied on
                NEMS estimates. The primary difference between those analyses and the
                NPRM (and this final rule), is that prior analyses applied the fleet
                share that was simulated for the baseline to all regulatory scenarios
                considered. Based on the fleet share functions in NEMS, NPRM corrected
                this internal inconsistency found in previous analyses. This approach
                also enables consistent sensitivity cases--where higher fuel prices
                produce fleets with more transitional passenger car body styles, for
                example--and ensures that the starting point (MY 2017) evolves in
                response to both fuel economy improvements and fuel prices in a way
                that is internally consistent.
                ---------------------------------------------------------------------------
                 \1648\ Id.
                ---------------------------------------------------------------------------
                 The agencies are making one change to the DFS function, which is
                the level of application. While NEMS intended the fleet shares to be
                defined by regulatory classes, vehicles are defined much more coarsely
                in NEMS than in the CAFE model, and manufacturers are not
                differentiated at all. In order to produce well-behaved fleet share
                projections with this model, the agencies applied the share functions
                to body-styles rather than regulatory classes. For many years, there
                was little overlap between nameplates in a manufacturer's passenger car
                regulatory class and its light truck regulatory class. However, with
                the recent emergence of smaller FWD SUVs and crossovers, it is
                increasingly common to have nameplates with model variants in both the
                passenger car and light truck regulatory classes, and it is also common
                for there to be only minor differences (like the presence of 4WD or
                AWD) between versions regulated as cars and versions regulated as light
                trucks. The agencies have modified the application of the fleet share
                equations to focus on body-style, rather than regulatory class, in
                recognition of the increased ambiguity between the regulatory class
                distinction for popular models like the Honda CR-V and Toyota RAV4,
                that sell more than 100K units in each regulatory class (typically
                using the same powertrain configuration). The Nissan Rogue sold more
                than 400K units in MY2017, and almost exactly half of them were in the
                light truck (LT) regulatory class. Applying the fleet share at the
                body-style level preserves the existing regulatory class splits for
                nameplates that straddle the class definitions. It also serves to
                minimize the deviation from the observed MY2017 regulatory class shares
                over time. Had the agencies applied the share equations at the
                regulatory class level, as some commenters incorrectly claimed the
                agencies were doing in the proposal, the passenger car regulatory class
                would have eroded much faster than we've seen in the real world and
                ceased to resemble the composition of the MY2017 fleet. Our
                implementation allows the passenger car (PC) regulatory class to
                continue evolving toward crossover-type cars, if that is what economic
                and policy conditions favor.\1649\
                ---------------------------------------------------------------------------
                 \1649\ The ``passenger car'' fleet for CAFE represents the
                combination of both imported passenger cars (IC) and domestic cars
                (DC). While Table VI-157 illustrates shares for the CAFE program,
                resulting shares under the tailpipe CO2 emissions
                standards are comparable.
                ---------------------------------------------------------------------------
                [[Page 24623]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.321
                 Table VI-157 shows the regulatory class shares under the baseline
                (augural standards), proposal (0 percent increase in stringency), and
                final rule (1.5 percent increase in stringency) between 2017 and 2030.
                The shares move relatively little between the classes in the baseline,
                with larger (but still small) deviations occurring in the least
                stringent alternative (0 percent increase) and the final rule. As the
                sensitivity cases show, the changes in shares (both over time and
                between regulatory classes) respond to the fuel price case, but remain
                internally consistent due to the inclusion of the DFS.
                 Some commenters encouraged the agencies to consider vehicle
                attributes beyond price and fuel economy when estimating a sales
                response to fuel economy/CO2 standards, and suggested that a
                more detailed representation of the new vehicle market would allow the
                agencies to simulate strategic mix shifting responses from
                manufacturers and diverse attribute preferences among consumers. Doing
                so would have required a discrete choice model (at some level), and
                below the reasons why the agencies have not chosen to employ that
                approach in this final rule.
                d) Using Vehicle Choice Models in Rulemaking Analysis
                 Some commenters argued that the NPRM's statistical model used to
                estimate changes in sales between alternatives was too highly
                aggregated and missed consumers' valuation of other vehicle attributes.
                CARB, Cities and States, and EDF all made some version of the argument
                that the sales model in the NPRM operated at too high a level of
                aggregation to estimate the real sales response, which primarily occurs
                at the model level where consumers are making decisions based on the
                comprehensive set of attributes and body styles available in the
                market. They also argued that a model must operate at the same level,
                such as a discrete choice model, in order to capture consumer response
                accurately. EPA's Science Advisory Board, Bento, Toyota, Automobile
                Alliance, RFF, and Bunch (writing on behalf of CARB) insisted that the
                best approach to estimating the change in sales across alternatives is
                to use a discrete choice model and embed it in the simulation.
                 Other commenters expressed different views on the importance of a
                consumer choice model. For example, while the Aluminum Association
                supported a consumer choice model, they suggested that total new
                vehicle sales may not change due to increases in price, but rather the
                attributes of new vehicles would shift, as consumers would likely shift
                their purchases toward lower content vehicles (in terms of safety,
                luxury, or other option content) when faced with generally higher
                prices. Other commenters, including UCS and CBD, strongly encouraged
                the agencies to avoid using consumer choice models; commenters asserted
                that consumer choice models have historically lacked reliability and
                predictive power.\1650\
                ---------------------------------------------------------------------------
                 \1650\ UCS, Technical Appendix, NHTSA-2018-0067-12039 at 50.
                ---------------------------------------------------------------------------
                 In general, these various comments present the agencies with
                considerably different suggestions on how to address these issues, and
                certain suggestions are in direct opposition to each other. That is,
                while some commenters argue that only micro-level consumer responses
                are relevant to the analysis, and that a consumer choice model is
                required to estimate these responses, others argue that it is
                inappropriate to use a discrete choice model--the method by which those
                responses are econometrically estimated--in a regulatory analysis.
                Adding to the confusion, some of the same commenters who argued against
                a consumer choice model,\1651\ also argued that it was necessary for
                the analysis to account for the influence of other vehicle attributes
                in purchasing decisions, which would require incorporating a discrete
                choice model.
                ---------------------------------------------------------------------------
                 \1651\ For example, see EDF, NRDC, RFF, NCAT, and CBD comments.
                ---------------------------------------------------------------------------
                 CARB argued that ``accurately capturing the relative impact of
                sales shifts versus no-buy decisions would require a more detailed
                consumer choice model, as recommended by the CAFE Model peer reviewers.
                The current new vehicle sales model has no
                [[Page 24624]]
                way of capturing these types of effects.'' \1652\
                ---------------------------------------------------------------------------
                 \1652\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 192.
                ---------------------------------------------------------------------------
                 David Bunch, writing for CARB, said, ``In fact, in previous
                versions of the CAFE model there were no attempts to directly simulate
                consumer response from within the CAFE model at all. Instead, NHTSA
                relied on fixed projections of future vehicle market behavior from
                multiple sources for the purpose of performing the required economic
                cost and benefit calculations. While this might possibly be less than
                ideal, this approach is only a problem if, in the real world, there
                [are] notable differences in future market behavior [that] occur under
                different regulation scenarios, and, moreover, that these differences
                would be large enough to compromise the validity of the net benefit
                comparisons.'' Bunch essentially argues that the old approach,
                asserting that standards can have no impact on sales, even at the
                individual model level, is more appropriate than trying to capture the
                general idea that when all new vehicles get more expensive, consumers
                are likely to buy fewer of them, all else being equal. The agencies
                disagree with that perspective.
                 There are a number of practical challenges to using estimates of
                consumer attribute preferences to simulate market responses. Discrete
                choice models typically rely on fixed effects (or alternative-specific
                constant terms) to account for the unobserved characteristics of a
                given model that influence purchasing decisions, such as styling,\1653\
                but are not captured by independent variables that represent specific
                vehicle attributes (horsepower, interior volume, or safety rating, for
                example). Ideally, these constant terms would contribute relatively
                little to the fit and performance of the model, assuming that the most
                salient characteristics are accounted for explicitly. In practice, this
                is seldom the case. While the fixed effects at the model level are
                statistically sound estimates of consumer preferences for the
                unobserved vehicle characteristics of the individual models, the
                estimates are inherently historical--based on observed versions of the
                specific vehicle models to which they belong. However, once the
                simulation starts, and new technologies are added to each
                manufacturer's product portfolio over successive generations, it is no
                longer obvious that those constant terms would still be valid in the
                context of those changes.
                ---------------------------------------------------------------------------
                 \1653\ Aesthetics such as styling are difficult, if it not
                impossible, to define in a manner that allows meaningful comparison
                between choices.
                ---------------------------------------------------------------------------
                 Another complication is that discrete choice models are highly
                dependent on their inputs and are unable to account for future market
                changes. For example, the Draft TAR relied on a MY 2014 market (for
                EPA's analysis) and a MY 2015 market (for NHTSA's analysis), while the
                NPRM used a MY 2016 fleet, and this final rule has updated the market
                characterization to a MY 2017 fleet. A discrete choice model estimated
                on any of those model years would probably produce different fixed
                effects estimates for each model variant in the fleet. Even assuming
                that no new variants of a given model are offered over time, new
                nameplates emerge as others are retired--and for those new nameplates
                and all of their model variants, no constant terms would exist. They
                would have to be imputed (either from comparable vehicles in the
                market, some combination of their attributes, or both). Some studies
                have attempted to estimate fixed effects for a single new entrant to
                the market,\1654\ but none have attempted to do so at the scale
                required to migrate a discrete choice model fit on an earlier model
                year to a newer model year for simulation.
                ---------------------------------------------------------------------------
                 \1654\ Berry, Steven, James Levinsohn, and Ariel Pakes (2004).
                Differentiated products demand systems from a combination of micro
                and macro data: The new car market. Journal of Political Economy
                112(1): 68-105.
                ---------------------------------------------------------------------------
                 Figure VI-65 shows the cumulative percentage of nameplates in the
                2017 new vehicle market by year of introduction. About ten percent of
                nameplates in 2017 have been around since the 1970s, but another ten
                percent have only existed since about 2010. This fact illustrates the
                likely necessity of constructing vehicle model fixed effects for the
                inevitable new entrants between the estimating fleet and the rulemaking
                fleet. But it also suggests another challenge. New model entrants are
                driven by the dynamics of the market, where some vehicle models succeed
                and others fail, but a simulated market with a discrete choice model
                can only simulate failure--where consumer demand for specific
                nameplates erode to the point that the nameplate volumes trend toward
                zero. It has no mechanism to generate new nameplates to replace those
                nameplates whose sales it estimates will erode beyond some minimal
                practical level of production.
                 Consumer choice models are typically fit on a single year of data
                (a cross-section of vehicles and buyers), but this approach misses
                relevant trends that build over time, such as rising GDP or shifting
                consumer sentiment toward emerging technologies. If such a model is
                used to estimate total sales, but lacks trends in GDP growth or
                employment, etc., it will have the wrong set (likely a smaller set) of
                new vehicle buyers and exaggerate price responses and attribute
                preferences. Consumer preferences change over time in response to any
                number of factors--given manufacturers' recent investments in electric
                powertrains, they are counting on this fact. But a choice model
                estimated on observed consumer preferences for EVs--or other vehicle
                attributes with comparatively little experience in the market--would
                necessarily disadvantage a technology that is currently (or only
                recently) unpopular, but gaining popularity. While these are problems
                that may not matter in the estimation process, where a researcher is
                attempting to measure revealed consumer preference for given attributes
                at a single point in time, they become material once that model is
                integrated into the simulation and dynamically carried forward for
                three decades. The agencies note that models that examine aggregate
                trends, such as the one utilized in this analysis, are able to side-
                step this issue by not placing a value on unique vehicle attributes.
                [[Page 24625]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.322
                 The agencies' compliance simulation model estimates the additional
                cost of technology required to achieve compliance, or to satisfy market
                demand for additional fuel economy. While it necessarily calculates
                these costs on a per-vehicle basis, estimating the cost of additional
                technologies as they are applied to each specific model in order to
                bring an entire fleet into compliance, it is agnostic about how these
                costs are distributed to buyers. Manufacturers have strategic, complex
                pricing models that rely on extensive market research and reflect each
                company's strategic interests in each segment. Automobile companies
                attempt to maximize profit from the sale of their vehicles, rather than
                solely focusing on minimizing the cost of compliance, as this
                rulemaking simulates. Lacking reliable data for each manufacturer on
                production costs and profit margins for each vehicle model in their
                portfolios, the most reasonable course of action is to simulate
                compliance as if OEMs are attempting to minimize costs, and, worth
                noting, this approach is also the one NHTSA takes in its rulemakings
                related to the FMVSS. However, it is obvious that some market segments
                and individual models are much less elastic than others.\1655\ As
                reflected in the prices of those models, consumers are able to bear a
                greater share of the total cost of compliance before negatively
                affecting sales and manufacturer profits.
                ---------------------------------------------------------------------------
                 \1655\ See, for example, Kleit, A.N. (2004), Impacts of Long-
                Range Increases in the Fuel Economy (CAFE) Standard. Economic
                Inquiry, 42: 279-294. doi:10.1093/ei/cbh060.
                ---------------------------------------------------------------------------
                 Several commenters (CARB, CBD, IPI, and Bento et al.) suggested
                that the agencies should employ a pricing model that allows
                manufacturers to vary prices in response to heterogeneous consumer
                preferences and different levels of willingness to pay for fuel
                economy, and other attributes, in the new vehicle market.
                Fundamentally, this would require the agencies to model strategic
                pricing for each manufacturer individually--no single pricing model
                would be appropriate for every manufacturer. Bento et al. stated that
                the agencies should simulate the market by allowing manufacturers to
                dynamically adjust vehicle prices to ensure compliance with the
                standards.\1656\ There is no reasonable expectation that the agencies
                could embed and utilize each manufacturer's pricing strategy, as this
                is an essential feature of competitive corporate behavior and that
                automakers closely hold pricing strategy information and the agencies
                have insufficient information to model manufacturer pricing strategies.
                Furthermore, models in the academic literature that commenters have
                suggested are superior because they allow prices to adjust, merely
                demonstrate that the mechanics of those adjustments work; they do not
                imply that the resulting prices are reasonable or realistic. Given the
                burden to estimate each manufacturer's standard under the attribute-
                based system, where the mix of vehicles sold defines not only the
                achieved fuel economy of each fleet but also the standard to which it
                is compared, the agencies are understandably reluctant to implement
                models that might shift a manufacturer's mix of vehicles sold within a
                market segment.
                ---------------------------------------------------------------------------
                 \1656\ NHTSA-2018-0067-12326 at 10.
                ---------------------------------------------------------------------------
                 Bunch suggested the agencies use a joint model of household vehicle
                holdings and sales that encompasses decisions to purchase new vehicles,
                retain existing ones, or reduce or augment current holdings of vehicles
                of all types and vintages in each period. Manufacturers would modify
                either new vehicle content, prices, or both to produce a supply of new
                vehicles that allowed them each to comply with standards. And,
                subsequently, households and manufacturers would iteratively interact
                until the market reached equilibrium. The model described by Bunch
                would face many of the same issues outlined above. There are
                significant econometric challenges associated with estimating a
                household's decision to buy a new vehicle instead of a used vehicle (of
                some vintage), or to maintain its current set. And integrating such a
                model would
                [[Page 24626]]
                require the agencies to simulate the dynamics of the used vehicle
                market--hundreds of unique nameplates for each of dozens of vintages--
                in order to provide the correct choice set in each simulated year. Such
                a model is beyond the scope of the current analysis.
                 While the agencies believe that these challenges provide a
                reasonable basis for not employing a discrete choice model in today's
                final rule analysis, the agencies also believe they are not
                insurmountable, and that some suitable variant of such models may yet
                be developed for use in future fuel economy and CO2
                emissions rulemakings. The agencies have not abandoned the idea and
                plan to continue experimenting with econometric specifications that
                address heterogeneous consumer preferences in the new vehicle market as
                they further refine the analytical tools used for regulatory analysis.
                 Operating at the level of individual auto and light truck model
                variants--the same level at which compliance is, necessarily,
                simulated--may not be tractable for rulemaking analyses. However,
                market shares for brands and manufacturers within market segments are
                more stable over time--even if the volumes of segments across the
                industry fluctuate. In the 2012 final rule, the agencies' analysis
                showed a new vehicle market where the share of passenger car body
                styles--sedans, coupes, hatchbacks--reached almost 70 percent of the
                new vehicle market by 2025, while light trucks, including many
                crossovers, accounted for the remaining 30 percent. Those results were
                consistent with the assumptions made in 2012, but the combination of
                low fuel prices and decreasing differences in fuel consumption between
                body styles has instead reduced the market share of those body styles
                significantly (only 40% in the MY 2017 fleet), and, thus eroded the
                value of the 2012 analysis to inform current decisions. Including a
                choice model that operated on existing market shares, albeit at a
                higher level of aggregation than specific nameplates, such as brand/
                segment/powertrain, may be able to improve internal consistency with
                the interaction of assumptions about fuel prices and regulatory
                alternatives. The agencies will continue to engage with the academic
                community and other stakeholders to ensure that future work on this
                question improves our analysis of regulatory alternatives.
                3) Scrappage
                a) The Impacts of New Vehicle Fuel Economy Standards on Fleet Turnover
                 Economic literature and theory indicate that the retirement (or
                scrappage) rates of existing vehicles slows when new vehicle fuel
                economy standards increase and cause new vehicle price increases.
                Slower retirement rates result in an older distribution of the on-road
                fleet. Today's on-road fleet is the oldest it has ever been,
                approaching an average of 12 years old.\1657\ Since older vehicles are,
                on average, less safe and less fuel efficient, modeling this reduction
                in the scrappage rates of existing vehicles has important implications.
                As mentioned in the sales section above, past quantitative analyses of
                CO2 and CAFE standards excluded the scrappage effect (though
                the agencies discussed the scrappage effect qualitatively), which could
                have resulted in an overestimate of the benefits of increasing
                standards.
                ---------------------------------------------------------------------------
                 \1657\ Bureau of Transportation Statistics (BTS). ``Average Age
                of Automobiles and Trucks in Operation in the United States.''
                Available at https://www.bts.gov/content/average-age-automobiles-and-trucks-operation-united-states.
                ---------------------------------------------------------------------------
                 For the NPRM, the agencies chose for the first time to model the
                change in existing vehicle retirement rates across regulatory
                alternatives. The agencies used a logistic function to estimate the
                instantaneous scrappage rate for vehicles of different body styles and
                model year vintages using registration data from Polk, the estimated
                durability of specific model year vintages, the prices of new vehicles,
                a measure of the cost of travel for the model year cohort versus new
                vehicles in any given calendar year, and other cyclical macroeconomic
                indicators.\1658\
                ---------------------------------------------------------------------------
                 \1658\ For a more detailed explanation of the NPRM model, see
                PRIA Chapter 8.10.
                ---------------------------------------------------------------------------
                 The agencies received many comments about the NPRM's scrappage
                model. While some commenters objected to the inclusion of a scrappage
                model, most commenters supported the inclusion of a dynamic scrappage
                model as an improvement in the agencies' analysis; these comments are
                discussed in Section VI.C.1.b)(3)(a)(ii). Other commenters raised
                concerns about the specific scrappage models used in the NPRM analysis;
                these are discussed in Section VI.C.1.b)(3)(b). Specifically,
                commenters raised concerns about overfitting in the models, the
                identification strategy, the modeling of new and used vehicle fuel
                economy in general, the exclusion of certain variables, about how the
                agencies captured macroeconomic effects, and about the lack of
                integration with the sales model.
                 The agencies contemplated all of the comments and suggestions made
                by commenters and, in response, have made several changes to final
                rule's model. First, the agencies changed the time-series strategy used
                in the model, as discussed in Section VI.C.1.b)(3)(c)(iii)(a). This
                change allows the agencies to simplify the models significantly,
                addressing commenters' concerns about potential overfitting of the
                model and difficulty of interpreting individual coefficient values
                (discussed in Section CI.C.1.b)(3)(b)(i)). Second, the agencies changed
                the modeling of the durability effect as discussed in Section
                VI.C.1.b)(3)(c)(iii)(c); this change reduces the reliance on the decay
                function and has the added benefit of addressing concerns about
                overfitting and out-of-sample projections discussed in Section
                VI.C.1.b)(3)(b)(i). Third, a portion of anticipated fuel savings from
                increased fuel economy are netted from new vehicle prices--meaning
                consumers are now assumed to value fuel economy at the time of purchase
                to a certain extent--as discussed in Section VI.C.1.b). This change is
                in response to comments discussed in Section VI.C.1.b)(3)(c)(iii)(d)
                and addresses inconsistent treatment of consumer valuation within the
                NPRM's analysis. Finally, the agencies consider the inclusion of
                additional or alternative variables in the scrappage model in response
                to comments discussed in Section VI.C.1.b)(3)(b)(ii). After extensive
                testing, the agencies concluded that these additional variables do not
                improve the model fits or would introduce autocorrelation in the error
                structures (see Sections VI.C.1.b)(3)(c)(iii)(e) and
                VI.C.1.b)(3)(c)(iii)(f) for further discussion). As such, the agencies
                rejected the additional terms suggested by commenters. Input from
                commenters was used to simplify the scrappage model, make it more
                consistent with modeling of new vehicle prices elsewhere in the
                analysis, and improve its predictions for the instantaneous scrappage
                rates of vehicles beyond age 20.
                i) Basis for `The Gruenspecht Effect'
                 Gruenspecht (1981) and (1982) recognized that since fuel economy
                standards affect only new vehicles, any increase in price (net of the
                portion of reduced fuel savings valued by consumers) will increase the
                expected life of used vehicles and reduce the number of new vehicles
                entering the fleet. The effects of differentiated regulation in the
                context of fuel
                [[Page 24627]]
                economy is often deemed the Gruenspecht Effect.\1659\ Jacobsen and van
                Bentham (2015) first quantified the Gruenspecht Effect, or the share of
                new vehicle fuel savings lost to the used vehicle fleet due to delayed
                scrappage, to be between 13 and 16 percent.\1660\
                ---------------------------------------------------------------------------
                 \1659\ Gruenspecht, H. ``Differentiated Regulation: The Case of
                Auto Emissions Standards.'' American Economic Review, Vol. 72(2),
                pp. 328-331 (1982).
                 \1660\ M. Jacobsen and A. van Benthem, ``Vehicle Scrappage and
                Gasoline Policy,'' American Economic Review, Vol. 105, pp. 1312-38
                (2015).
                ---------------------------------------------------------------------------
                 As discussed in the write up of the sales model, fuel economy
                standards increase the cost of acquiring new vehicles, but also improve
                the quality of those vehicles by increasing their fuel economy. The
                CAFE analysis assumes that consumers value 30 months of fuel savings,
                so that the quality-adjusted change in new vehicle prices is the
                increase in regulatory costs less 30 months of fuel savings. As long as
                the quality-adjusted price is positive,\1661\ it becomes more expensive
                for manufacturers to produce vehicles and, as a result, prices of new
                vehicles increase. From a supply and demand perspective, this equates
                to the supply curve for new vehicles moving inwards or to the left and
                a corresponding increase in the equilibrium price and decrease in the
                equilibrium quantity of new vehicles purchased.
                ---------------------------------------------------------------------------
                 \1661\ The quality adjusted price is positive when regulatory
                compliance costs exceed 30 months of fuel savings.
                ---------------------------------------------------------------------------
                 New and used vehicles are substitutes. When the price of a good's
                substitute increases, the demand curve for that good shifts upwards and
                the equilibrium price and quantity supplied also increases. Thus,
                increasing the quality-adjusted price of new vehicles will result in an
                increase in equilibrium price and quantity of used vehicles. Since, by
                definition, used vehicles are not being ``produced'' but rather
                ``supplied'' from the existing fleet, the increase in quantity must
                come via a reduction in their scrappage rates. Practically, when new
                vehicles become more expensive, demand for used vehicles increases (and
                they become more expensive). Because used vehicles are more valuable in
                such circumstances, they are scrapped at a lower rate, and just as
                rising new vehicle prices push marginal prospective buyers into the
                used vehicle market, rising used vehicle prices force marginal
                prospective buyers of used vehicles to acquire older vehicles or
                vehicles with fewer desired attributes.
                ii) Commenter Response to the Inclusion of the Gruenspecht Effect
                (a) Many Commenters Support the Inclusion of the Effect
                 Academic researchers and automakers widely agree with the existence
                and direction of the Gruenspecht Effect. For example, RFF commented,
                ``There's good evidence supporting the scrappage effect.'' \1662\ The
                Auto Alliance stated that the agencies ``made significant strides
                toward improving their modeling of consumer behavior by adding new
                modules to estimate new vehicle sales and in-use vehicle scrappage in
                response to changes to new vehicle prices.'' \1663\ FCA agreed ``that
                an outcome of the current augural stringency of the CAFE/
                [CO2] emission regulations may be a decreasing trend in
                vehicle scrappage rates as consumers delay purchases [. . .] forc[ing]
                consumers to hold their current vehicles for additional time.'' \1664\
                ---------------------------------------------------------------------------
                 \1662\ RFF, Comments EPA NHTSA, NHTSA-2018-0067-11789, at 4.
                 \1663\ Auto Alliance, Full Comment Set, NHTSA-2018-0067-12073,
                at 47.
                 \1664\ FCA, Comments for CAFE-GHG NPRM Final Public Version,
                NHTSA-2018-0067-11943, at 22.
                ---------------------------------------------------------------------------
                 Other commenters agreed with the existence of the effect, but took
                issue with the implications of the combination of the sales and
                scrappage models. Mark Jacobsen stated ``while we agree that the
                scrappage effects we study will mitigate changes in the used fleet, we
                do not believe they could be strong enough to reverse completely the
                direction of change in the used fleet.'' \1665\ Jacobsen's contention
                was echoed by many commenters; the main point was that they believed
                that the prices of both new and used vehicles should be less expensive
                in the NPRM's preferred alternative than the augural standards, and
                that this should, if anything, result in a larger fleet in the NPRM's
                preferred alternative. This issue is further discussed in Section
                (b)(iv) with other comments about integrating the sales and scrappage
                models and the incremental fleet size across alternatives. Here it is
                important to note that this concern does not suggest that a scrappage
                model should not exist, but takes issue with the specific modeling of
                scrappage and/or sales implemented in the NPRM analysis.
                ---------------------------------------------------------------------------
                 \1665\ Mark Jacobsen and Arthur van Benthem, Letter Describing
                Scrappage Effects, NHTSA-2018-0067-7788, at 2.
                ---------------------------------------------------------------------------
                b) Some Commenters Worry About the Shift in Agency Perspective
                 Some commenters argued that the agencies modeling of sales and
                scrappage in the NPRM analysis contradicted previous positions that
                these effects were too uncertain to model. For example, the Center for
                Biological Diversity (CBD) commented:
                 In the 2012 rulemaking for fuel economy and [CO2]
                standards, both NHTSA and EPA stated that analysis of the standards'
                impact on new vehicles sales and on the ``scrappage'' of used
                vehicles was too uncertain to be used in the rulemaking. The
                agencies reiterated this position in their 2016 technical assessment
                of the standards.\1666\
                ---------------------------------------------------------------------------
                 \1666\ CBD, Appendix A, NHTSA-2018-0067-12000, at 171.
                ---------------------------------------------------------------------------
                 They further stated:
                 The agencies have not provided a meaningful rationale or
                justification for the change in position regarding their ability to
                present quantified estimates of the impact of the standards on new
                vehicle sales and the scrappage of used vehicles.\1667\
                ---------------------------------------------------------------------------
                 \1667\ CBD, Appendix A, NHTSA-2018-0067-12000, at 178.
                 To respond to these comments, it is useful to look at the reasons
                the agencies gave for not considering fleet turnover effects on pages
                ---------------------------------------------------------------------------
                845-46 of the 2012 rulemaking:
                 If the value of fuel savings resulting from improved fuel
                efficiency to the typical potential buyer of a new vehicle outweighs
                the average increase in new models' prices, sales of new vehicles
                will rise, while scrappage rates of used vehicles will increase
                slightly. This will cause the ``turnover'' of the vehicle fleet--
                that is, the retirement of used vehicles and their replacement by
                new models--to accelerate slightly, thus accentuating the
                anticipated effect of the rule on fleet-wide fuel consumption and
                CO2 emissions. However, if potential buyers value future
                fuel savings resulting from the increased fuel efficiency of new
                models at less than the increase in their average selling price,
                sales of new vehicles will decline, as will the rate at which used
                vehicles are retired from service. This effect will slow the
                replacement of used vehicles by new models, and thus partly offset
                the anticipated effects of the final rules on fuel use and
                emissions.
                 Because the agencies are uncertain about how the value of
                projected fuel savings from the final rules to potential buyers will
                compare to their estimates of increases in new vehicle prices, we
                have not attempted to estimate explicitly the effects of the rule on
                scrappage of older vehicles and the turnover of the vehicle
                fleet.\1668\
                ---------------------------------------------------------------------------
                 \1668\ 77 FR 62,623, 63,112-13 (emphasis added).
                The agencies' reason for not modeling the fleet turnover effects in
                prior rulemakings was not uncertainty about the direction or impact of
                vehicle prices on sales or scrappage rates, but rather uncertainty
                about how consumers value fuel savings. The agencies now have
                sufficient knowledge regarding the amount of fuel savings consumers are
                assumed to value at the time they purchase new vehicles and make these
                [[Page 24628]]
                assumptions in the technology application simulation. With this
                assumption, it becomes possible to model the fleet turnover effects,
                including the scrappage effect.
                c) Some Commenters Think the Effects Are Uncertain
                 Other commenters argue that the sales and scrappage effects are too
                uncertain to include in a rulemaking analysis. For example, CBD argued
                that ``the models are attempting to evaluate the small and uncertain
                effects of changes in vehicle standards on certain dynamics--vehicle
                sales, scrappage rates, and vehicle usage--which are largely determined
                by much stronger forces, such as the state of the economy.'' \1669\
                ---------------------------------------------------------------------------
                 \1669\ CBD, Appendix A, NHTSA-2018-0067-12000, at 177.
                ---------------------------------------------------------------------------
                 The agencies agree that there is uncertainty around the magnitude
                of the sales and scrappage response, but do not agree that sign of
                either effect is uncertain. Importantly, excluding modeling of the
                sales and scrappage effects would only make sense if there was a
                legitimate existential concern--the sales and scrappage effects are
                founded in very basic economic theory, as noted above, in Section
                VI.C.1.b)(3)(a)(i). Furthermore, the agencies believe that assessing
                the magnitudes of the sales and scrappage effects is a tractable task
                for researchers and sufficient data exists to quantify these effects.
                Thus, excluding these effects would be a serious omission that limits
                accurate accounting of the costs and benefits of fuel economy
                standards. Other stakeholders commented that the NPRM analysis did not
                thoroughly consider the uncertainty around the magnitudes of the sales
                and scrappage responses. These comments and the agencies response is
                discussed in Section VI.C.1.b)(3)(b)(i), below. The agencies believe it
                is better to consider a range of the scrappage and sales response to
                address concerns about uncertainty, and that excluding them would be
                inappropriate.\1670\ The agencies did just that with the proposal
                through sensitivity analyses--including seeking comment and having the
                scrappage modeling peer reviewed--and continue to do so for the final
                rule.
                ---------------------------------------------------------------------------
                 \1670\ See, e.g. Ctr. for Biological Diversity v. Nat'l Highway
                Traffic Safety Admin., 538 F.3d 1172, 1203 (9th Cir. 2008), (finding
                that NHTSA inappropriately assigned no value to reducing carbon
                emissions when the value for doing so was ``certainly not zero.'').
                ---------------------------------------------------------------------------
                b) Summary of Notice, Request for Comments, and the Agencies' Response
                 The comments related to the scrappage model are summarized here
                into five major categories: Overfitting and identification strategies,
                modeling fuel economy and new vehicle prices, consideration of other
                additional variables, integration with sales or VMT, and evaluations of
                associated costs and benefits due to changes in scrappage rates within
                the CAFE model. Specific modeling decisions the agencies have made or
                considered in response to the public comments summarized in this
                section are discussed in Sections VI.C.1.b)(3)(c)(ii)(d) and
                VI.C.1.b)(3)(c)(iii).
                i) Overfitting and Identification Strategy
                 Several commenters argued that the NPRM scrappage model did not
                have a clear identification strategy, or that the model over-fit the
                data. These commenters suggest that the NPRM model may not capture a
                causal relationship, but picks up other correlation or noise within the
                data. This section outlines the specific claims made by commenters.
                a) Overfitting and the Use of Lagged and Interactions Terms
                 Several commenters argued that the results presented in the NPRM
                could be driven by the specific structure of the price effect used in
                the scrappage models that were implemented into the CAFE Model. IPI,
                California States et. al., CARB, and other commenters suggested that
                the NPRM model is over-fit. CARB outlined its argument that the
                agencies overfit the data in the following passage:
                 [T]he model appears to be significantly overfit and to suffer
                from multicollinearity. An overfit model means that the model is
                able to precisely replicate past trends, but only through the use of
                too many variables. An overfit model fits the data too well, fitting
                the noise or errors in the data in addition to the underlying
                relationships between the variables of interest. Because an overfit
                model also fits the noise and errors of the data, the out-of-sample
                predictions are unreliable. Comments from Jeremy Michalek and Katie
                Whitefoot suggest that choice of specification of the scrappage
                model could result in substantially different predictions, and that
                the agencies should make only those claims that are robust to
                reasonable variations in the model specifications.\1671\
                ---------------------------------------------------------------------------
                 \1671\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 245.
                 The agencies agree that it is important that the scrappage model
                results are robust across those specifications that meet a set of
                econometric criteria (these criteria are discussed further in Section
                VI.C.1.b)(3)(c)(iii)). However, the agencies acknowledge that the NPRM
                could have provided further evidence that the specification did not
                drive the results. In the analysis for the final rule the agencies have
                presented more than one specification of the price effect as evidence
                that the specification chosen here does not drive the results of the
                analysis. Further, claims that the specification of the scrappage
                response in the NPRM is inconsistent with economic theory are false.
                 Theoretically, changes in average new prices may have longer-term
                trends that can be picked up by including lagged terms, and/or be non-
                linear with age, so that vehicles of different ages have different
                elasticities of scrappage (relative to changes in average new vehicle
                prices). Further, sometimes the effect of one independent variable on
                the dependent variable depends on the magnitude of another independent
                variable--this is called an interaction effect. Regression analysis can
                capture these interaction effects by defining a new variable using some
                combination of independent variables.\1672\ It is necessary to retain
                such interaction terms when doing so.\1673\ For example, it is not
                obvious that the elasticities of scrappage rates to changes in new
                vehicle prices should be constant for all vehicle ages, or put another
                way, the older a vehicle is, the higher likelihood the vehicle will be
                scrapped instead of being retained or resold.
                ---------------------------------------------------------------------------
                 \1672\ Davis, J. B., Statistics using SAS enterprise guide.
                Cary, NC: SAS Institute, pp. 411-415 (2012).
                 \1673\ As explained in more detail in Section
                I.A.1.a)(1)(a)(ii)(a), below, the agencies perform several
                sensitivity analyses to ensure the model captures the correct impact
                of interactive effects.
                ---------------------------------------------------------------------------
                 Michalek and Whitefoot, Honda, and other commenters, argued that
                the fact that some of the interaction terms were not statistically
                significant was evidence that the response measured is uncertain. CBD
                in particular claimed that the ``scrappage model is poorly constructed,
                and its results are not statistically significant.''
                 In response to such comments, it is important to note that when
                interaction terms are included, the significance of the overall effect
                of a variable should be tested by performing a restricted F-test, which
                simultaneously tests that all coefficients of the variable of interest
                are jointly indistinguishable from zero. The insignificance of one term
                of the interaction does not imply that the effect is indistinguishable
                from zero.\1674\
                ---------------------------------------------------------------------------
                 \1674\ Davis, J. B., Statistics using SAS enterprise guide.
                Cary, NC: SAS Institute, pp. 411-415 (2012).
                ---------------------------------------------------------------------------
                 Commenters also noted the lagged terms and age interactions make
                the new vehicle price effect difficult to interpret. IPI argued that
                ``[t]he inclusion of interaction variables make it very difficult to
                evaluate the results of the regression for an individual variable
                [[Page 24629]]
                of interest.'' Michalek and Whitefoot suggested ``using a Monte Carlo
                analysis to understand the distribution of scrappage outcomes implied
                by uncertainty of the value of the coefficients in the model regression
                and reporting 95% confidence intervals.''
                 We agree that the inclusion of lags and age interactions of new
                vehicle prices can make interpreting the sign and magnitude of the
                price effect difficult. It also makes it difficult to use the
                confidence intervals on the coefficients as a way to capture
                uncertainty, since the interaction variables are jointly estimated.
                Thus, for the NPRM analysis, the agencies could not independently
                sample each coefficient from the confidence intervals and perform a
                Monte Carlo analysis.
                 While the agencies think that the inclusion of lags and interaction
                terms is theoretically plausible, in response to commenter and peer
                reviewer concerns about overfitting and the difficulty of interpreting
                coefficients, the agencies reconsidered the time series approach. The
                agencies found that new vehicle prices are integrated to order one and
                that the dependent variable is stationary (as discussed in Section
                VI.C.1.b)(3)(c)(iii)(a)). It is therefore sufficient to fit the first
                difference of new vehicle prices within the models. Thus, the agencies
                have simplified the central model of the response of scrappage rates to
                changes in new vehicle prices to exclude lags of the effect. The
                agencies further simplified the central scrappage models to exclude
                interaction of new vehicle prices and vehicle age; this allows the
                agencies to take the 95 percent confidence intervals as a low and high
                range for the magnitude of the price effect for the sensitivity
                analysis. The agencies also include a sensitivity analysis which
                includes interaction terms between new vehicle price and vehicle age to
                allow the elasticity of scrappage to changes in new vehicle price to
                vary by vehicle age.
                 Commenters also noted that the model did not perform well for
                vehicles beyond age 20. The agencies noted in the PRIA that the Polk
                dataset for older vehicles was limited and likely led to the inability
                to estimate the scrappage rates for older ages.\1675\
                ---------------------------------------------------------------------------
                 \1675\ FR, Vol 83, No. 165, August 24, 2018, p.43097.
                ---------------------------------------------------------------------------
                 The final rule dataset includes almost 30 percent more data for
                vehicles fifteen years or older than the NPRM, which improves estimates
                of the scrappage rate of vehicles aged 20 to 30 (see Table VI-158). The
                agencies are still unable to capture the scrappage trends for vehicles
                over 30, as the dataset is still limited for the oldest ages of
                vehicles, and still rely on the decay function used in the NPRM for
                vehicles over the age of 30. The limited data explains the inability to
                predict the scrappage rates for older vehicles. However, including
                model year fixed effects and including the share of the initial cohort
                remaining does improve predictions of the final share remaining in the
                final rule models. These changes are discussed in Section
                VI.D.1.b)(c)(i)(c).
                b) Reduced Form and Endogenous Prices
                 California States et. al., CARB, EDF, IPI and academic commenters
                expressed concerns that the NPRM analysis fit a reduced form of the
                scrappage model, rather than a structural model. In other words,
                instead of explicitly modeling new and used vehicle prices in
                equilibrium under different regulatory alternatives and applying a
                measurement of the elasticity of scrappage to the resulting used
                vehicle prices, the agencies modeled the elasticity of scrappage from
                changes to new vehicle prices. For example, California States et. al.,
                argued that the model ``does not link the new and used vehicle markets
                as required by economic theory, nor does it attempt to measure used
                vehicle prices, which form the basis of scrappage theory.''
                 While the agencies recognize that there are certain advantages to a
                structural model, they disagree that the sales of new and used vehicles
                must be modeled simultaneously. The agencies do link the new and used
                car markets by including new vehicle prices as an independent variable
                in scrappage regression equation. However, it would be inappropriate to
                include used vehicle prices in this equation due to endogeneity
                concerns. A change in used vehicle prices may change scrappage rates,
                but also an exogenous shock to scrappage rates may cause used car
                prices to vary.
                 Furthermore, the agencies are unaware of a viable structural model
                for the scrappage effect. The agencies performed an extensive review of
                economic of literature, both before creating the scrappage model for
                the proposal and revising it for the final rule, but were unable to
                find such a model or any insights on how to construct one. The agencies
                note that commenters did not suggest a structural model that the
                agencies should use or give any indication of whether such a model
                exists.
                 In order to understand why such a model is difficult to construct,
                it is important to understand what a structural model of the sales and
                scrappage responses would entail. A hypothetical structural model for
                the new vehicle market can be represented by the following simultaneous
                demand and supply equations:
                DNew = [beta]0 + [beta]1 * PNew +
                [beta]2 * PUsed + [beta]3 * PTransit +
                [beta]4 * Income + [beta]5 * Households
                SNew = [beta]6 + [beta]7 * PNew +
                [beta]8 * Production CostNew
                The demand equation for new vehicles in a given year is determined by
                the annual price of owning and operating new vehicles, the annual price
                of owning and operating used vehicles, the annual price of other
                substitutes, average household income, and the number of households.
                The supply equation is made up of the average price of new vehicles and
                the average cost to produce them.
                 As noted in the sales model write up, reducing required fuel
                economy stringency reduces the cost of producing new vehicles, and
                shifts the supply curve to the right. This results in an increase in
                the quantity supplied of new vehicles.
                 The structural model for the used vehicle market can be represented
                by the following simultaneous demand and supply equations:
                DUsed = [gamma]0 + [gamma]1 * PUsed +
                [gamma]2 * PNew + [gamma]3 * PTransit +
                [gamma]4 * Income + [gamma]5 * Households
                SUsed = [gamma]6 + [gamma]7 * PUsed +
                [gamma]8 * Maint RepairUsed + [gamma]9 * Scrap
                ValueUsed
                 The aggregate demand equation for used vehicles is determined by
                the price of owning and operating used vehicles, the price of owning
                and operating new vehicles, the price of other transit substitutes,
                average income, and the number of households. The supply curve equation
                for used vehicles is determined by the price of used vehicles, the cost
                to repair and maintain them in service, and the opportunity cost of the
                scrappage value of doing so. Relaxing new vehicle standards reduces new
                vehicle prices and shifts the demand curve for used vehicles downward,
                which reduces demand for used vehicles and the equilibrium price and
                quantity of used vehicles, and increases the annual scrappage rate.
                 Modeling the structural equations would require that the agencies
                predict new and used vehicle prices in equilibrium, allowing prices of
                new and used vehicles be determined simultaneously from estimates of
                the supply and demand curves for each market. As CARB stated in the
                following comment, new and used vehicle prices are endogenous--the
                equilibrium prices of each good are simultaneous:
                [[Page 24630]]
                 Because both scrappage rates and new vehicle prices may
                influence one another, the Agencies would need to utilize different
                statistical techniques to credibly identify the impact of new
                vehicle prices on scrappage rates. For example, the Agencies would
                need to identify an instrumental variable that impacts new vehicle
                price but that does not impact the scrappage rate. Models that
                suffer from endogeneity problems will have biased estimates. In
                other words, the estimates from these models cannot be used to
                inform policy, because they do not actually tell us how new vehicle
                prices impact scrappage.
                 CARB suggested a way to correct for endogeneity: Using an
                instrumental variable in a two-stage least squares methodology where
                the instrumental variable is correlated with new vehicle prices, but
                not scrappage rates.\1676\ The agencies could also address the
                potential for endogeneity in two steps: First, they could model the
                impacts of exogenous changes in new vehicle prices on used vehicle
                prices, and second, they could model the impacts of exogenous changes
                in used prices on scrappage rates. Implementing the first step would
                require using an instrumental variable to isolate exogenous shifts to
                the new vehicle supply curve, and then using the predicted values of
                new vehicle prices to model changes in prices for used vehicles of all
                ages. Because prices and scrappage rates are jointly determined in the
                market for used vehicles, predicting the elasticity of scrappage with
                respect to price variation also requires isolating exogenous changes in
                used vehicle price via the use of an instrumental variable.
                ---------------------------------------------------------------------------
                 \1676\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 244.
                ---------------------------------------------------------------------------
                 There is one literature example that approaches the structural
                model that some commenters would like the agencies to implement.
                Jacobsen and van Bentham \1677\ developed a structural model that
                simultaneously solves for prices that clear new and used vehicle
                supplies, and then applies an elasticity of scrappage measure that
                corrects for potential endogeneity of used vehicle values and scrappage
                rates using an instrumental variable methodology. Specifically, they
                use changes in fuel prices as an instrumental variable; changes in fuel
                prices shift the demand for different vehicle models, but not the cost
                of supplying them. This should capture exogenous changes in value, so
                that an exogenous measure of the scrappage elasticity can be isolated
                in the second stage of the two-staged least squares method.
                ---------------------------------------------------------------------------
                 \1677\ M. Jacobsen and A. van Benthem, ``Vehicle Scrappage and
                Gasoline Policy,'' American Economic Review, Vol. 105, pp. pp. 1312-
                38 (2015).
                ---------------------------------------------------------------------------
                 While Jacobsen and van Bentham are able to correct for potential
                endogeneity between used vehicle values and their scrappage rates,
                their structural model to set new and used vehicle values
                simultaneously makes some presumptions that the agencies are not
                comfortable making. First, they calibrate their constant elasticity of
                substitution (CES) utility function using 1999 data from GM's internal
                model. This type of model would estimate elasticities of specific
                vehicle models and require a pricing strategy other than allotting all
                additional technology costs to the vehicle models to which they are
                applied. The agencies have avoided a pricing strategy for the reasons
                cited in the sales model write up. Second, by relying on GM's internal
                model, Jacobsen and van Bentham used elasticities calculated using only
                1999 data of the GM fleet. The agencies do not expect that elasticities
                estimated from 20-year old data from a single OEM's portfolio of
                vehicles would translate to the entirety of the current vehicle
                fleet.\1678\ Finally, Jacobsen and van Bentham represent total vehicle
                demand of a representative consumer from a composite vehicle. This
                approach precludes the realistic consideration that a household may
                prefer two used vehicles over one new vehicle, which is accounted for
                in the agencies' functional equations.
                ---------------------------------------------------------------------------
                 \1678\ Kleit, Andrew N., 2004. ``Impacts of Long-Range Increases
                in the Corporate Average Fuel Economy (CAFE) Standard.'' Economic
                Inquiry 42:279-94.
                ---------------------------------------------------------------------------
                 Jacobsen's and A. van Benthem's model is not a household level
                choice model, and is not meant to determine fleet size, as noted in
                their comment:
                 In summary, while the Jacobsen and van Benthem (2015) paper
                cannot inform by how much the total vehicle fleet would expand under
                a CAFE rollback (since we do not estimate by how much it shrinks
                under CAFE), all the evidence and economic logic points to a larger
                total vehicle fleet under a rollback, at odds with NHTSA's fleet
                turnover model.\1679\
                ---------------------------------------------------------------------------
                 \1679\ Mark Jacobsen and Arthur van Benthem, Letter Describing
                Scrappage Effects, NHTSA-2018-0067-7788, at 2.
                 The agencies agree that the long-term fleet should be smaller in
                the augural case, as fewer new vehicles flow into the used car market
                (because of lower sales), but do think it is plausible that in the
                short term the fleet size could increase under augural standards if in
                some cases consumers substitute two used vehicles for one new one or
                choose to retain an additional vehicle on the margin because the higher
                value makes doing so a more reasonable investment (at the annual
                level). This sort of outcome is not possible with the Jacobsen and van
                Bentham 2015 model, because the overall demand for vehicles is set by
                the annual rent prices of a composite vehicle. The updates to the
                scrappage model for the final rule are consistent with this view, but
                do show a smaller fleet size under the augural standards relative to
                the proposal. This is discussed further in Section
                VI.C.1.b)(3)(b)(iv)(b).
                 Fitting the reduced form equation requires that endogenous
                variables are excluded from the model to avoid biased coefficients. As
                a result, used vehicle prices were omitted by design, because used
                vehicle prices and scrappage rates are endogenous.\1680\ Some
                commenters argue that new vehicle prices and scrappage rates are also
                endogenous; CARB argued that ``the model tries to rely solely on new
                vehicle prices to predict scrappage rates without realizing or
                controlling for the fact that scrappage rates may also affect new
                vehicle prices.'' \1681\
                ---------------------------------------------------------------------------
                 \1680\ Hill, R. C., Griffiths, W. E., & Lim, G. C. Chapter 11:
                Simultaneous Equation Models. In Principles of Econometrics (3rd
                ed., pp. 303-24). Hoboken, NJ: John Wiley & Sons, Inc. (2008).
                 \1681\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 244.
                ---------------------------------------------------------------------------
                 Commenters provided neither evidence nor an explanation as to why
                there may be some degree of ``reverse causality'' or endogeneity
                between new vehicle prices and scrappage rates. Two potential
                econometric explanations for such endogeneity could be that: (1) These
                variables are jointly or simultaneously determined, so each one
                influences the other; or (2) the model omitted a variable that causes
                covariance between new vehicle prices and scrappage rates. The agencies
                believe the first source of potential endogeneity can be dismissed, as
                any causal relationship between scrappage rates and new vehicle prices
                would necessarily flow through the used car market, which are
                substitute products for new vehicles, and specifically through the
                mechanism of used car prices. For example, an exogenous shock to
                scrappage rates might cause the supply curve in the market for the
                lowest-price used vehicles to shift, and the resulting change in their
                price might cause price responses in higher-price segments of the used
                vehicle market, which in turn might eventually filter up to the new
                vehicle market and affect the prices for new vehicles. This chain of
                events suggests omitted variable bias might be a concern, rather than
                simultaneity.
                 The agencies believe that supply and demand for used vehicles (or
                some measure of their interaction, such as
                [[Page 24631]]
                used vehicle prices) are the most likely sources of any potential
                omitted variable bias. If an omitted variable is causing bias in the
                estimates, then the bias is observable. Whether endogeneity--through an
                omitted variable--is causing bias is an empirical question, which can
                be answered by conducting common empirical test--the Durbin-Wu-Hausman
                test. The Durbin-Wu-Hausman test requires identifying a suitable
                instrument(s)--a variable--that is correlated with new vehicle prices
                but not with scrappage rates, so any effect exerted on scrappage rates
                by the instrument will occur through their association with prices for
                new vehicles.\1682\ The agencies tested a few alternative approaches,
                which included using the change in new vehicle prices during the
                preceding time period and the level of prices during the current period
                as instrumental variables for the change in prices during the current
                period, and another test using the current-period growth rate in GDP as
                an instrument for the change in new vehicle prices during the current
                period. Each of these tests fails to reject the null hypothesis that no
                endogeneity is present at the 0.05 level of significance.
                ---------------------------------------------------------------------------
                 \1682\ For a conceptual overview of this test, see https://www.statisticshowto.datasciencecentral.com/hausman-test/. For a more
                detailed description of the logic underlying the test and how to
                interpret its results, see http://personal.rhul.ac.uk/uhte/006/ec2203/Lecture%2015_IVestimation.pdf.
                ---------------------------------------------------------------------------
                 For both theoretical and empirical reasons, the agencies are
                therefore skeptical about both the likelihood that scrappage rates will
                affect prices for new vehicles, and the extent to which they might do
                so. The agencies find the theoretical underpinnings for endogeneity to
                be tenuous, and believe the empirical evidence suggests such
                endogeneity is not an issue for today's analysis.
                 The agencies chose not to fit a model predicting used vehicle
                prices directly from new vehicle prices for the proposal because
                currently-available time-series data on the prices of used vehicles of
                a given vintage going back to 1975 is limited. EDF cited the lack of
                available data as the reason not to fit the structural model:
                 In the absence of any data or analysis, NHTSA did not describe
                the extent to which changes in new vehicle prices affect used
                vehicle prices of varying age, condition, etc. \1683\
                ---------------------------------------------------------------------------
                 \1683\ EDF, Appendix B, NHTSA-2018-0067-12108, at 56.
                The agencies note that acquisition, assembly, and cleaning of a
                nationally representative database for calendar years 1974 to 2017 on
                used vehicle prices by vintage from Kelly Blue Book (or a similar
                source) would take months to years, and would push the final rule
                beyond the necessary April 2020 lead time requirement to set MY 2022
                standards. Kelly Blue Book data is readily searchable for current
                prices, but without a time series of used vehicle prices the data
                cannot be used to answer the causal relationship of changes in used
                vehicle prices over time on vehicle retirement rates. Even assembling a
                nationally representative sample of used vehicle prices by vintage
                would be a major undertaking. This is not to suggest that doing so is
                out of scope for future analyses; the agencies plan to consider further
                the possibility of conducting additional analysis on the relationship
                between new and used vehicle prices.
                 The agencies considered use of the Consumer Expenditure Survey
                (CEX), which has reported vehicle transaction data annually since
                1984.\1684\ However, the sample of used vehicle purchase prices aged
                twenty and older is severely limited. For vehicles purchased between
                1996 and 2017, the average number of transaction prices reported for
                vehicles aged 20 is 58, and for vehicles aged 25 is 18. Any computation
                of average used vehicle prices from such a small sample would not be
                reliable, and in fact, would be quite noisy. The agencies do not think
                that estimates of a structural model based on such limited sampling
                would improve the prediction of the scrappage effects over use of the
                reduced form equation.
                ---------------------------------------------------------------------------
                 \1684\ U.S. Bureau of Labor Statistics. (2016). Consumer
                Expenditures and Income: Collections & Data Sources. Retrieved from
                https://www.bls.gov/opub/hom/cex/data.htm.
                ---------------------------------------------------------------------------
                 EDF argued that modeling the impact of changes in new vehicle
                prices directly on used vehicle scrappage may not capture the fact that
                changes in used vehicle prices impact vintages differently. Further,
                they argue that if new and used vehicle prices change by the same
                proportion, the effect will have a very small impact on the prices of
                the oldest used vehicles. They argue that these small changes are not
                enough to change the scrappage decisions:
                 Given that vehicles can sell for as little as a couple of
                hundred dollars and new vehicle prices average over $30,000, used
                vehicle prices can be as little as 1% of that of a new vehicle.
                Given that the largest increase in new vehicle prices projected by
                NHTSA in the NPRM is less than $3000, and assuming that its effect
                on used vehicle prices is likely to be roughly proportional to
                current relative prices, this might mean that the value of a very
                old vehicle or one in poor condition might only increase by $30
                (decline by $30 under the proposal). It is difficult to see how such
                a change in value would have a measurable impact on scrappage. Of
                course, the impact of an increase in new vehicle prices on used
                vehicle prices might be more or less than proportional to their
                current relative values. However, NHTSA has done nothing to show
                which might be the case. The probability of any realistic change in
                used vehicle prices to induce the scrappage of used vehicles is
                still a complete mystery.\1685\
                ---------------------------------------------------------------------------
                 \1685\ EDF, Appendix B, NHTSA-2018-0067-12108, at 52.
                 However, the age interaction on the new vehicle price effect allows
                that the elasticity of scrappage to changes in new vehicle prices may
                not be constant for all ages. Allowing the scrappage elasticity to new
                vehicle prices to vary by age incorporates the fact that the elasticity
                of scrappage of used vehicles and the cross-price elasticity of used
                vehicle demand to new vehicle prices may not be constant with age. At
                some point, the thirty-dollar increase EDF cited could be the
                difference in keeping a marginally used vehicle on the road; it would
                be a 10 percent increase in the price of a used vehicle, and may cover
                State registration fees on a marginally scrapped vehicle.
                (c) Time Series
                 The scrappage model utilizes panel data. Panel data observes
                multiple individuals or cohorts over time. The data employed by the
                scrappage model observes the scrappage rates of individual model year
                cohorts between successive calendar years. The model allows for the
                isolation of trends over time and across individuals.\1686\ Since the
                scrappage model uses aggregate model year cohorts to estimate scrappage
                rates by age and time-dependent variables (new vehicle prices, fuel
                prices, GDP growth rate, etc.) panel data is necessary to estimate the
                model. A major challenge to using panel data is that the data structure
                requires consideration of potential violations of econometric
                assumptions necessary for consistent and unbiased estimates of
                coefficients both across the cross-section and along the time
                dimension. The cross-section of the scrappage data introduces potential
                heterogeneity bias--where model year cohorts may have cohort-specific
                scrappage patterns. \1687\ Another way to put this is that each model
                year may have its own inherent durability. The NPRM captured this
                potential bias by including model year as a continuous variable, but
                the model amended for the final rule includes the more traditional
                [[Page 24632]]
                individual fixed effects. This is discussed in Section
                VI.C.1.b)(3)(c)(iii)(a). The time dimension of a panel introduces a set
                of potential econometric concerns present in time series analysis. The
                agencies considered potential autocorrelation in the error structures
                and included lags of the dependent and specific independent variables
                to correct for it; this is not an uncommon practice in dynamic panel
                models.\1688\ Some commenters argued that time series approaches were
                not appropriate in the scrappage model at all. CARB stated the
                following:
                ---------------------------------------------------------------------------
                 \1686\ Cambridge University Press. (1989). Analysis of Panel
                Data. New York, NY.
                 \1687\ Cambridge University Press. (1989). Analysis of Panel
                Data. New York, NY.
                 \1688\ Bun, M. J. G., & Sarafidis, V. (2015). Dynamic Panel Data
                Models. In The Oxford Handbook of Panel Data (pp. 76-110). New York,
                NY: Oxford University Press.
                 Time-series analysis for modeling scrappage is also
                inappropriate for the same reasons as it was for the new vehicle
                sales model--particularly because time-series analysis does not
                capture structural changes, which the scrappage model seeks to
                illustrate.\1689\
                ---------------------------------------------------------------------------
                 \1689\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 243.
                 The agencies disagree with CARB's assessment. The potential
                scrappage effect can only be measured with a time series dimension; the
                agencies are interested in how changes in new vehicle prices over time
                impact the retirement rate of the on-road fleet over time. In order to
                isolate this effect, the agencies need multi-period data on the
                scrappage rates of used vehicles and prices of new vehicles.
                 The literature on vehicle scrappage rates utilizes panel data, but
                most research has ignored potential autocorrelation issues caused by
                the structural properties of independent variables that vary along the
                time dimension. With the NPRM analysis, the agencies found evidence of
                auto-correlated errors, which were corrected by including three lagged
                terms of the dependent variable.\1690\ While in a pure time series
                analysis, this can be an appropriate methodology to account for
                autocorrelation in the error structure; estimates of the coefficients
                of the lagged dependent variable are biased downwards when applied in
                fixed or random effects panel models. The reason for this is that the
                constant individual specific terms are correlated with the lagged
                dependent variable (by definition, since the individual specific terms
                are constant for all time periods, including the previous period),
                creating a bias in the estimate of the coefficient on the lagged
                dependent variable, and potentially other measures.\1691\ The eponymous
                bias was first discussed in a paper written by Nickell in 1982.\1692\
                There is an increasing body of work developing estimators built
                specifically for dynamic panel data (DPD), or panel data where there is
                an autoregressive component to the data-generating process. In other
                words, the previous value of the dependent variable impacts the current
                value.
                ---------------------------------------------------------------------------
                 \1690\ FR, Vol 83, No. 165, August 24, 2018, p.43097.
                 \1691\ Allison, P., Don't Put Lagged Dependent Variables in
                Mixed Models, (2015, June 2). Retrieved June 1, 2019, from https://statisticalhorizons.com/lagged-dependent-variables.
                 \1692\ Nickell, Stephen. ``Biases in Dynamic Models with Fixed
                Effects.'' Econometrica, vol. 49, no. 6, 1981, pp. 1417-26. JSTOR,
                www.jstor.org/stable/1911408.
                ---------------------------------------------------------------------------
                 Further research into this literature (discussed above), comments
                on the NPRM, and peer review comments prompted the agencies to
                reconsider the approach developed for the NPRM. The NPRM analysis did
                not use fixed effects for specific model years, but instead imposed a
                parametric logarithmic relationship of successive model years. This
                parametric model year term will still result in biased estimates of the
                lagged dependent variable because it also does not vary over time for
                the same model year, and is therefore correlated with the
                autoregressive term. Since the autoregressive term carries through
                effects from the previous period (the new vehicle price effect), this
                will also bias the predicted Gruenspecht effect in the NPRM model.
                Updates to the model used for the final rule correct this issue by more
                deliberately considering the time series properties of both the
                dependent and independent variables.
                 In reconsidering the appropriate way to address the time series
                properties of the scrappage model, the agencies first consider the
                stationarity of dependent and independent variables. This was suggested
                in James Sallee's peer review:
                 In contrast to the new vehicle sales regression reported in the
                PRIA's section 8.6, the discussion of the scrappage regressions does
                not include any discussion of the time series properties of the
                estimators. It is important to test for non-stationarity, for
                example.\1693\
                ---------------------------------------------------------------------------
                 \1693\ CAFE Model Peer Review (Report No. DOT HS 812 590).
                Washington, DC--National Highway Traffic Safety Administration, B-
                64.
                Importantly, the agencies find that the instantaneous scrappage rate is
                stationary, so that there is no longer term information in the
                scrappage rates to recover with an autoregressive term. This means that
                a DPD model is not necessary to correct for potential autocorrelation
                in the model. This also implies that the autocorrelation in the errors
                is a result of non-stationarity in some or all of the regressors, and
                not the independent variable. The solution to this problem is to
                identify the order of integration of each regressor and difference
                until each is non-stationary. Table VI-160 in Section
                VI.C.1.b)(3)(c)(iii)(a) shows the order of integration of variables
                considered in the scrappage modelling.
                (ii) Modeling Fuel Economy
                (a) Counterintuitive Signs
                 In the NPRM analysis, the agencies controlled for the changes in
                the relative fuel economy of new and used vehicles by including the
                cost per mile of travel in the current period and the previous period
                for both new vehicles and the model year cohort whose scrappage is
                being predicted. This allowed fuel prices to alter the scrappage rates
                of existing vehicles, meaning model year cohorts with lower-than-
                average fuel economies were impacted by increases to fuel prices to a
                greater extent than cohorts with higher-than-average average fuel
                economies. It also allowed increases in the fuel economy of new
                vehicles to impact the scrappage rates of existing vehicles; the idea
                is that when new vehicles have a higher average fuel economy, holding
                price constant, the demand for new vehicles should increase relative to
                used vehicles, and scrappage rates should increase. While this was a
                plausible way of controlling for changes in the relative fuel cost per
                mile of usage of new and used vehicles, the agencies noted in the NPRM
                that some of the signs on new vehicle cost per mile were
                counterintuitive, so that increases in the average new vehicle fuel
                economy of certain body styles actually increased the scrappage rates
                of existing vehicles.
                 IPI, CARB, CBD, Natural Resources Defense Council (NRDC), and other
                commenters argued that these results were driven more by modeling
                decisions than by actual relationships within the data. NRDC suggested
                that the conclusions from the NPRM model should be treated with
                suspicion until validated by further research:
                 [A]n increase in fuel price for a given level of fuel economy
                results in longer vehicle retention even though operational costs
                per mile increase. While it is not possible to rationalize this
                response without significant additional research, it is indicative
                of the fact that the algorithm response functions may not be
                properly defined.\1694\
                ---------------------------------------------------------------------------
                 \1694\ NRDC, Attachment 3: CAFE Model Activity Review, NHTSA-
                2018-0067-11723, at 20.
                 The agencies agree that the results were counter-intuitive--having
                identified this issue in the NPRM and
                [[Page 24633]]
                specifically seeking comment on the matter--and considered multiple
                alternative methods of capturing the fuel economy improvements of new
                vehicles within the scrappage model in response to comments. Among the
                changes considered were alternate forms of modeling the form of new
                ---------------------------------------------------------------------------
                vehicle fuel economy, as suggested by IPI:
                 A paper by Shanjun Li et al., provides a useful example of how
                the agencies could include fuel efficiency in their regression
                without raising the econometric concerns that may be leading to
                their nonsensical results. Li et al. include fuel price and vehicle
                fuel efficiency (gallons per mile) of used vehicles as well as a
                variable that captures the interaction of fuel efficiency of used
                vehicles and fuel price in their regression as explanatory
                variables. Unlike the agencies' model, the regression analysis used
                in the Li et al. paper found results that are consistent with
                economic theory: A decrease in overall demand for vehicles and an
                increase in demand for more fuel-efficient cars.\1695\
                ---------------------------------------------------------------------------
                 \1695\ IPI, Policy Integrity Comments: NHTSA Final--Appendix,
                NHTSA-2018-0067-12213, at 72.
                 The NPRM included changes in new vehicle cost-per-mile, but did not
                include separate variables for fuel prices or fuel economy. This could
                potentially have conflated changes in the cost-per-mile of new vehicles
                from changes in fuel prices and changes in new vehicle fuel economy.
                The agencies considered including changes in fuel prices and new
                vehicle fuel economy as separate measures, as suggested in IPI's
                comment above, but opted for a different method of addressing the
                concern of how to include changes to new vehicle fuel economy in the
                scrappage model. However, specifications considering this approach are
                shown in Section VI.C.1.b)(3)(c)(iii)(d).
                (b) New Vehicle Prices Net of Fuel Savings
                 UCS, CBD, NRDF, EDF, and other commenters expressed concern that
                quality adjustments were not included in the price series used to fit
                the NPRM model. In particular, commenters suggested that the valuation
                of fuel savings at the time of purchase should be deducted from the new
                vehicle price increases. For example, CBD argued:
                . . . [T]he agencies rely heavily on work by Howard Gruenspecht
                regarding the scrappage effect, and the NPRM acknowledges that
                Gruenspecht considered the effect of an increase in price ``net of
                the portion of reduced fuel savings valued by consumers.'' Yet
                consumer valuation of fuel savings is excluded from the scrappage
                model, as well.\1696\
                ---------------------------------------------------------------------------
                 \1696\ CBD, Appendix A, NHTSA-2018-0067-12000, at 177.
                 The scrappage model cannot include both independent variables on
                the fuel economy and cost-per-mile of new vehicles, and adjust the new
                vehicle prices by the value of fuel savings considered at the time of
                purchase, which would account for the improvement of the fuel economy
                of new vehicles twice. Thus, the agencies must choose between these
                methods to capture the value improvement of new vehicles when their
                fuel economy increases. The agencies show both methods in Section
                VI.C.1.b)(3)(c)(iii)(d). However, additional comments give reason to
                prefer a methodology that does not model the fuel economy or cost per
                mile of new model year cohorts directly, but instead adjusts the new
                vehicle price series by the amount of fuel savings valued at the time
                of purchase.
                 IPI expressed concern that the cost-per-mile measure was included
                in the scrappage model, but not in the sales model:
                 [T]he CPM results in the scrappage model are inconsistent with
                the agencies' sale model. In the sales module, the agencies have
                chosen to ignore consumer demand for fuel economy and significantly
                boosted the price impact of the baseline standards as a result. But
                in the scrappage model, the agencies have incongruously allowed
                consumer valuation of fuel economy to drive a significant portion of
                the estimated fatalities.\1697\
                ---------------------------------------------------------------------------
                 \1697\ IPI, Policy Integrity Comments: NHTSA Final--Appendix,
                NHTSA-2018-0067-12213, at 79.
                The agencies note that the fuel economy of new vehicles was not
                included in the sales model because the signs were statistically
                insignificant when it was included, and the fit of the overall model
                was not improved. It was not excluded because the agencies do not think
                that new vehicle fuel economy does not affect their sales. One way to
                consider the value of increased fuel economy in both the sales and the
                scrappage model (in the same way) is to adjust the price of new
                vehicles by the amount of fuel savings consumers value at the time of
                purchase in both models. This is also consistent with how the CAFE
                model applies technology in the absence of CAFE standards, or when a
                manufacturer is already in compliance with existing standards. In
                response to comments about the counterintuitive signs of the change in
                new vehicle cost per mile for some body styles, and about the
                disconnect in how the fuel economy of new vehicles is modelled in the
                sales and scrappage models, the agencies have adjusted the new vehicle
                price series in both models by the amount of fuel savings consumers are
                assumed to value at the time of purchase (30 months of fuel savings).
                As noted in Section VI.C.1.b)(3)(b)(ii)(a), alternatives to this
                solution are presented in Section VI.C.1.b)(3)(c)(iii)(d). The agencies
                also discuss consideration of other quality improvements over
                successive model years in Section VI.C.1.b)(3)(b)(iii)(d).
                (iii) Consideration of Other Additional Variables
                 Some commenters expressed concern that the scrappage model
                implemented in the NPRM analysis omitted several theoretically
                important variables in predicting the scrappage rates of the existing
                vehicle fleet. To understand these comments more fully it is useful to
                recall that existing vehicle owners can be private households/
                individuals, businesses, or dealerships. They supply the used vehicle
                (in the sense of making it available for use) to the market either by
                reselling them, or continuing to own the vehicle for their own use.
                Theoretically an existing owner will supply a used vehicle for
                additional use if the value of the vehicle (net of the opportunity cost
                of its value as scrap metal and used parts) exceeds the cost of
                maintenance, repair, insurance, and registration fees for the vehicle.
                If a seller does not perform necessary repair or maintenance services
                on the vehicle prior to sale, the value of the vehicle should be offset
                by the cost of those services. Accordingly, the scrappage threshold for
                a vehicle should remain the same regardless of whether the seller or
                buyer pays for any necessary maintenance or repair services on the
                vehicle.
                 Under this framework, commenters have argued that the agencies
                should include maintenance and repair costs, the value of the used
                vehicle when scrapped, and other costs to purchase the vehicle, all of
                which were excluded in the NPRM version of the scrappage models. IPI
                stated the following:
                 The agencies should include the variables that Gruenspecht and
                others have traditionally included in their scrappage analysis,
                including price of vehicles indexed by maintenance and repair costs,
                the price of scrap metal, and interest rates.\1698\
                ---------------------------------------------------------------------------
                 \1698\ IPI, Policy Integrity Comments: NHTSA Final--Appendix,
                NHTSA-2018-0067-12213, at 91.
                The agencies agree that these variables are relevant to determining the
                scrappage rates of existing vehicles, but have concerns that the level
                of aggregation of available series related to each of these factors may
                obscure the ability of a statistical model to capture their impact on
                vehicle scrappage rates.
                [[Page 24634]]
                Below, the agencies discuss commenter concerns about the omission of
                maintenance and repair costs, scrap steel prices, and interest rates,
                in turn. This rulemaking then outline the agencies' further
                consideration of each factor in this final rule analysis, and why each
                chose whether to consider each factor in the analysis for the final
                rule. Empirical results of models considering these factors are shown
                in Sections VI.C.1.b)(3)(c)(iii)(e) and VI.C.1.b)(3)(c)(iii)(f); the
                decision to exclude them from the primary analysis is further explained
                in these sections.
                (a) Maintenance and Repair Costs
                 EDF, IPI, California States et. Al., CARB, CBD, and other
                commenters suggest that the omission of maintenance and repair costs by
                the agencies was not justified, and that the measure should be included
                in future models. CARB claimed that:
                parameters for repair costs and used vehicle prices towards the end
                of life should likely be included in a scrappage model. However,
                neither of these variables appear in the Agencies' model.\1699\
                ---------------------------------------------------------------------------
                 \1699\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 244.
                The agencies agree that the theoretically ideal model of scrappage
                would include maintenance and repair costs. For this reason, the
                agencies explored several methods for explicitly incorporating
                maintenance and repair costs. Section VI.C.1.b)(3)(c)(iii)(f) reports
                model results both with and without a maintenance and repair variable.
                Since the variable is integrated of order one, (see Table VI-158), the
                models including it take the first difference; in this form, increases
                in maintenance and repair costs result in an increase in the scrappage
                rate of existing vehicles, as expected. The sign is also statistically
                significant. While the agencies would prefer a maintenance and repair
                price series that varies by calendar year and vintage, such a series is
                not currently available. The agencies hope to continue to improve this
                variable in future work on the scrappage model, but respond to comments
                by including the first difference of the maintenance and repair series
                in some of the models considered for the model used for the final rule.
                 Commenters were apparently confused about the agencies' discussion
                of the impact of fuel economy standards on durability. The agencies
                discussed a finding from the Greenspan and Cohen (1996) paper that
                suggested that higher EPA emission standards actually decreased the
                durability of certain model years. The discussion from the PRIA
                follows:
                 In addition to allowing new vehicle prices to affect cyclical
                vehicle scrappage [agrave] la the Gruenspecht effect, Greenspan &
                Cohen also note that engineering scrappage seems to increase where
                EPA emission standards also increase; as more costs goes towards
                compliance technologies, it becomes more expensive to maintain and
                repair more complicated parts, and scrappage increases. In this way,
                Greenspan and Cohen identify two ways that fuel economy standards
                could affect vehicle scrappage--(1) through increasing new vehicle
                prices, thereby increasing used vehicle prices, and finally,
                reducing on-road vehicle scrappage, and (2) by shifting resources
                towards fuel-saving technologies--potentially reducing the
                durability of new vehicles by making them more complex.\1700\
                ---------------------------------------------------------------------------
                 \1700\ PRIA at 1000.
                EDF and IPI misinterpret the agencies' discussion of findings from
                Greenspan and Cohen's work to imply that the fuel efficiency variable
                is meant to control for changes in maintenance and repair costs. The
                ---------------------------------------------------------------------------
                following quote from IPI exemplifies their confusion:
                 In addition, the agencies have explicitly excluded several
                theoretically important explanatory variables (e.g., the cost of
                maintenance and repair), which are potentially correlated with fuel
                efficiency. [Footnote 405: Id. at 1000 (indirectly making this point
                with respect to fuel efficiency and maintenance and repair costs
                when emphasizing that `Greenspan & Cohen also note that engineering
                scrappage seems to increase where EPA emission standards also
                increase; as more costs goes towards compliance technologies, it
                becomes more expensive to maintain and repair more complicated
                parts, and scrappage increases'). In other words, maintenance and
                repair costs are correlated with respect to fuel efficiency and
                scrappage rates.]\1701\
                ---------------------------------------------------------------------------
                 \1701\ IPI, Policy Integrity Comments: NHTSA Final--Appendix,
                NHTSA-2018-0067-12213, at 78.
                The agencies did not mean to imply that including some measure of the
                fuel economy of a model year cohort (cost per mile, in the NPRM model)
                would control for variation in maintenance and repair costs over time.
                The discussion of Greenspan and Cohen's results was intended only to
                demonstrate that durability and standards that increase technological
                complexity may be correlated, so that durability increases may not be
                independent of CAFE/CO2 standards.
                 Maintenance and repair costs for a given model year cohort likely
                are correlated with the fuel saving technologies applied to that
                cohort, but there is also a dimension of maintenance and repair costs
                that are correlated with other macroeconomic factors (i.e., wages,
                materials, etc.). Controlling for fuel economy would not capture
                calendar-year-specific changes to maintenance and repair costs that are
                caused by factors other than fuel economy. It also does not seem likely
                that variation in maintenance and repair costs from different fuel
                savings technology would be linearly related to fuel consumption, so
                that even model year variation in maintenance and repair costs could
                not be captured by including some measure of fuel economy or fuel
                consumption. As noted above, the agencies agree that maintenance and
                repair prices exist in the theoretically ideal scrappage model, and
                consider the variable in some of the models presented in Section
                VI.C.1.b)(3)(c)(iii)(f).
                (b) Scrap Values
                 In the NPRM model, the agencies considered inclusion of the BLS
                scrap steel CPI series. The agencies gave the following reasons for
                excluding the measure in the final NPRM models in the PRIA:
                 As noted by Parks (1977), the value of a scrapped vehicle can be
                derived either from the value of recoverable scrap metal or from the
                value of sellable used parts. There are several issues with using
                the BLS scrap steel CPI. First, as in Park's work, the coefficient
                on scrap steel is statistically insignificant--model results
                including the CPI of scrap steel are not shown, as there were other
                theoretical problems with the measure. The material composition and
                mass of vehicles has changed over time so that the absolute amount
                of recoverable scrap steel is not constant over the series. The
                average weight of recoverable steel by vintage would have to be
                known, and this measure would still be missing any other recoverable
                metals and other materials. Further, projecting the future value of
                the recoverable scrap metal would involve computing the amount of
                recoverable steel under all scenarios of fuel economy standards,
                where mass and material composition are assumed to vary across all
                alternatives. This value is not calculated explicitly in the current
                model, which is another reason some estimate of the value of
                recoverable metal is not included in the preferred model
                specification.\1702\
                ---------------------------------------------------------------------------
                 \1702\ PRIA at 1012.
                The concerns the agencies raised in the NPRM continue to be present for
                the model used for the final rule. The BLS scrap steel CPI will not
                have the same effect on the opportunity cost (the scrap value) of
                keeping an existing vehicle on the road as opposed to scrapping it for
                successive model year cohorts. The average weight of vehicles has
                changed over successive model years, as has the average steel
                composition.
                 Even considering the limitation of using the BLS scrap steel price
                series, commenters expressed concern about the exclusion of a variable
                to capture changes in the value of a vehicle as
                [[Page 24635]]
                scrapped metal and/or used vehicle parts. As noted in Section
                VI.C.1.b)(3)(b)(iii)(a), IPI suggested that ``the price of scrap
                metal'' should be included, while CARB suggested the model include
                ``used vehicle prices towards the end of life.'' The agencies made
                several further attempts to capture this component of vehicle
                scrappage, and address commenters' concerns, in the scrappage models
                used in the final rule. The agencies continue to consider models which
                include the BLS iron and scrap steel CPI series; results of these
                considerations are shown in Section VI.C.1.b)(3)(c)(iii)(f).
                (c) Interest Rates
                 IPI and EDF expressed concerns that changes in the real interest
                rates of vehicle loans had not been included in the final NPRM
                scrappage model. EDF commented the following:
                 NHTSA's model also does not include interest rates or the cost
                of financing a vehicle, another variable which NHTSA acknowledges
                affects scrappage. NHTSA itself states that ``[a]s the real interest
                rate increases so does the cost of borrowing and the opportunity
                cost of not investing. For this reason, it is expected that as real
                interest rates increase that vehicle scrappage should decline.
                Consumers delay purchasing new vehicles because the cost of
                financing increases. Conversely, as real interest rates decrease,
                vehicle scrappage should increase . . . . Yet, NHTSA chooses not to
                include interest rates in its model since inclusion of interest
                rates yields results that are opposite to what is expected--``as
                real interest rates increase, so does the scrappage rate'' in
                NHTSA's model. As discussed above, this is yet another indication
                that the model is flawed and cannot be relied upon.\1703\
                ---------------------------------------------------------------------------
                 \1703\ EDF, Appendix A, NHTSA-2018-0067-12108, at 41.
                 The agencies considered real interest rates in the NPRM analysis.
                Increasing the cost of purchasing a vehicle should increase the
                incentive for households to hold onto existing vehicles (as opposed to
                purchasing one) and scrappage rates should decline. The agencies
                excluded real interest rates from the final NPRM model for the reasons
                ---------------------------------------------------------------------------
                stated in the PRIA:
                 Table 8-14, Table 8-15, and Table 8-16 include interest rates
                and maintenance and repair CPI for cars, vans/SUVs, and pickups,
                respectively. For cars, as shown in Table 8-8, real interest rate is
                of the opposite sign than expected; as real interest rates increase,
                so does the scrappage rate--this model is also a worse fit by
                measures of AIC and BIC relative to the preferred model.\1704\
                ---------------------------------------------------------------------------
                 \1704\ PRIA at 1028.
                 In response to commenters' concerns, the agencies continue to
                consider interest rates in the model used for the final rule, as shown
                in Section VI.C.1.b)(3)(c)(iii)(e). However, interest rates only affect
                scrappage rates where a household might be unable to finance the
                purchase of a new or used vehicle and instead decides to maintain an
                existing vehicle that would have otherwise been scrapped. The most
                likely substitute for a marginal scrapped vehicle would not be a
                vehicle that could be financed. Accordingly, the relationship between
                interest rates and scrappage rates may be weaker than that between new
                vehicle prices and scrappage rates. The most likely substitutes for new
                vehicles are vehicles just off lease, and the resulting increase in
                residual values will affect slightly older vehicles. Eventually, the
                price of the most likely substitutes for marginally scrapped vehicles
                will also increase, so that scrappage rates will also be affected.
                (d) Other Vehicle Quality Adjustments
                 CARB and other commenters expressed concerns that the NADA series
                used by the agencies in development of the NPRM scrappage model did not
                make quality adjustments. CARB made the following specific comment:
                 By only including new vehicle prices and no other controls for
                vehicle quality, the Agencies' scrappage model omits variables that
                are important predictors of scrappage rates and of vehicle prices.
                Prior work that has relied on new vehicle prices to estimate
                scrappage rates have also included some aspects of quality
                improvements, meaning considering that the vehicle is improving in
                some way. For example, Greenspan and Cohen (1996) include both the
                Bureau of Labor Statistics (BLS) new vehicle price index and the BLS
                cost of repair index.\1705\
                ---------------------------------------------------------------------------
                 \1705\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 244.
                 The NADA average new vehicle transaction price does not control for
                other average characteristics that may change over successive model
                years. The agencies considered controlling for average body style and
                model year characteristics in the scrappage model as an alternative to
                including fixed effects in the model. The considered characteristics
                included: Horsepower to weight, zero to sixty acceleration time, and
                average curb weight. However, performing the pFtest implementation of
                an F-test of goodness-of-fit, from the ``plm'' R package, suggested
                that fixed effects are necessary to control for heterogeneity across
                model years.\1706\ For this reason, average characteristics that are
                constant over calendar years for a given model year cohort cannot be
                included in the model. The agencies do present specifications that
                include the ratio of new to used vehicle performance (since this has
                calendar year level variation and can be included with model year fixed
                effects) in Section VI.C.1.b)(3)(c)(iii)(f).
                ---------------------------------------------------------------------------
                 \1706\ Croissant, Y., Millo, G., & Tappe, K. (2019, September
                7). Package `plm.' Retrieved from https://cran.r-project.org/web/packages/plm/plm.pdf.
                ---------------------------------------------------------------------------
                (iv) Integration of Sales and/or VMT, Total Fleet Size, and Total VMT
                 Some commenters believe the ideal model of how CAFE/CO2
                standards affect sales, scrappage, and usage would be a joint household
                choice model. RFF makes the following comment:
                 The agencies can fix those problems by making two changes.
                First, they can jointly model VMT and vehicle holdings (i.e.,
                scrappage and new-vehicle purchases). The literature provides many
                examples of such modeling for guidance (see citations above).
                Jointly modeling these choices will make the analysis internally
                consistent and will account for the fact that households do not make
                scrappage and vehicle use decisions in isolation. If the model
                predicts that weaker standards cause more scrappage, it will
                simultaneously estimate any increase in VMT for the remaining
                vehicles.\1707\
                ---------------------------------------------------------------------------
                 \1707\ RFF, Comments EPA NHTSA, NHTSA-2018-0067-11789, at 14.
                 The advantage of such a model is that sales, scrappage, and usage
                would be jointly determined so that the impacts on scrappage is
                conditional on how increased new vehicle prices affect sales and
                vehicle prices, and usage is dependent on both effects. The agencies
                agree that this type of model would better capture the joint nature of
                the choices of which vehicles to buy, which to sell or scrap, and how
                much to use each than modelling each effect separately. However, the
                agencies are not aware of any national dataset that would allow sales,
                scrappage and usage to be jointly predicted, nor are they confident of
                such a model's ability to predict better than carrying current market
                shares forward.
                 The papers cited in the RFF comment, Linn and X. Dou, 2018; \1708\
                Berry, Levinsohn, and Pakes, 1995; \1709\ and Jacobsen and van Bentham,
                2015,\1710\ either use the CEX or the NADA transaction price series
                merged with the Polk registration counts. The CEX is a relatively small
                sample of households (about 160,000), their vehicle holdings,
                [[Page 24636]]
                vehicle purchases, and usage. However, it does not report retirement
                rates, but only when a vehicle exits a household's fleet (most often it
                is sold or traded in). Thus, at best, the CEX could be used to build a
                household consumer vehicle holdings and usage model, but the vehicles
                that are scrapped would be implied; scrappage would not be modeled
                directly, nor would it be attached to the number of miles on a vehicle.
                The NADA and Polk datasets used by Jacobsen and van Bentham links
                vehicles prices and scrappage rates, but does not track individual
                household decisions. The Jacobsen and van Bentham paper relies instead
                on a model of the new and used vehicle market which takes cross-price
                elasticities as an assumption derived from the outputs of a 1997 GM
                consumer choice model.1728 1711 The agencies will continue
                investigating whether a consumer/household choice model can serve as an
                alternative to aggregate estimates of sales and scrappage, but are
                skeptical about the ability of such models to predict future model
                shares accurately.
                ---------------------------------------------------------------------------
                 \1708\ J. Linn and X. Dou, ``How Do US Passenger Vehicle Fuel
                Economy Standards Affect Purchases of New and Used Vehicles?''
                (Washington, DC: Resources for the Future, 2018).
                 \1709\ Berry, S., J. Levinsohn, and A. Pakes, ``Differentiated
                Product Demand Systems from a Combination of Micro and Macro Data:
                The New Car Market,'' Journal of Political Economy 112(1) (2004):
                68-105.
                 \1710\ M. Jacobsen and A. van Benthem, ``Vehicle Scrappage and
                Gasoline Policy,'' American Economic Review 105 (2015): 1312-38.
                 \1711\ Kleit, Andrew N., 2004. ``Impacts of Long-Range Increases
                in the Corporate Average Fuel Economy (CAFE) Standard.'' Economic
                Inquiry 42:279-94.
                ---------------------------------------------------------------------------
                 As was the case with the 2012 final rule and the 2016 TAR, the
                agencies again note there is no credible consumer choice model which
                can be implemented in the CAFE model. Literature comparing the
                performance of consumer choice models to holding manufacturers constant
                suggest that the latter predicts future market shares better than the
                former. NCAT raises this point in their comment below:
                 Academic and other researchers have developed a number of
                vehicle demand (consumer choice) models for the new and/or used
                vehicle markets to look at effects on sales and fleet mix. Rarely
                has there been any effort to validate these models, either for
                consistency across models, or for ability to predict out of sample.
                Recent academic research, as well as work by EPA, has found that
                these models commonly perform worse, especially in the short run,
                than simply holding market shares constant.\1712\
                ---------------------------------------------------------------------------
                 \1712\ NCAT, NCAT Comments, NHTSA-2018-0067-11969, at 11.
                For these reasons, the agencies have not used a consumer choice model
                to capture the sales and/or scrappage impacts, but have built reduced
                form equations from aggregate data instead.
                 NCAT and CBD also refer to EPA attempts to develop a consumer
                choice model in conjunction with Oak Ridge National Labs, and note that
                the agencies did not use this model for the NPRM analysis. This
                specific choice model, as referenced in the excerpted NCAT comment
                above, has not predicted future market shares as well as projecting
                current shares forward. For this reason the model was not deemed fit to
                include in the policy analysis. NHTSA also worked to develop a consumer
                choice model, but when implemented, the model predicted that some OEM's
                would have unrealistic declines in total sales. The limitations of the
                consumer choice models the agencies have considered is overlooked in
                the following comments from CBD:
                 The sales model the agencies use is not the consumer-choice
                model that EPA has been developing and refining for almost a decade.
                Rather, both it and the scrappage model appear to have been
                developed by NHTSA in just the last two years. Neither model has
                been peer-reviewed, nor even released publicly until the publication
                of this NPRM.\1713\
                ---------------------------------------------------------------------------
                 \1713\ CBD, Appendix A, NHTSA-2018-0067-12000, at 175.
                The agencies did not use the consumer choice models either agency
                developed because the predictions are not reliable--which has
                disappointed not only the commenters mentioned above, but the agencies
                and researchers who have spent significant resources attempting to
                develop models for these purposes. Instead, the agencies have modelled
                the effects from reduced form equations from aggregate data.
                (a) Integration With Sales Model
                 The NPRM models did not include any direct linkage between the
                sales, scrappage, and usage functions, as noted by the agencies. Here,
                the agencies consider comments from stakeholders about the lack of
                integration of the scrappage model with sales (and the effect on total
                fleet size), and the lack of integration with the vehicle usage
                schedules (and the effects on total VMT).
                 NCAT, EDF, CBD, CARB, and other commenters argued that the sales
                and scrappage models should be directly linked, and that their
                independence predicts the higher fleet size and total VMT under the
                augural standards. CBD makes the following statement:
                 The agencies now, irrationally, decouple those two effects, such
                that the number of new vehicles sold (or left unsold) has no effect
                on the number of vehicles scrapped. Relying on the deeply flawed
                scrappage model, the agencies have predicted a massive ballooning of
                fleet size under the existing standards that leads, automatically
                under their model, to a massive increase in VMT. \1714\
                ---------------------------------------------------------------------------
                 \1714\ CBD, Appendix A, NHTSA-2018-0067-12000, at 185.
                 The agencies note that the structural model presented in Section
                VI.C.1.b)(3)(b)(i)(b) demonstrates that both the equilibrium quantity
                and the price of new vehicles sold are changed when the production cost
                of new vehicles changes under different regulatory alternatives.
                Specifically, under relaxed standards, the equilibrium price is lower
                and equilibrium sales are higher than the counterfactual augural
                standards. Controlling for other variables that might shift the new
                vehicle supply or demand curves, either new vehicle prices or sales
                could enter the used vehicle demand equation (as in the structural
                model, there is a functional relationship between the two, again,
                controlling for factors that shift the supply and demand curves for new
                vehicles). Thus, the agencies could use either new vehicle sales or
                prices to control for changes in the new vehicle equilibrium solution
                in the scrappage equation. It is important to control for factors that
                affect the demand for vehicles overall (business cycle conditions,
                etc.). The agencies present the preferred models using either new
                vehicle prices or new vehicles sales in Section
                VI.C.1.b)(3)(c)(iii)(d). Since there should be a collinearity between
                the two, it would be inappropriate to include both variables
                simultaneously.
                (b) Total Fleet Size
                 NCAT, EDF, CBD, CARB, UCS, IPI, California et. al., academic
                commenters, and other stakeholders argue that the fleet size should not
                change much with new vehicle prices. Some commenters go further to
                argue that higher vehicle prices under the augural standards should
                result in a smaller fleet size in the augural case relative to the
                proposal. The agencies agree that the long-term impact of higher new
                vehicle prices should be a slight reduction in fleet size, but do not
                agree that the short-term impacts of the standards on fleet size are
                obvious.
                 Many examples from the literature make assumptions that ensure that
                the fleet size under different regulatory alternatives remain constant.
                UCS cites this assumption in the original Gruenspecht works (their
                emphasis):
                 Though the agencies cite the Gruenspecht effect for its basis
                for the scrappage model, they ignore a central constraint of
                Gruenspecht's work--namely, his assumption that FLEET SIZE AND TOTAL
                VMT ARE INSENSITIVE TO PRICE.\1715\
                ---------------------------------------------------------------------------
                 \1715\ UCS, UCS MY2021-2026 NPRM: Technical Appendix, NHTSA-
                2018-0067-12039, at 60.
                Other works ensure the same conclusion with different assumptions.
                Within the
                [[Page 24637]]
                Jacobsen and van Bentham, 2015 and Goulder et. al., 2012 framework, a
                household first chooses the number of vehicles to own based on the
                average price of all vehicles subject to a budget constraint. After
                choosing the number of vehicles to hold, the household chooses the
                specific type and age of vehicles to hold. However, for some households
                the choice of how many and which vehicles to hold is not disjoint, so
                that a household may choose to hold two used vehicles as a second
                choice to one new vehicle. When new vehicle prices increase, under the
                same budget constraint, they may choose to hold two vehicles instead of
                one. If enough households make this choice, the fleet size could
                slightly increase.
                 IPI gives a literature example of a model that does not ensure this
                outcome with initial assumptions. This model directly predicted fleet
                size, and not sales and scrappage. The fleet size in the CAFE model is
                the result of the sales and scrappage models, and not the result of a
                single of the models. Small and Van Dender, 2007 finds that higher new
                vehicle prices are associated with lower total vehicle stock, as IPI
                states in the quote below: \1716\
                ---------------------------------------------------------------------------
                 \1716\ Auto Alliance, Attachment 1: NERA Evaluation, NHTSA-2018-
                0067-1207, at D-3.
                 In their 2007 study estimating the rebound effect caused by
                changes in fuel efficiency, Kenneth Small and Kurt Van Dender
                derived estimates of the relationship between vehicle price and
                fleet size. By simultaneously estimating a system of equations for
                VMT per capita, fleet size, and fuel efficiency for the United
                States from 1966 to 2001, Small and Van Dender also found that an
                increase in new vehicle price has a negative, statistically
                significant effect on total vehicle stock.\1717\
                ---------------------------------------------------------------------------
                 \1717\ IPI, Policy Integrity Comments: NHTSA Final--Appendix,
                NHTSA-2018-0067-12213, at 70.
                However, it is worth noting that Hymel, Small, and Van Dender in 2010
                published a study finding a statistically insignificant result of the
                opposite sign.\1718\ The general framework of the two papers are very
                similar, so that the updated results show that the fleet size impact is
                ambiguous.
                ---------------------------------------------------------------------------
                 \1718\ Hymel, Kent M. & Small, Kenneth A. & Dender, Kurt Van,
                2010. ``Induced demand and rebound effects in road transport,''
                Transportation Research Part B: Methodological, Elsevier, vol.
                44(10), pages 1220-1241.
                ---------------------------------------------------------------------------
                 Toyota and the Automobile Alliance mentioned that NERA built sales
                and scrappage models, and requested that the agencies ``review the NERA
                econometric study's methodologies for adoption or to refine their own
                models.'' The agencies considered the NERA scrappage model, but note
                that the model merges the data for all vehicle types, so that the
                scrappage relationship by age for pickups is adjusted by the same
                constant for all ages. However, the agencies note that each body style
                has a unique functional form with age--as evidenced in Section
                VI.C.1.b)(3)(c)(iii)(c))--so that it does not seem appropriate to merge
                them. Further, it does not seem likely that the elasticity of scrappage
                is the same for all vehicle types.
                 While the agencies think there are reasons not to adopt the NERA
                scrappage model as is, this suggested general approach does support
                simplifying the model as further suggested in Section
                VI.C.1.b)(3)(b)(i). Also, this research supports the notion that the
                relative fleet size of the proposed and augural standards is not a
                given. NERA's comments about their model provided:
                 The separate changes in new vehicle sales and changes in
                scrappage rates would lead to differences in the overall fleet size
                for the CAFE standard alternatives. The net effects of these two
                changes did not have a substantial effect on the overall fleet
                population under any of the three CAFE alternatives (never more than
                0.25% change in fleet size compared to the augural standards).\1719\
                ---------------------------------------------------------------------------
                 \1719\ Auto Alliance, Attachment 1: NERA Evaluation, NHTSA-2018-
                0067-1207, at D-3.HONDA.
                The NERA model shows the same directional fleet impacts as the NPRM
                sales and scrappage model. This lends some further support to the
                notion that the fleet impacts are not as certain as some commenters
                suggest.
                 Another empirical model predicts a larger total fleet size under
                the augural standards than under the proposed standards. Comments by
                David Bunch offer an extended comparison of the sales, fleet size, and
                retirement rate results of the Department of Energy's National Energy
                Modeling System (NEMS) model under the proposed and augural standards.
                NEMS predicts fleet size from input assumptions about the size of the
                on-road fleet, endogenous new vehicle sales estimates, and exogenous
                assumptions about scrappage.\1720\ However, in his comments Bunch said:
                ---------------------------------------------------------------------------
                 \1720\ From page 109 of 2016 NEMS documentation ``exogenously
                estimated vehicle scrappage and fleet transfer rates.'' https://www.eia.gov/outlooks/aeo/nems/documentation/archive/pdf/m070(2016).pdf.
                 Scrappage is an implied behavior determined by projecting total
                fleet size and new vehicle sales. Through this mechanism, all else
                equal, an increase in new vehicle sales would yield an increase in
                scrappage.\1721\
                ---------------------------------------------------------------------------
                 \1721\ David Bunch, Bunch-UC Davis: Consumer Behavior Modeling,
                at 77.
                 NEMS does not project total fleet size endogenously in their model
                as Bunch assumes. Nor is scrappage an implied behavior determined by
                fleet size and new sales projections. Instead, total fleet size is
                implied from an endogenous sales model, and constant age- and body-
                style-specific scrappage rates. The difference between the CAFE Model
                and NEMS is that the CAFE model has both endogenous new vehicles sales
                and scrappage rates--scrappage rates are not assumed to be constant for
                all regulatory alternatives. Fleet size is the implied variable in both
                models.
                 Bunch finds that the NEMS model also predicts a larger fleet size
                under the augural standards than the proposed standards. Specifically,
                he finds the following:
                 The differences are initially about 100K, increasing linearly
                from 2031 from 200K to 1.8M in 2050. Because even the Existing
                standards remain at the same level after 2025, this would seem to
                represent a very different effect from what might be going on in the
                CAFE model results.\1722\
                ---------------------------------------------------------------------------
                 \1722\ David Bunch, Bunch-UC Davis: Consumer Behavior Modeling,
                at 69.
                 Bunch goes on to discuss the relationship between sales, scrappage
                ---------------------------------------------------------------------------
                and fleet size in NEMS in the following passage:
                 New vehicle sales generally are growing in both scenarios, so
                economic theory suggests that fleet sizes should also be growing
                (they are). Specifically, although the Gruenspecht effect logic
                suggests that increasing new vehicle sales should lead to increased
                used vehicle scrap rates, the total ``value'' of the fleet is
                increasing, so this would suggest an increase in the fleet size.
                Moreover, new vehicle sales are higher under Existing, so the fleet
                size should be also.\1723\
                ---------------------------------------------------------------------------
                 \1723\ David Bunch, Bunch-UC Davis: Consumer Behavior Modeling,
                at 71.
                 Bunch makes several claims that are not consistent with available
                data and the agencies' understanding of how the NEMS model. First, he
                states that because sales are growing fleet size should also be
                growing. However, change in fleet size is the result of new vehicle
                sales less the number of existing vehicles scrapped; if new vehicle
                sales and used vehicle scrappage rates both increase, the fleet size is
                not necessarily increasing. Second, he states that the `Gruenspecht
                effect logic' suggests that increasing new vehicle sales results in
                increasing scrappage rates. However the NEMS model does not change
                vintage-specific scrappage rates endogenously, but takes them as an
                exogenous input. Thus, the NEMS model does not capture the Gruenspecht
                effect, and its fleet size projections can only vary from changes in
                new vehicle sales. Any differences in the projected total fleet
                scrappage rates Bunch considers later are due to
                [[Page 24638]]
                different initial sales of each body style, and therefore a different
                weighting of the constant body-style- and vintage-specific scrappage
                rates. This makes the comparison of the fleet size and scrappage rates
                of the two models not particularly meaningful. However, the difference
                in the projected sales impacts are worth a second glance. NEMS predicts
                prices that are at most about $1,000 higher in the Augural than the
                proposed standards, while the CAFE model predicts prices that are up to
                approximately $2,500 higher. The difference in the projected costs to
                meet the CAFE standards is likely the main reason for the difference in
                the sales outcomes--if the average fuel savings exceed the average
                incremental cost of the augural standards (relative to the proposal) in
                the NEMS model, the expected outcome is that sales should be higher in
                the augural case, as shown.
                 It is also worth noting Bunch's discussion of the empirical results
                of the CAFE scrappage model. Bunch purports to calculate the scrappage
                elasticity relative to new vehicle price increases, but his point of
                comparison does not hold constant other factors that might impact used
                vehicle scrappage rates. Instead, Bunch calculates the inter-annual
                percentage change in the scrappage rates for each regulatory
                alternative, then calculates the inter-annual change in new vehicle
                prices for each regulatory alternative, and finally takes the quotient.
                However, for inter-annual changes in scrappage rates, different
                projected GDP growth rates and fuel prices will have also played a
                critical role in the scrappage rates. The better point of comparison
                would be the incremental percentage decrease in scrappage rates for the
                augural standard relative to the proposal, over the incremental
                percentage increase in new vehicle price in the augural standard
                relative to the proposal for each calendar year. This ensures that the
                point of comparison holds constant all other factors that determine
                scrappage, as the regulatory alternatives use the same GDP growth rate
                and fuel price projections. When computing the implied scrappage
                elasticity in this way, the implied elasticities vary between
                approximates -0.1 and -1.1, with the average being approximately -0.5--
                which is more in line with what Bunch determines reasonable for his
                incorrect calculations of the NEMS model scrappage elasticities, as
                cited below:
                 Finally, the average values are -0.90 and -0.88 for the Existing
                and Rollback scenarios, respectively. On one hand, these are
                reasonably close to the Jacobsen and van Benthem (2015) estimate for
                scrap elasticity with respect to used vehicle prices. On the other
                hand, the Bento et al. (2018) estimate was -0.4, and one might
                expect the elasticity with respect to new vehicle price to be
                smaller. In any case, these results are not unreasonable.\1724\
                ---------------------------------------------------------------------------
                 \1724\ David Bunch, Bunch-UC Davis: Consumer Behavior Modeling,
                at 79.
                The implied elasticities from the NEMS model are approximately zero,
                which is not a surprise since these are merely the result of different
                new vehicle sales affecting the relative weighting of NEMS' constant
                age-specific scrappage rates. Figure VI-66, below, shows a comparison
                of fleet sizes under the baseline, preferred alternative, and AEO 2019.
                The agencies see that, as commenters believed likely, the fleet size
                under the preferred alternative (where sales are larger in many years
                and scrappage rates higher) is eventually larger than in the baseline.
                However, those differences are minimal in the early years of the
                simulation where policy differences produce only small differences in
                sales and scrappage. Furthermore, the agencies see that the magnitudes
                of the fleet sizes in today's rule are generally similar to those
                produced by the AEO 2019 model. NEMS tends to produce growth that is
                more linear, leading to slightly smaller fleet sizes than those
                simulated by the CAFE Model through the 2030's and slightly larger
                fleet sizes through the 2040's. However, these differences are at most
                three percent of fleet size, and typically closer to one or two
                percent.
                [[Page 24639]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.323
                 As discussed above, commenters offered NERA's model and NEMS as
                points of comparison for NHTSA's sales and scrappage models and their
                combined implied fleet size. However, since NEMS does not model the
                scrappage effect, but takes static scrappage rates, it is not a fair
                point of comparison. NERA's model shows a larger fleet under the
                Augural standards, providing evidence that the impacts of the sales and
                scrappage models are ambiguous.
                (c) Integration With VMT
                 In the NPRM the agencies noted that the average VMT by age is
                constant regardless of instantaneous or cumulative scrappage rates. The
                agencies noted that this was a limitation of the model, and sought
                comment on ways to integrate the two effects:
                 [O]ur scrappage model assumes that the average VMT for a vehicle
                of a particular vintage is fixed--that is, aside from rebound
                effects, vehicles of a particular vintage drive the same amount
                annually, regardless of changes to the average expected lifetimes.
                The agencies seek comment on ways to further integrate the survival
                and mileage accumulation schedules.\1725\
                ---------------------------------------------------------------------------
                 \1725\ EDF, Appendix B, NHTSA-2018-0067-12108, at 51.
                Several commenters suggest that the lack of integration between VMT and
                scrappage rates is not justified. Some commenters suggested that the
                VMT should be determined from a household holdings model, while others
                suggested merely that delayed scrappage under higher standards should
                increase average mileage accumulation, which will have some feedback
                for the next year's scrappage rates.
                 Joshua Linn and other commenters suggest that VMT is determined at
                the household level and should thus be modelled as such. EDF makes the
                following comment, which seems to reflect a fundamental
                misunderstanding of the type of model used to predict the scrappage
                effect:
                 When describing the process whereby a potential new vehicle
                purchaser chooses to forego buying a new vehicle and continues to
                drive their existing vehicle, NHTSA's scrappage model ignores the
                fact that this action shifts VMT from a new vehicle with a higher
                average mileage per year to a used vehicle with a lower average
                mileage. Either the driver of this vehicle will drive their older
                vehicle less, causing overall VMT to decline, or the average mileage
                of the used vehicle will increase without any need to affect
                scrappage. By focusing solely on scrappage, and focusing the change
                in scrappage on those vehicles with the worst fuel economy (i.e.,
                the oldest vehicles), NHTSA essentially shifts new vehicle VMT to
                the oldest vehicles. According to NHTSA's own rationale, much of the
                lost VMT from new vehicles will be replaced by vehicles only a few
                years old. The VMT of these relatively new used vehicles which is
                then replaced by VMT from older used vehicles, and so on.\1726\
                ---------------------------------------------------------------------------
                 \1726\ EDF, Appendix B, NHTSA-2018-0067-12108, at 51.
                 The agencies' scrappage model does not capture household choices,
                but uses aggregate data to predict new vehicle sales and age-specific
                scrappage rates in response to changes in new vehicle prices. In
                addition, the scrappage rates of all ages change in response to
                increases in new vehicle prices, not just the oldest vehicles. Further,
                the household that does not buy a new vehicle but holds onto an
                existing vehicle instead, in EDF's example, results in one fewer used
                vehicle supplied to the used market--this will result in an increased
                price for used vehicles and potentially lead to some used vehicles not
                being scrapped. Because the VMT schedules the agencies use in modelling
                show usage declining with age, the agencies' model does assume that
                younger vehicles that are not scrapped are driven more than older
                vehicles that are not scrapped.
                [[Page 24640]]
                 EDF, IPI, and Honda further argue that mileage accumulation should
                not be constant under all scrappage rates. Specifically, they suggest
                that the assumption that average VMT accumulation by age is constant
                even when scrappage rates decline, results in an overestimate of VMT.
                IPI suggests that the marginally unscrapped vehicles should drag down
                the average VMT accumulation under higher standards in the following
                comment:
                 Because those schedules assume each vehicle of a certain age and
                type in the fleet drives a set amount of miles without any
                adjustment for the increase in total fleet size or vehicle quality
                (i.e., wear and tear and durability), the finding that the standards
                cause the fleet size to increase results in a significant increase
                in total VMT.\1727\
                ---------------------------------------------------------------------------
                 \1727\ IPI, Policy Integrity Comments: NHTSA Final--Appendix,
                NHTSA-2018-0067-12213, at 61.
                The agencies note that mileage accumulation and scrappage are not
                disjoint. A vehicle that is driven more miles is more likely to be
                scrapped. However, since the National Vehicle Population Profile (NVPP)
                data does not track individual vehicles, there is no obvious way to
                merge individual vehicle odometer readings with those that are
                scrapped. The agencies explored different data sources that could be
                used to capture the joint relationship of the two effects, but
                unfortunately were unable to identify a workable dataset. Furthermore,
                the agencies note that while commenters could be correct about the
                relationship between mileage accumulation and scrappage, they did not
                provide the agencies with any empirical evidence supporting their
                assertions.\1728\ In the meantime, the agencies have adjusted the final
                rule analysis to conservatively assume that total demand for VMT, not
                including the rebound effect, should be constant for all regulatory
                alternatives, as discussed in Section VI.C.1.b)(3)(b)(iv)(d), below.
                This requires that the VMT schedules are no longer constant for all
                fleet sizes.
                ---------------------------------------------------------------------------
                 \1728\ EDF, Appendix B, NHTSA-2018-0067-12108, at 54.
                ---------------------------------------------------------------------------
                (d) Total VMT
                 Many commenters think that total VMT, not considering rebound
                miles, should be constant, regardless of the number of new vehicles
                sold and used vehicles scrapped. NCAT, Global, Auto Alliance, CBD, EDF,
                IPI, CARB, and Honda all make this argument. CARB makes the following
                statement suggesting that even a larger fleet size should not increase
                aggregate demand for VMT (again, not including rebound miles):
                 A change in the overall fleet size due to the Augural standards
                might not in and of itself be problematic, as long as the VMT
                schedules are adjusted to account for overall travel activity that
                is distributed over a larger number of vehicles. However, the As-
                Received version of the [scrappage] model does not adjust VMT
                schedules, with the result that the additional unscrapped vehicles
                inflate total VMT proportionally.\1729\
                ---------------------------------------------------------------------------
                 \1729\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 238.
                The agencies agree that the aggregate demand for VMT should be roughly
                constant across alternatives, and stated this in the NPRM, where the
                differences in non-rebound VMT were on the order of 0.4%.
                 NERA's modelling efforts found similar small decreases in VMT in
                regulatory alternatives where the standards are relaxed. The Alliance
                stated:
                 Under all three scenarios, vehicle miles traveled (``VMT'')
                decreases relative to the augural standards. This is due primarily
                to rebound effects. Because NERA was only examining vehicles through
                MY 2029, the difference in VMT between the alternatives and the
                augural standards decreases over time, since fewer of the MY 2029
                and earlier vehicles are on the road in those later years.\1730\
                ---------------------------------------------------------------------------
                 \1730\ Auto Alliance, Full Comment Set, NHTSA-2018-0067-12073,
                at 11.
                 NERA's model used similar assumptions as the NPRM analysis and,
                like the NPRM results, the NERA model results suggest that it is
                plausible that total VMT could decline under less stringent standards.
                A key assumption common to NERA's model and the NPRM analysis is that
                the VMT schedules are constant under all scrappage rates. However, as
                discussed in Section VI.C.1.b)(3)(b)(iv)(c), this can potentially
                overestimate total VMT in the augural case, where vehicles that were
                marginally scrapped in the proposal are kept on the road.
                 Presumably, vehicles that are scrapped in the proposal, but not in
                the augural, are in more disrepair than others in the same age cohort.
                As a result, these vehicles would on average be driven less, bringing
                down the average usage of the entire age cohort. This effect could
                alter the relative size of total VMT under the regulatory alternatives,
                as Honda notes in the following comment:
                 According to our calculations, if the impact of lowering the
                average cohort's utility is even 0.2% the augural standards would
                become safer than the preferred alternative. We believe that the
                agencies should consider VMT behavior change as part of an effort to
                mature and refine the scrappage model.\1731\
                ---------------------------------------------------------------------------
                 \1731\ Honda, Honda Comment, NHTSA-2018-0067-11818, at 18.
                As Honda suggests, a relatively small reduction in the average VMT
                schedules for the more stringent regulatory alternatives could result
                in a change in the direction of the safety impact. This shows the
                importance of investigating the linkage between usage and scrappage
                rates, but also shows that small changes to the total VMT assumptions
                can have meaningful impacts on the predicted effects of the analysis.
                Other commenters make similar points.
                 As noted above, the difference in total non-rebound VMT in the NPRM
                analysis was only 0.4%. However, CBD notes that this relatively small
                change in VMT across the alternatives in a single year can result in a
                large number of cumulative additional miles in more stringent
                regulatory alternatives:
                 While 0.4% sounds small, when the scrappage model's effect it is
                multiplied by all the VMT that NHTSA includes in its analysis,
                spanning decades, it becomes highly significant--at least 692
                billion additional VMT under the CAFE standards and 894 billion
                under the CO2 program, both relative to the preferred
                alternative.\1732\
                ---------------------------------------------------------------------------
                 \1732\ CBD, Appendix A, NHTSA-2018-0067-12000, at 180.
                Since VMT is related to many of the costs and benefits of the program,
                differences in cumulative VMT of this magnitude can have meaningful
                impacts on the incremental net benefit analysis. This point was implied
                by comments from CBD, EDF, NCAT, EAO, and in a paper published by
                academics after the issuance of the NPRM.\1733\ For this reason, the
                agencies have opted to constrain total non-rebound VMT across
                regulatory alternatives.
                ---------------------------------------------------------------------------
                 \1733\ Bento, Antonio M., et al. ``Flawed Analyses of U.S. Auto
                Fuel Economy Standards.'' Science, vol. 362, no. 6419, 2018, pp.
                1119-21., doi:10.1126/science.aav1458.
                ---------------------------------------------------------------------------
                 Such a constraint was suggested by EDF, IPI and other commenters.
                EDF states the following:
                 A sophisticated model is not needed to correct this problem. One
                only needs to adjust the VMT added by the ``scrappage model'' so
                that it matches the VMT lost by the sales response model. Put
                another way, used vehicles would be used to the same extent as new
                vehicles since they meet the identical demand (possibly minus a
                rebound effect). \1734\
                ---------------------------------------------------------------------------
                 \1734\ EDF, Appendix B, NHTSA-2018-0067-12108, at 49.
                EDF goes on to suggest some potential issues with implementing this
                ---------------------------------------------------------------------------
                constraint:
                 Even this adjustment would still be in favor of the proposal, as
                it assumes that all the VMT lost from fewer new vehicle sales would
                be replaced by used vehicle VMT. This assumes that travel is
                inelastic. This is
                [[Page 24641]]
                clearly not the case given NHTSA's position on the rebound effect.
                NHTSA must first justify the used vehicle response to any change in
                new vehicle sales. Then, in the unlikely event that this can be
                done, NHTSA must link the scrappage model to the sales response
                model to ensure that the combination of the two models does not
                increase VMT in any calendar year (and probably show a decrease, as
                the overall cost of driving will have increased).\1735\
                ---------------------------------------------------------------------------
                 \1735\ EDF, Appendix B, NHTSA-2018-0067-12108, at 49.
                The agencies disagree that lost new vehicle sales would impact the VMT
                of the new vehicles that are sold. The agencies do, however, as EDF
                notes, adjust the VMT of new vehicles to consider changes in the cost
                per mile of travel. In fact, when fuel prices increase, the agencies
                assume that owners of all existing vehicles drive less; the reduction
                will be greater when the vehicles on the road are less efficient, which
                seems consistent with what EDF suggests in the last sentence above. The
                agencies have justified the scrappage effect throughout this
                discussion, above.
                 EDF identifies another reason the agencies think a constraint on
                total VMT is reasonable for purpose of the final rule analysis. The
                scrappage, sales, and VMT models each have a certain amount of
                uncertainty associated with it (the uncertainty of the scrappage model
                is discussed in Section VI.C.1.b)(3)(b)(i)(a)), so that when the three
                models are combined, the uncertainty is compounded. EDF characterizes
                these results as being inconsistent with economic theory in the comment
                below:
                 We are not aware of any economic arguments which would support
                such an increase. All that can be said is that NHTSA put data from a
                variety of sources through a statistical regression and never
                bothered to see if the results were reasonable or consistent with
                its own economic theory. \1736\
                ---------------------------------------------------------------------------
                 \1736\ EDF, Appendix B, NHTSA-2018-0067-12108, at 57.
                The NPRM analysis discussed total fleet size and VMT at length; the
                agencies noted that the fleet was 1.5% bigger for the augural standard
                than the proposal, resulting in 0.4% additional non-rebound VMT in
                CY2050.\1737\ However, given the amount of uncertainty around each of
                the models, and considering that differences in total VMT can have
                meaningful impacts on the cost benefit analysis, the agencies are
                conservatively assuming for the final rule analysis that non-rebound
                VMT is constant, to constrain the outputs derived from the combination
                of the three models.
                ---------------------------------------------------------------------------
                 \1737\ FR, Vol 83, No. 165, August 24, 2018, p.43099.
                ---------------------------------------------------------------------------
                (v) Comments on the Evaluation of Associated Costs and Benefits
                (a) Presentation and Valuation of Non-Rebound Miles
                 IPI and EDF argued that it was inconsistent to exclude the costs
                and benefits of additional rebound driving but include them for the
                sales and scrappage effect. For example, EDF stated:
                 [W]henever a vehicle is driven an additional mile, there is
                value associated with that travel. NHTSA completely ignores the
                value of any additional travel which occurs due to reduced
                scrappage. Including this value would not be an adequate surrogate
                for the additional repair costs required to keep older vehicles on
                the road. Just as NHTSA is now recognizing that rebound VMT is due
                to drivers' express decision to drive more, any driving of older
                vehicles in lieu of new vehicles is due to the same choice. To treat
                these identical choices in 180 degree different manners is of course
                manifestly arbitrary. \1738\
                ---------------------------------------------------------------------------
                 \1738\ EDF, Appendix B, NHTSA-2018-0067-12108, at 58.
                 The agencies agree that there is value associated with additional
                miles driven. The NPRM did not directly attribute costs for the loss of
                additional miles in the scrappage analysis when the fleet size shrank.
                The final rule analysis addresses this issue by holding non-rebound
                total VMT constant across regulatory alternatives. However, contrary to
                what EDF suggests above, the cost of additional maintenance and repair
                for otherwise-scrapped vehicles are not directly related to the
                additional miles. The cost of additional maintenance and repair is
                incurred because the value of used vehicles has increased. The increase
                in value of the used vehicles should at least offset the maintenance
                and repair costs.
                 Holding aggregate non-rebound VMT constant across alternatives
                addresses IPI's and EDF's concerns that additional miles due to a
                larger fleet size were not adequately valued. However, on average newer
                vehicles tend to be safer, more efficient, more powerful, and more
                spacious than used vehicles. Because of this, driving a newer vehicle
                will be more enjoyable, and provide more utility per mile, than driving
                a used vehicle. Even disregarding trends in vehicle quality, the
                utility of a mile driven in a newer vehicle is on average higher than
                that driven in an older vehicle because the average newer vehicles in
                better condition. The regulation is responsible for the shift in the
                distribution of miles driven at each vehicle age. Including the
                additional safety risks and fuel costs accrued from more miles being
                driven by older vehicles accounts for part of the reduction in the
                utility of the average mile under more stringent standards. Quantifying
                the remaining change in utility of more miles being driven by older
                vehicles is currently beyond the scope of this rulemaking analysis and
                will require extensive future research. The agencies do not think
                excluding other sources of changes in the utility of driving
                (performance, comfort, etc.) will significant change the outcome of the
                analysis.
                (b) Increase in Maintenance and Repair Costs and Used Vehicle Values
                 EDF and others also commented that the agencies should include the
                value of additional maintenance and repair costs and the increase in
                value for used vehicles explicitly in the cost and benefit analysis.
                They state the following:
                 ``It is important to note that NHTSA fails to account for three
                large economic impacts occurring during this process.
                 1. The increase in value of the entire used vehicle fleet from
                2017-2050. This is a windfall gain for all current vehicle owners
                that is completely ignored;
                 2. The cost of repairing and maintaining the older vehicles
                which are no longer scrapped;
                 3. The value of the additional driving that these vehicles
                provide.
                 NHTSA only counts the costs related to the additional driving
                performed by the non-scrapped vehicles. Again, NHTSA's decision to
                only include this cost maximizes monetary costs related to the
                current standards and minimizes those related to the proposal.''
                \1739\
                ---------------------------------------------------------------------------
                 \1739\ EDF, Appendix B, NHTSA-2018-0067-12108, at 50.
                 As discussed above, in Section VI.D.1.b)(3)(a)(a), the agencies
                hold the non-rebound fleetwide VMT constant to an exogenous projection
                of aggregate VMT. This addresses EDF's third concern, above. Without a
                model of the used vehicle market it is impossible for the agencies to
                estimate the value increase of used vehicles due to a substitution
                towards used vehicles when new vehicle prices increase. However, the
                maintenance and repair costs should be less than or equal to the
                increase in vehicle value (or the current owner would not pay to
                maintain the vehicle). Not including the additional maintenance and
                repair costs should at least partially offset not including the
                increase in the value of used vehicles. The remaining increase in
                vehicle value should be a transfer between the seller and buyer of a
                used vehicle so that it should be both a cost and benefit exactly
                offsetting. Thus, the total costs and benefits are understated by the
                same amount, and including them
                [[Page 24642]]
                should not affect the reported net benefits of the rule.
                (c) Scrappage Effects From MY2030 and Beyond
                 The NPRM analysis considered cost per mile as a continuous
                variable, and new vehicle prices in discrete levels. This means that
                persistently higher new vehicle prices in more stringent standards
                would continue to suppress the scrappage rate of existing vehicles. It
                also means that higher fuel economies in more stringent scenarios would
                continue to affect the scrappage rates as well. EDF noted that the cost
                and benefit accounting that considered the costs and benefits accruing
                to the remaining lifetimes of MYs 1977-2029 included some of the costs
                of the scrappage effect due to the higher prices of MYs beyond 2030,
                but did not include the benefits of the reduced fuel economy for these
                MYs. EDF proposed that the agencies consider a CY analysis instead of
                the model year presented in the NPRM:
                 [A] 2017-50 CY analysis would include the operation of 2017-2029
                MY vehicles through CY 2050. This would include the any scrappage
                effects on these vehicles through 2050, consistent with the
                inclusion of new 2050 MY vehicles in the analysis. Some of the
                operation of all the 2017-2029 MY vehicles would be excluded from
                the analysis, as these vehicles are not assumed to be scrapped in
                the Volpe Model until CY 2052-2068. Such an analysis would include
                the benefits over the clear majority of the operation of 2017-2029
                MY vehicles compared to both the shorter calendar year analysis and
                NHTSA's 1977-2029 MY analysis. It would also include the scrappage
                effects caused by 2017-2050 MY vehicles through CY 2050. Any
                scrappage effects would be applied to 2030-2050 MY vehicles, as well
                as 2017-2029 MY vehicles.\1740\
                ---------------------------------------------------------------------------
                 \1740\ EDF, Appendix B, NHTSA-2018-0067-12108, at 22.
                However, as the commenter also notes, a CY analysis would exclude some
                of the lifetime costs and benefits of improving the fuel economy of MYs
                impacted by the rule (MYs 2017-2029). For this reason, the agencies do
                not think that a CY analysis should supplant the MY perspective shown
                in the NPRM.
                 EDF presents an alternative to switching to a CY analysis which
                would exclude the scrappage effects due to differences in the prices
                and fuel efficiencies of MYs not included in the cost benefit analysis
                (MY 2030 and beyond):
                 An alternative that keeps the model year structure of NHTSA's
                1977-2029 MY analysis would be to modify it by removing any
                scrappage effects occurring in 2030 CY and beyond. This analysis
                would still have the disadvantage of barely including any vehicles
                which reflect full compliance with the current and proposed
                standards in 2025. However, it would at least remove the primary
                problem with NHTSA's current MY analysis. The impact of including
                the scrappage effects caused by 2030 and later MY vehicles simply
                and straightforwardly increases the VMT of used vehicles under the
                current standards.\1741\
                ---------------------------------------------------------------------------
                 \1741\ EDF, Appendix B, NHTSA-2018-0067-12108, at 23.
                The agencies note that previous analyses have not considered the costs
                and benefits of MYs beyond those which could be a response to the
                change in the considered set of standards. Part of the reason for this
                was that future standards are unknown, and without existing standards
                in place, manufacturers may choose to shift application of fuel saving
                technologies to increases in vehicle performance or safety. The CAFE
                model does not currently simulate such actions, so that including MYs
                too far into the future may overstate the costs and benefits of the
                rule.
                 While the agencies disagree that excluding cost and benefits of MYs
                beyond 2030 is an issue for the cost benefit analysis, the agencies
                agree that allowing persistently higher prices and fuel economies of
                future MYs to impact the scrappage of the on-road fleet but not
                considering the costs and benefits of those MYs is inconsistent.
                However, changes to the scrappage model mitigate this issue. As noted
                in Section VI.C.1.b)(3)(b)(i)(c) and VI.C.1.b)(3)(b)(ii), updates to
                the time series strategy and the way that new vehicle fuel economy is
                modelled in the FRM scrappage model change the form of how new vehicle
                prices and fuel economy enter the equation. First, addressing the
                autocorrelation by taking the first difference of variables with first
                order integration instead of including lags of the dependent variables
                means that cost per mile variables and new vehicle prices are captured
                as changes rather than in levels. This means that constant, but higher,
                new vehicle prices in the augural standards will not continue to impact
                the scrappage rates of existing vehicles. More specifically, higher
                prices of MYs 2030 and beyond in the augural case will no longer result
                in lower scrappage rates for prior MYs. Further, since new vehicle cost
                per mile is no longer explicitly included, but rather the amount of
                fuel savings consumers of new vehicles value at the time of purchase is
                excluded from the new vehicle prices series, differences in new vehicle
                fuel economies for MYs beyond 2029 will no longer impact the scrappage
                rates of earlier MYs. This naturally takes care of the concern raised
                by several commenters that the accounting for costs and benefits due to
                changes in MYs 2030 and beyond was inconsistent due to the scrappage
                model.
                (c) Estimation of the FRM Scrappage Models
                (i) Framing Dynamic Scrappage Models in the Literature
                (a) How Fuel Economy Standards Impact Vehicle Scrappage
                 As noted above, any increase in price (net of the portion of
                reduced fuel savings valued by consumers) will increase the expected
                life of used vehicles and reduce the number of new vehicles entering
                the fleet (the Gruenspecht effect). In this way, increased fuel economy
                standards slow the turnover of the fleet and the entrance of any
                regulated attributes tied only to new vehicles. Gruenspecht tested his
                hypothesis in his 1981 dissertation using new vehicle price and other
                determinants of used car prices as a reduced form to approximate used
                car scrappage in response to increasing fuel economy standards.
                 Greenspan and Cohen (1996) offer additional foundations from which
                to think about vehicle stock and scrappage. Their work identifies two
                types of scrappage: Engineering scrappage and cyclical scrappage.
                Engineering scrappage represents the physical wear on vehicles which
                results in their being scrapped. Cyclical scrappage represents the
                effects of macroeconomic conditions on the relative value of new and
                used vehicles--under economic growth the demand for new vehicles
                increases and the value of used vehicles declines, resulting in
                increased scrappage. In addition to allowing new vehicle prices to
                affect cyclical vehicle scrappage [agrave] la the Gruenspecht effect,
                Greenspan and Cohen also note that engineering scrappage seemed to
                increase where EPA vehicular-criteria pollutant emissions standards
                also increased; as more costs went towards compliance technologies,
                scrappage increased. In this way, Greenspan and Cohen identify two ways
                that fuel economy standards could affect vehicle scrappage: (1) Through
                increasing new vehicle prices, thereby increasing used vehicle prices,
                and finally, reducing on-road vehicle scrappage, and (2) by shifting
                resources towards fuel-saving technologies--potentially reducing the
                durability of new vehicles.
                [[Page 24643]]
                (b) Aggregate vs. Atomic Data Sources in the Literature
                 One important distinction in literature on vehicle scrappage is
                between those that use atomic vehicle data (data following specific
                individual vehicles), and those that use some level of aggregated data
                (data that counts the total number of vehicles of a given type). The
                decision to scrap a vehicle is made on an individual vehicle basis, and
                relates to the cost of maintaining a vehicle, and the value of the
                vehicle both on the used car market, and as scrap metal. Generally, a
                used car owner will decide to scrap a vehicle when the value of the
                vehicle is less than the value of the vehicle as scrap metal, plus the
                cost to maintain or repair the vehicle. In other words, the owner gets
                more value from scrapping the vehicle than continuing to drive it, or
                from selling it.
                 Recent work is able to model scrappage as an atomic decision due to
                the availability of a large database of used vehicle transactions. Work
                by authors including Busse, Knittel, and Zettelmeyer (2013), Sallee,
                West, and Fan (2010), Alcott and Wozny (2013), and Li, Timmins, and von
                Haefen (2009) consider the impact of changes in gasoline prices on used
                vehicle values and scrappage rates. In turn, they consider the impact
                of an increase in used vehicle values on the scrappage rate of those
                vehicles. They find that increases in gasoline prices result in a
                reduction in the scrappage rate of the most fuel efficient vehicles and
                an increase in the scrappage rate of the least fuel efficient vehicles.
                This has important implications for the validity of the average fuel
                economy values linked to model years, and assumed to be constant over
                the life of that model year fleet within this study. Future iterations
                of such studies could further investigate the relationship between fuel
                economy, vehicle usage, and scrappage, as noted in other places in this
                discussion.
                 While the decision to scrap a vehicle is made atomically, the data
                available to NHTSA on scrappage rates and variables that influence
                these scrappage rates are aggregate measures. This influences the best
                available methods to measure the impacts of new vehicle prices on
                existing vehicle scrappage. The result is that this study models
                aggregate trends in vehicle scrappage, and not the atomic decisions
                that make up these trends. Many other works within the literature use
                the same data source and general scrappage construct, including those
                by Walker (1968), Park (1977), Greene and Chen (1981), Gruenspecht
                (1981), Gruenspecht (1982), Feeney and Cardebring (1988), Greenspan and
                Cohen (1996), Jacobsen and van Bentham (2015), and Bento, Roth, and
                Zhuo (2016.). These works all use aggregate vehicle registration data
                as the source to compute vehicle scrappage.
                 Walker (1968) and Bento, Roth and Zhuo (2016) use aggregate data
                directly to compute the elasticity of scrappage from measures of used
                vehicle prices. Walker (1968) uses the ratio of used vehicle Consumer
                Price Index (CPI) to repair and maintenance CPI. Bento, Roth, and Zhuo
                (2016) use used vehicle prices directly. While the direct measurement
                of the elasticity of scrappage is preferable in a theoretical sense,
                the CAFE model does not predict future values of used vehicles, only
                future prices of new vehicles. For this reason, any model compatible
                with the current CAFE model must estimate a reduced form similar to
                Park (1977), Gruenspecht (1981), and Greenspan and Cohen (1996), who
                use some form of new vehicle prices or the ratio of new vehicle prices
                to maintenance and repair prices to impute some measure of the effect
                of new vehicle prices on vehicle scrappage.
                (c) Historical Trends in Vehicle Durability
                 Waker (1968), Park (1977), Feeney and Cardebring (1988), Hamilton
                and Macauley (1999), and Bento, Ruth, and Zhuo (2016) all note that
                vehicles change in durability over time. Walker (1968) simply notes a
                significant distinction in expected vehicle lifetimes pre- and post-
                World War I. Park (1977) discusses a `durability factor' set by the
                producer for each year, so that different vintages and makes will have
                varying expected lifecycles. Feeney and Cardebring (1988) show that
                durability of vehicles appears to have generally increased over time
                both in the U.S. and Swedish fleets using registration data from each
                country. They also note that the changes in median lifetime between the
                Swedish and U.S. fleet track well, with a 1.5-year lag in the U.S.
                fleet. This lag is likely due to variation in how the data is
                collected--the Swedish vehicle registration requires a title to
                unregister a vehicle, and therefore gets immediate responses, where the
                U.S. vehicle registration requires re-registration which creates a lag
                in reporting further discussed in Section VI.C.1.b)(3)(c)(ii)(b).
                 Hamilton and Macauley (1999) argue for a clear distinction between
                embodied versus disembodied impacts on vehicle longevity. They define
                embodied impacts as inherent durability similar to Park's producer
                supplied `durability factor' and Greenspan's `engineering scrappage'
                and disembodied effects as those which are environmental, not unlike
                Greenspan and Cohen's `cyclical scrappage.' They use calendar year and
                vintage dummy variables to isolate the effects--concluding that the
                environmental factors are greater than any pre-defined `durability
                factor.' Some of their results could be due to some inflexibility of
                assuming model year coefficients are constant over the life of a
                vehicle, and also some correlation between the observed life of the
                later model years of their sample and the `stagflation' \1742\ of the
                1970's. Bento, Ruth, and Zhuo (2016) find that the average vehicle
                lifetime has increased 27 percent from 1969 to 2014 by sub-setting
                their data into three model year cohorts. To implement these findings
                in the scrappage model incorporated into the CAFE model, this study
                takes pains to estimate the effect of durability changes in such a way
                that the historical durability trend can be projected into the future;
                for this reason, the agencies include a continuous `durability' factor
                as a function of model year vintage.
                ---------------------------------------------------------------------------
                 \1742\ Continued high inflation combined with high unemployment
                and slow economic growth.
                ---------------------------------------------------------------------------
                (ii) Polk/IHS Registration Data
                 As in the NPRM, NHTSA uses proprietary data on the registered
                vehicle population from IHS/Polk for the scrappage models. IHS/Polk has
                annual snapshots of registered vehicles counts beginning in calendar
                year (CY) 1975 and continuing until CY2017. Notably, the data
                collection procedure changed in CY2002, which requires some special
                consideration (discussed below). The data includes the following
                regulatory classes as defined by NHTSA: Passenger cars, light trucks
                (classes 1 and 2a), and medium and heavy-duty trucks (classes 2b and
                3). Polk separates these vehicles into another classification scheme:
                cars and trucks. Under their schema, pickups, vans, and SUVs are
                treated as trucks, and all other body styles are included as cars. In
                order to build scrappage models to support the model year (MY) 2021-
                2026 light duty vehicle (LDV) standards, it was important to separate
                these vehicle types in a way compatible with the existing CAFE model.
                (a) Choice of Aggregation Level: Body Style
                 Two compatible methods existed by which the agencies could
                aggregate scrappage rates: By regulatory class or by body style. Since,
                for CAFE
                [[Page 24644]]
                purposes, vans/SUVs are sometimes classified as passenger cars and
                sometimes as light trucks (depending upon vehicle-specific attributes)
                and there was no simple way to reclassify some SUVs as passenger cars
                within the Polk dataset, the agencies chose to aggregate survival
                schedules by body style. This approach is also preferable because it is
                consistent with the level of aggregation of the VMT schedules. Since
                usage and scrappage rates are not independent of each other, if average
                usage rates are meaningfully different at the level of body style, it
                is likely that scrappage rates are as well.
                 Once stratified into body style level buckets, the data can be
                aggregated into population counts by vintage and age. These counts
                represent the population of vehicles of a given body style and vintage
                in each calendar year. The difference between the counts of a given
                vintage and vehicle type from one calendar year to the next is assumed
                to represent the number of vehicles of that vintage and type scrapped
                in each year.
                (b) Greenspan and Cohen Correction
                 One issue with using snapshots of registration databases as the
                basis for computing scrappage rates is that vehicles are not removed
                from registration databases until the last valid registration expires--
                for example, if registrations are valid for a year, vehicles will still
                appear to be registered in the calendar year in which they are
                scrapped. To correct for the scrappage that occurs during a calendar
                year, a similar correction as that in Greenspan and Cohen (1996) is
                applied to the Polk dataset. It is assumed that the real on-road count
                of vehicles of a given MY registered in a given CY is best represented
                by the Polk count of the vehicles of that model year in the succeeding
                calendar year (PolkCY+1). For example, the vehicles scrapped
                between CY2000 and CY2001 will still remain in the Polk snapshot from
                CY2000 (PolkCY2000), as they will have been registered at
                some point in that calendar year, and therefore exist in the database.
                Using a simplifying assumption that all States have annual registration
                requirements,\1743\ vehicles scrapped between July 1st, 1999 and July
                1st, 2000 will not have renewed registration between July 1st, 2000 and
                July 1st, 2001, and will not show up in PolkCY2001. The
                vehicles scrapped during CY2000 are therefore represented by the
                difference in count from the CY2000 and CY2001 Polk datasets:
                PolkCY2001-PolkCY2000.
                ---------------------------------------------------------------------------
                 \1743\ In future analysis, it may be possible to work with
                State-level information and incorporate State-specific registration
                requirements in the calculation of scrappage, but this correction is
                beyond the initial scope of this rulemaking analysis. Such an
                approach would be extraordinarily complicated as States can have
                very different registration schemes, and, further, the approach
                would also require estimates of the interstate and international
                migration of registered vehicles.
                ---------------------------------------------------------------------------
                 For new vehicles (vehicles where MY is greater than or equal to
                CY), the count of vehicles will be smaller than the count in the
                following year--not all of the model year cohort will have been sold
                and registered. For these new model years, Greenspan and Cohen assume
                that the Polk counts will capture all vehicles which were present in
                the given calendar year and that approximately one percent of those
                vehicles will be scrapped during the year. Importantly, this analysis
                begins modeling the scrappage of a given model year cohort in: CY =
                MY+2,\1744\ so that the adjustment to new vehicles is not relevant in
                the modeling because it only considers scrappage after the point where
                the on-road count of a given MY vintage has reached its maximum.
                ---------------------------------------------------------------------------
                 \1744\ Calculating scrappage could begin at CY=MY+1, as for most
                model year the vast majority of the fleet will have been sold by
                July 1st of the succeeding CY, but for some exceptional model years,
                the maximum count of vehicles for a vintage in the Polk data set
                occurs at age 2.
                ---------------------------------------------------------------------------
                (c) Polk Data Collection Changes
                 Prior to calendar year 2002, Polk vehicle registration data was
                collected as a single snapshot on July 1st of every calendar year. All
                vehicles that are in the registration database at that date are
                included in the dataset. For calendar years 2002 and later, Polk
                changed the timing of the data collection process to December 31st of
                the calendar year. In addition to changing the timing of the data
                collection, Polk updated the process to a rolling sample. That is, they
                consider information from other data sources to remove vehicles from
                the database that have been totaled in crashes before December 31st,
                but may still be active in State registration records.
                 The switch to a partially rolling dataset will mean that some of
                the vehicles scrapped in a calendar year will not appear in the dataset
                and their scrappage will wrongly be attributed to the year prior to
                when the vehicle is scrapped. While this is less than ideal, these
                records represent only some of the vehicles scrapped during crashes and
                scrappage rates due to crashes should be relatively constant over the
                2001 to 2002-time period. For these reasons, the agencies expect the
                potential bias from the switch to a partially rolling dataset to be
                limited. Thus, the Greenspan and Cohen adjustment applied does not
                change for the dataset complied from Polk's new collection procedures.
                As indicated in Figure VI-67, the scrappage counts computed from the
                old Polk snapshot series represent vehicles scrapped between July 1st
                of a given calendar year and the succeeding July 1st, and is computed
                for CY1976-2000. The new Polk snapshot series represents vehicles
                scrapped between December 31st of a given calendar year and the
                succeeding calendar year, and is computed for CY2002-2016.
                [[Page 24645]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.324
                 There is a discontinuity between the old and new methods so that
                the computed scrappage for calendar year 2001 represents the difference
                between the vehicle count reported in PolkCY2002 and
                PolkCY2001. PolkCY2001 represents all vehicles on
                the road as of July 1st, 2000, and PolkCY2002 represents all vehicles
                on the road as of December 31, 2001. For this one timespan, the
                scrappage will represent vehicles scrapped over a 17-month time period,
                rather than a year. For this reason, the CY2001 scrappage data point is
                dropped, and because of the difference in the time period of vehicles
                scrapped under the old and new collection schemes, an indicator for
                scrappage measured before and after CY2001 was considered; however,
                this indicator is not statistically significant, and is dropped from
                the preferred model.
                (d) Updated FRM Dataset
                 As noted in section II.A.1, some commenters expressed concern about
                the inability of the scrappage model to predict the scrappage rates of
                vehicles over age 20. The inability was in large part due to the
                limited data on the scrappage rates of older vehicles. NHTSA has worked
                with Polk/IHS to construct some of the historical registration
                databases using the new methodology for the purposes of other research.
                As a result, the agency has registration data using both Polk
                collection methods for CY's 2001-2012. Importantly, the old Polk
                dataset censored data on older vehicles, with CY's 1975-1993 including
                vehicles ages 0-15 and each successive CY past 1993 adding one
                additional age to the dataset--so that by 2000 ages 0-22 are included.
                The new datasets do not censor data on older vehicles, giving these
                datasets an advantage over the old datasets--for this reason, NHTSA
                uses as many years of the new data as is available.
                 The NPRM analysis also used all of the available data using the new
                methodology at the time of publication (CY's 2005-2015). Since the NPRM
                was published, NHTSA has gained access to registration data using
                Polk's new methodology for CY's 2002-2005 and CY's 2016-2017. Table VI-
                158 shows the calendars years of data in the NPRM and the final rule
                datasets by age, as well as the total number of data points for each
                age. There are a total of 330 and 420 data points for ages over 15 in
                the NPRM and final rule datasets, respectively. That represents almost
                a 30 percent increase in the number of data points for vehicles over
                15, and a 50 percent increase in the number of data points for the
                oldest vehicles considered in the dataset (ages 27-39). This additional
                data on older vehicles allows the new scrappage models to better
                predict the survival rates of older vehicles than the NPRM models.
                [[Page 24646]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.325
                (e) Models of the Gruenspecht Effect Used in Other Policy
                Considerations
                 This is not the first estimation of the `Gruenspecht Effect' for
                rulemaking policy considerations. In their Technical Support Document
                (TSD) for its 2004 proposal to reduce emissions from motor vehicles,
                CARB outlined how they utilized the CARBITS vehicle transaction choice
                model in an attempt to capture the effect of increasing new vehicle
                prices on vehicle replacement rates. They considered data from the
                National Personal Transportation Survey (NPTS) as a source of revealed
                preferences and a University of California (UC) study as a source of
                stated preferences for the purchase and sale of household fleets under
                different prices and attributes (including fuel economy) of new
                vehicles.
                 The transaction choice model represents the addition and deletion
                of a vehicle from a household fleet within a short period of time as a
                ``replacement'' of a vehicle, rather than as two separate actions.
                CARB's final data set consists of 790 vehicle replacements, 292
                additions, and 213 deletions; they do not include the deletions, but
                assume any vehicle over 19 years old that is sold is scrapped. This
                allowed CARB to capture a slowing of vehicle replacement under higher
                new vehicle prices. That said, because their model does not include
                deletions, it does not explicitly model vehicle scrappage, but assumes
                all vehicles aged 20 and older are scrapped rather than resold. CARB
                calibrated the model so that the overall fleet size is benchmarked to
                Emissions FACtors (EMFAC) fleet predictions for the starting year; the
                simulation then produced estimates that match the EMFAC predictions
                without further calibration.
                 The CARB study captures the effect on new vehicle prices on the
                fleet replacement rates, and offers some precedence for including an
                estimate of the Gruenspecht Effect. However, because vehicles that
                exited the fleet without replacement were excluded, the agencies do not
                learn the effect of new vehicle prices on scrappage rates where the
                scrapped vehicle is not replaced. New and used vehicles are
                substitutes, and therefore the agencies expect used vehicle prices to
                increase with new vehicle prices. And because higher used vehicle
                prices will lower the number of vehicles whose cost of maintenance is
                higher than their value, the agencies expect the replacements of used
                vehicles to slow, but the agencies also expect that some vehicles that
                would have been scrapped without replacement under lower new vehicle
                prices will now remain on the road because their value will have
                increased. The agencies' aggregate measures of the Gruenspecht effect
                includes changes to scrappage rates both from slower replacement rates,
                and from slower non-replacement scrappage rates.
                (f) Car Allowance Rebate System (`Cash for Clunkers')
                 On June 14, 2009, the Car Allowance Rebate System (CARS) became
                law, with the intent to stimulate the economy through automobile sales
                and accelerate the retirement of older, less fuel efficient and less
                safe vehicles. The program offered a $3,500 to $4,500 rebate for
                vehicles traded-in for the purchase of a new vehicle. Vehicles were
                subject to several program eligibility criteria: First, the vehicle had
                to be drivable and continuously registered and insured by the same
                owner for at least one year; second, the vehicle had to be less than 25
                years old; third, the MSRP had to be less than $45,000; and finally,
                the new vehicle purchased had to be more efficient than the trade-in
                vehicle by a specified margin. The fuel economy improvement
                requirements by body style for specific rebates are presented in Table
                VI-159.
                [[Page 24647]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.326
                 The program was originally budgeted for $1 billion dollars and to
                end on November 1, 2009, but that amount was spent far more quickly
                than expected and the program received an additional $1.85 billion in
                funding. Even with that additional funding, the program only lasted
                through August 25, 2009, expending $2.85 billion on 678,359 eligible
                transactions. To ensure that the replaced vehicles did not remain on
                the road, the vehicles were scrapped at the point of trade-in by
                destroying the engine. While the program resulted in the replacement of
                more vehicles and at a faster rate than expected, critics have argued
                that many of the trade-ins would have happened even if the program had
                not been in place, so that any economic stimulus to the automobile
                industry during the crisis cannot be attributable to the CARS program.
                Further, others have argued that forcing the scrappage of vehicles that
                could still remain on the road has negative environmental impacts that
                could outweigh any environmental benefits of the reduced fuel
                consumption from the accelerated retirement of these less efficient
                vehicles.
                 Li, Linn, and Spiller (2010) use Canada as a counterfactual example
                to identify the portion of CARS trade-ins attributable to the policy,
                i.e., trade-ins that would not have happened anywhere if the program
                were not in place. They argue that the Canadian market is largely
                similar to the U.S. market, in part based upon the fact that 13 to 14
                percent of households purchased new vehicles one year pre-recession in
                both countries. They also argue that the economic crisis affected the
                Canadian economy in a similar manner as it affected the U.S. economy.
                While they note that Canada offered a small rebate of $300 to vehicles
                traded in during January, 2009, hey further note that only 60,000
                vehicles were traded in under that program. Using those assumptions,
                Li, et al., applied a difference-in-difference methodology to isolate
                the effect of the CARS program on the scrappage of eligible vehicles.
                Li, et al., found a significant increase in the scrappage only for
                eligible U.S. vehicles, suggesting they isolated the effect of the
                policy. They conclude that of the 678,359 trade-ins made under the
                program, 370,000 of those would not have happened during July and
                August 2009. They conclude that the CARS program reduced gasoline
                consumption by 0.9-2.9 billion gallons, at $0.89-$2.80 per gallon
                saved.
                 The agencies find the evidence from Li, et al., persuasive toward
                the inclusion of a control for the CARS program during calendar year
                2009. The importance is discussed further both in the data section,
                Section VI.C.1.b)(3)(c)(ii), which provides more evidence for the
                effect of the CARS program, and in the model specifications Section
                VI.C.1.b)(3)(c)(iii), which describes the control used for the effect
                of the program. This ensures that the measurements of other determining
                factors are not biased by the exceptional scrappage observed in
                calendar year 2009.
                (iii) Updated Final Rule Modeling
                 The agencies contemplated all of the comments and suggestions made
                by commenters and, in response, have made several changes to final
                rule's model. First, the agencies changed the time-series strategy used
                in the model, as discussed in Section VI.C.1.b)(3)(c)(iii)(a). This
                change allows the agencies to simplify the models significantly,
                addressing commenters' concerns about potential overfitting of the
                model and difficulty of interpreting individual coefficient values
                (discussed in Section VI.C.1.b)(3)(b)(i)). Second, the agencies changed
                the modeling of the durability effect as discussed in Section
                VI.C.1.b)(3)(c)(iii)(c); this change reduces the reliance on the decay
                function and has the added benefit of addressing concerns about
                overfitting and out-of-sample projections discussed in Section
                VI.C.1.b)(3)(b)(i). Third, a portion of anticipated fuel savings from
                increased fuel economy are netted from new vehicle prices--meaning
                consumers are now assumed to value fuel economy at the time of purchase
                to a certain extent--as discussed in Section VI.C.1.b)(3)(c)(iii)(d).
                This change is in response to comments discussed in Section
                VI.C.1.b)(3)(b)(ii) and addresses inconsistent treatment of consumer
                valuation within the NPRM's analysis. Finally, the agencies consider
                the inclusion of additional or alternative variables in the scrappage
                model in response to comments discussed in Section
                VI.C.1.b)(3)(b)(iii). After extensive testing, the agencies concluded
                that these additional variables do not improve the model fits or would
                introduce autocorrelation in the error structures (see Sections
                VI.C.1.b)(3)(c)(iii)(e) and VI.C.1.b)(3)(c)(iii)(f) for further
                discussion). As such, the agencies rejected the additional terms
                suggested by commenters. Input from commenters was used to simplify the
                scrappage model, make it more consistent with modeling of new vehicle
                prices elsewhere in the analysis, and improve its predictions for the
                instantaneous scrappage rates of vehicles beyond age 20.
                (a) Changes to the Time Series Strategy
                 As discussed in Section VI.D.1.b)(3)(b)(i)(c), the agencies
                reconsidered the time series strategy for the final rule in response to
                comments. The first step in doing so is to test the time series
                properties of the dependent and independent variables. The agencies use
                the Augmented Dickey-Fuller (ADF) unit root test implemented in the
                `CADFtest' R package to test for stationarity.\1745\ The agencies find
                that the logistic scrappage rate is I(0), or stationary in levels.
                Since the dependent variable is stationary, there is no long-term trend
                in scrappage rates to capture. Lags of dependent variables need not be
                included, but their stationary forms should be used in the regressions.
                The following table summarizes the order of integration of each of the
                considered regressions; the
                [[Page 24648]]
                regression forms represent the form of the variable that is included in
                the considered models.\1746\ All the variables considered are either
                I(0) or I(1), meaning that they should be run in either levels or first
                differences, respectively. This significantly simplifies the
                regressions. Two unintended, positive outcomes of this change in time
                series strategy are that the coefficients on variables are easier to
                interpret and the models are less likely to be overfit. In this way,
                the shift to address concerns about the time series strategy (discussed
                in Section VI.D.1.b)(3)(b)(i)(c)) also addresses commenter concerns
                outlined in Section VI.D.1.b)(3)(b)(i)(a).
                ---------------------------------------------------------------------------
                 \1745\ Lupi, Claudio (2019, September 7). Package `CAFtest.'
                Retrieved from https://cran.r-project.org/web/packages/CADFtest/CADFtest.pdf.
                 \1746\ Note: Some of these variables were considered or added in
                response to comments presented in Sections I.A.1.a)(1)(b)(ii),
                I.A.1.a)(1)(b)(iii), and I.A.1.a)(1)(b)(iv), and may not be present
                in the NPRM.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.327
                (b) Final Rule Preferred and Sensitivity Specifications
                 After consideration of comments on, and subsequent peer review of,
                the NPRM analysis, the agencies updated the scrappage model
                specifications for the final rule. Section VI.C.1.b)(3)(c)(iii)(a)
                through VI.C.1.b)(3)(c)(iii)(f) discuss other considered specifications
                and variables. The equation below represents the final
                [[Page 24649]]
                form of the scrappage equation included in the central and sensitivity
                analysis:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.328
                 Here, ``S'' represents the instantaneous scrappage rate in a
                period, so that the dependent variable is the logit form of the
                scrappage rates. Logit models ensure that predicted values are
                bounded--in this case between zero and one. It is not possible to scrap
                more than all the remaining vehicles, nor fewer than zero percent of
                them, which is illustrated in the graph below:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.329
                [GRAPHIC] [TIFF OMITTED] TR30AP20.330
                Solving for instantaneous scrappage yields the following:
                [[Page 24650]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.331
                 In the equation above, [Sigma][beta]iXi represents the right-hand
                side of the above model specification. Within the right-hand side of
                the equation, Age represents the age of the model year cohort in a
                specific calendar year, defined by the Greenspan and Cohen adjustment
                discussed in Section VI.C.1.b)(3)(c)(ii)(b). The coefficient on the
                cubic age term is assumed to be zero for the van/SUV and pickup
                specifications as this term is not necessary to capture the general
                scrappage trend for these body styles. Share Remaining represents the
                share of the original cohort remaining at the start of the period.
                These two components represent the engineering portion of scrappage--
                the inherent durability of a model year and the natural life cycle of
                how vehicles scrap out of a model year cohort as the cohort increases
                with age. The determination of these specific forms is discussed in
                detail in Section VI.C.1.b)(3)(c)(iii)(g).
                 New Price--FS represents the average price of new vehicles minus 30
                months of fuel savings for all body styles. The central analysis
                assumes the coefficient on the age interactions for this term are zero
                for all body styles, but a sensitivity case allows the elasticity of
                scrappage to vary with age. Fuel Price represents the real fuel prices,
                weighted by fuel share of the model year cohort being scrapped. CP100M
                represents the cost per 100 miles of travel for the specific body style
                of the model year cohort being scrapped under the current period fuel
                prices and using fuel shares for that model year cohort. These measures
                capture the response of scrappage rates to new vehicle prices, fuel
                savings, and to changes in fuel prices that make the used model year
                cohort more or less expensive to operate. Because these measures are
                all I(1), as discussed above in 0, the first difference of all of these
                variables is used in modelling. The other specific modelling
                considerations that resulted in this form of modelling the new and used
                vehicles markets are discussed in Section VI.C.1.b)(3)(c)(iii)(d).
                 GDP Growth represents the GDP growth rate for the current period.
                This captures the cyclical components of the macro-economy. Section
                VI.C.1.b)(3)(c)(iii)(e) discusses how this specific measure was chosen,
                and what other measures were considered as alternative or additional
                independent variables.
                 CY2009 and CY2010 represent calendar year dummies for 2009 and 2010
                when the CARS program was in effect; this controls for the impact of
                the program. [Age = 25] represents an indicator for vehicles
                25 years and older. The interaction of the calendar year dummies with
                this indicator allows for the effect of the CARS program to be
                different for vehicles under 25 versus vehicles 25 and older. Since
                only vehicles under 25 were eligible for the program (see the
                discussion of the program in Section VI.C.1.b)(3)(c)(ii)(f)), this
                flexibility is important to correctly control for the program.
                 Finally, FE represents a set of model year fixed effects used to
                control for heterogeneity across different model years. This is related
                to the durability and engineering scrappage. The NPRM model did not
                include fixed effects because it fit a parametric relationship to model
                year as a continuous variable as a way to capture durability. This
                change in how the durability effect is modelled is discussed further in
                Section. Further, Section VI.C.1.b)(3)(c)(iii)(g) discusses trends in
                the fixed effects and how these are projected forward within the CAFE
                model.
                (c) Modeling Durability Trends Over Time
                 As noted in the NPRM, the durability of successive model years
                generally increases over time. However, this trend is not constant with
                vehicle age--the instantaneous scrappage rate of vehicles is generally
                lower for later vintages up to a certain age, but increases thereafter
                so that the final share of vehicles remaining converges to a similar
                share remaining for historically observed vintages. The NPRM
                parameterized this trend by using the natural log of the model year as
                a continuous variable interacted with a polynomial form of the age
                variable--this predicted an increasing but diminishing trend in vehicle
                durability for younger ages. The analysis for the final rule makes a
                change that allows more flexibility in durability trends. Below, the
                agencies consider the survival and scrappage patterns by body style.
                 Figure VI-69 to Figure VI-71 shows the survival and scrappage
                patterns of different vintages with vehicle age for cars, SUVs/vans and
                pickups, respectively. Cars have the most pronounced durability
                pattern. Figure VI-69 shows that newer vintages scrap slower at first,
                but that scrap more heavily so that the final share remaining of cars
                is more or less constant by age 25 for all vintages.
                [[Page 24651]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.332
                 SUVs/vans have a less pronounced durability pattern. Model year
                1980 actually lives longer than model years 1985 and 1990. This is
                likely due to a switch of SUVs/vans to be based on car chassis rather
                than pickup chasses over time. However, through the later model years,
                the durability trend is more like that of cars. The lack of a
                continuous trend in durability of SUVs/vans make how this trend is
                captured particularly important. Below the agencies discuss a change in
                how the durability trend is modelled for the final rule, which is more
                flexible than the NPRM model.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.333
                 There is no clear trend in durability for pickups. Like SUVs/vans,
                this makes parameterizing by using a form of vintage as a continuous
                variable problematic. Such a parametric form does not allow for each
                model year to have its own durability pattern.
                [[Page 24652]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.334
                 As noted above, the NPRM model used the natural log of model year
                as a continuous variable interacted with age to capture an increasing
                but diminishing trend of vehicle durability for the younger ages.
                However, enforcing a parametric form on a continuous model year
                excluded the possibility of including model year specific fixed effects
                and required that durability have a parametric trend with successive
                vintages. As seen above, SUVs/vans and pickups certainly do not follow
                such a trend, so that this constraint was too restrictive, at least for
                these body styles. The final rule analysis makes an adjustment that
                allows for an initial increase in the durability of a model year to
                persist, while including fixed effects and relaxing the parametric
                assumption.
                 Instead of regressing the natural log of the vintage share in the
                remaining models, shown in Table VI-161 through Table VI-163, the
                agencies use the share remaining in the previous period as an
                independent variable. Since the logistic instantaneous scrappage rate
                is stationary (it is independent of the previous periods' logistic
                instantaneous scrappage rate), the share remaining should not be
                endogenous. The share remaining models for the final rule include model
                year specific fixed effects and project a linear trend in durability by
                fitting a regression through the fixed effects. This latter part still
                requires a parametric assumption about durability (discussed in Section
                VI.C.1.b)(3)(c)(iii)(g)), but not while jointly estimating other
                coefficients. In this way, the other coefficients should not be biased
                by projecting the durability trend forwards in the implementation of
                the scrappage regressions within the CAFE model.
                BILLING CODE 4910-59-P
                [[Page 24653]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.335
                [[Page 24654]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.336
                [[Page 24655]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.337
                BILLING CODE 4910-59-C
                 As Table VI-161 shows, the NPRM specification and both the constant
                and the quadratic forms of the age interaction with the share remaining
                variable to capture the durability effect show evidence of
                autocorrelation. The linear form of the interaction of age and share
                remaining does not show evidence of autocorrelation and also has the
                lowest AIC and highest adjusted R-squared. For these reasons, this is
                the preferred specification of the durability effect. Since the share
                remaining coefficient is negative and larger than the positive
                coefficient on the share remaining interacted with age, a cohort that
                has a higher share remaining at an early age will have a lower
                instantaneous scrappage rate in this period until a certain age and
                then a higher scrappage rate after that age. To find the age where the
                sign of the share remaining coefficient will switch from predicting a
                lower instantaneous scrappage rate to a higher one, the agencies must
                take the ratio of the coefficient on the share remaining variable to
                the share remaining interacted with age--this suggests that at age 19,
                the sign of the share remaining variable flips. That is, the
                instantaneous scrappage rate of cars is predicted to be lower if the
                share remaining is higher until age 18, after which a higher share
                remaining predicts a higher instantaneous scrappage rate.
                 As Table VI-162 shows, the linear interaction of age and share
                remaining is the only specification of the durability effect for SUVs/
                vans that do not show autocorrelation in the error structure. The
                linear interaction of age and share remaining has the lowest AIC and
                highest R-squared; for this reason, this is the preferred specification
                of the durability effect for SUVs/vans. The signs for share remaining
                and share
                [[Page 24656]]
                remaining interacted with age show a similar trend as that to cars.
                Taking the ratio again of the share remaining to the share remaining
                interacted with age, for ages 0 to 18 a higher share remaining predicts
                lower instantaneous scrappage, and for ages beyond 18 it predicts a
                higher instantaneous scrappage rate.
                 As Table VI-163 shows, all but the NPRM specification of the
                durability effect for pickups do not show autocorrelation in the error
                structures. However, similar to cars and SUVs/vans, the linear
                interaction of age and share remaining has the lowest AIC and highest
                adjusted R-squared. For this reason, this is the preferred
                specification for all body styles. Taking the ratio of the coefficient
                on share remaining to share remaining interacted with age shows that a
                higher share remaining will predict a lower instantaneous scrappage
                rate in the next period for ages 0 through 14, but a higher
                instantaneous scrappage rate for ages 15 and older.
                 Using the preferred forms of the engineering scrappage rates for
                each body style as the reference point, Section VI.C.1.b)(3)(c)(iii)(d)
                considers different forms to predict the Gruenspecht effect for each
                body style. Section VI.C.1.b)(3)(c)(iii)(e) uses the preferred
                engineering and Gruenspecht forms to consider alternative macroeconomic
                variables to predict the effects of the business cycle. Finally,
                Section VI.C.1.b)(3)(c)(iii)(f) uses the preferred engineering,
                Gruenspecht and business cycle forms to consider the inclusion of other
                additional independent variables.
                (d) Modeling Impacts of New Vehicle Market on Used Scrappage Rates
                 Table VI-164 through Table VI-166 show the relationship between
                car, SUV/van, and pickup scrappage rates and changes in new vehicle
                price and fuel economies. The agencies consider two methods in response
                to comments outlined in Section VI.C.1.b)(3)(b)(ii). (1) changes in
                average new vehicle prices net of 30 months of fuel savings (consistent
                with the technology selection and sales model) and (2) change in
                average new vehicle prices, change in average fuel prices, changes in
                new vehicle cost per mile and changes in new vehicle fuel consumption.
                The agencies allow the elasticity of average new vehicle prices net of
                30 months of fuel savings to vary by age by including interaction
                terms.
                BILLING CODE 4910-59-P
                [[Page 24657]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.338
                [[Page 24658]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.339
                [[Page 24659]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.340
                 For all body styles, the specification of the Gruenspecht effect as
                the change in new vehicle prices net of fuel savings does not show
                signs of auto-correlated errors. However, for cars and vans/SUVs, the
                specification which separates the effect of new vehicle prices and fuel
                economy does show evidence of autocorrelation. For this reason, the
                changes in new vehicle fuel prices net of fuel savings is the preferred
                specification of the Gruenspecht effect.
                 The agencies consider the interaction of the change in average new
                vehicle prices with vehicle age. This relaxes an assumption that the
                elasticity of scrappage rates to change in new vehicle prices is
                constant. For cars and
                [[Page 24660]]
                vans/SUVs the linear interaction of change to new vehicle prices net of
                fuel savings show evidence of autocorrelation. The quadratic
                interaction of age with change in new vehicle prices shows
                autocorrelation with cars. For this reason, the agencies consider the
                constant elasticity of scrappage rates to changes in new vehicle prices
                to be the preferred specification (as the only specification that does
                not show evidence of autocorrelation for all body styles). However, the
                agencies do consider the quadratic form of the elasticity with age as a
                sensitivity case (even though there is evidence of autocorrelation (but
                only in the car specification)). This allows the agencies to test the
                impact of relaxing the assumption around constant elasticity on CAFE
                model outcomes.
                (e) Considering Alternative/Additional Macroeconomic Indicators
                 Table VI-167 through Table VI-169 show alternative macroeconomic
                indicators for cars, vans/SUVs and pickups, respectively. The agencies
                consider unemployment rate and per capita personal disposable income as
                alternatives to GDP growth rate to capture the cyclical component of
                the macro economy. The unemployment rate and the per capita personal
                disposable income are both I(1), so that the first difference of each
                is the form included. For the car and van/SUV specifications, the
                specifications replacing GDP growth rate show evidence of
                autocorrelation in the error structures. For this reason, the GDP
                growth rate is the preferred specification for the cyclical components
                of instantaneous scrappage rates, as in the NPRM models.
                 As discussed in Section VI.D.1.b)(3)(b)(iii)(c), some commenters
                were concerned with the exclusion of interest rates. In response, the
                agencies considered including the change in interest rates for the
                otherwise preferred specification. For vans/SUVs the model has a higher
                AIC and shows evidence of autocorrelation in the error structures. For
                pickups, the sign changes on the change in cost per mile when the
                interest rate is included, which would be an implausible result.
                Finally, the AIC for cars is nearly identical regardless as to whether
                the interest rate is included. For these reasons, the agencies continue
                to exclude the interest rate from the preferred specification.
                [[Page 24661]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.341
                [[Page 24662]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.342
                [[Page 24663]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.343
                [[Page 24664]]
                (f) Considering Other Additional Variables
                 Table VI-170 through Table VI-172 show specifications that consider
                additional variables not included in the preferred specifications. As
                discussed in Section VI.D.1.b)(3)(b)(iii)(a), some commenters
                criticized the fact that maintenance and repair costs were excluded
                from the scrappage models. In response to comments, and since the
                maintenance and repair costs are I(1), the agencies considered
                including the difference in maintenance and repair costs. When
                included, changes in maintenance and repair costs show the expected
                sign--when maintenance and repair costs are higher, instantaneous
                scrappage rates are predicted to be higher (as used vehicles are more
                expensive to maintain). When included, the AIC is higher for the car
                and van/SUV specifications. That is, including the change in
                maintenance and repair costs does not improve the fit of the models.
                Because of this, and because there is no obvious way to predict future
                change to maintenance and repair costs (as discussed in the NPRM), the
                preferred specification continues to exclude maintenance and repair
                costs.
                 As discussed in Section VI.D.1.b)(3)(b)(iii)(b), some commenters
                criticized the exclusion of steel and iron scrap prices from the
                scrappage models. In response to comments, and since this variable is
                also I(1), the agencies considered including the change in steel and
                iron scrap prices. When included, the AIC of cars and vans/SUVs is
                higher. Further, the car specification includes evidence of
                autocorrelation in the error structures. In addition, there is no known
                projection of steel and iron scrappage prices, so that the agencies
                would have to make projections to include this variable in the
                scrappage models. Accordingly, the central case continues to exclude
                steel and iron scrap prices.
                 As discussed in Section VI.D.1.b)(3)(b)(iii)(d), some commenters
                and peer reviewers suggested that controlling for aggregate measures of
                model year cohorts, such as performance, might correct some unexpected
                signs. The preferred specification already addresses these concerns.
                Further, because fixed effects are included for model years, the
                agencies cannot include aggregate model year specific attributes that
                are constant over the lifetime of the cohort. The agencies do consider
                the ratio of the average horsepower to weight of a model year cohort to
                the new vehicle cohort, as this will change along with changes to the
                horsepower to weight ratio over successive calendar years. Including
                this variable results in a higher AIC for cars and vans/SUVs and shows
                evidence of autocorrelation in the errors for these two body styles.
                For this reason, the preferred specification excludes this metric.
                 The agencies also considered including new vehicles sales directly
                as a predictor of instantaneous scrappage rates. Since new vehicle
                sales are I(1), the difference in new vehicle sales is the included
                form. Including the change in new vehicle sales results in a higher AIC
                for cars and vans/SUVs. It also introduces evidence of autocorrelation
                in the error structure for the car model, and reduces the effect of the
                change in fuel prices by two orders of magnitude for vans/SUVs. It
                seems unlikely that the magnitude of the effect of fuel prices would so
                drastically vary between body styles. For these reasons, the preferred
                specifications exclude the change in new vehicles sales. The agencies
                also considered including changes in vehicle stock, but this similarly
                did not improve the fit of the scrappage models--and doing so limited
                the ability to link the sales and scrappage models as some commenters
                suggested (see Sections (b)(iv)(a) and (b)(iv)(b)).
                [[Page 24665]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.344
                [[Page 24666]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.345
                [[Page 24667]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.346
                [GRAPHIC] [TIFF OMITTED] TR30AP20.347
                BILLING CODE 4910-59-C
                (g) Projecting Durability in the CAFE Model
                 The left graphs in Figure VI-72 through Figure VI-74 show the fixed
                effects for the preferred scrappage specifications for cars, vans/SUVs,
                and pickups, respectively. For all body styles there is a general
                downward trend in the fixed effects. This suggests an increase in the
                durability of successive model years. However, since the panel datasets
                are not balanced, there is likely potential bias for the fixed effects
                that include only certain ages. This makes projecting the durability
                increase from the fixed effects a little more complicated than merely
                fitting to all fixed effects. First, the agencies must determine what
                part of this trend is likely due to increases in vehicle durability
                (and should be projected forward) and which part of the trend may
                conflate other factors.
                 The right graphs in Figure VI-72 through Figure VI-74 show the
                average observed logistic scrappage rates by model year for all ages
                where data exists. As can be seen, the average observed scrappage rates
                decline dramatically for model years after 1996 for all body styles.
                There are two reasons this trend exists. First, as Figure VI-72 through
                Figure VI-74 show, the instantaneous scrappage rate generally follows
                an inverted u-shape with respect to vehicle age. The instantaneous
                scrappage rates generally peak between ages 15 and 20 for all body
                styles. Model year 1996 is the first model year which will be at least
                age 20 at the last date of available data (calendar year 2016). This
                means that all model years newer than 1996 have likely not yet reached
                the age where the instantaneous scrappage rate will be the highest for
                the cohort. Accordingly, the fixed effects could be biased downwards
                (consistent with the sharper downward slope in the fixed effects for
                most body styles for model years beyond 1996) because of the unbalanced
                nature of the panel, and not because of an actual increase in inherent
                vehicle durability for those model years.
                [[Page 24668]]
                 The second reason the average logistic scrappage rates for model
                years before 1996 is more stable is because each data point in the
                average has increasingly less effect on the average as more data
                exists. For model years 1996 and older there are at least 18 data
                points (we start the scrappage at age 2, by which point effectively all
                of a model year has been sold), and each will have a smaller effect on
                the average than for newer model years with fewer observations. For
                these reasons, the average observed logistic scrappage rate is more
                constant for model years before 1996. As a result, the agencies do not
                consider the trend in fixed effects after model year 1996 to rely on
                enough historical data to represent a trend in vehicle durability, as
                opposed to a trend in the scrappage rate with vehicle age.
                 In considering which model year fixed effects should be considered
                in projecting durability trends forward, another important factor is
                whether there are discrete shifts in the types of vehicles that are in
                the market or category of each body style over time. For cars, an
                increasing market share of Japanese automakers which tend to be more
                durable over time might result in fixed effects for earlier model years
                being higher. This trend is shown in the fixed effects in Figure VI-72,
                which follow a steeper trend before model year 1980.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.348
                 For vans/SUVs, earlier model years are more likely to be built on
                truck chassis (body-on-frame construction) instead of car chassis
                (unibody construction). Since pickups tend to be more durable, the
                earlier fixed effects are likely to be lower for vans/SUVs for earlier
                model years. The 1984 Jeep Cherokee was the first unibody construction
                SUV.\1747\ As Figure VI-73 shows, the fixed effects before 1986 show
                inconsistent trends; these are likely due to changes in what was
                considered a van/SUV over time. For this reason, the agencies build the
                trend of fixed effects from model years 1986 to 1996.
                ---------------------------------------------------------------------------
                 \1747\ https://www.autoguide.com/auto-news/2018/01/10-interesting-facts-from-the-history-of-the-jeep-cherokee.html.
                ---------------------------------------------------------------------------
                [[Page 24669]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.349
                [GRAPHIC] [TIFF OMITTED] TR30AP20.350
                 While the trend for pickups and cars could be extrapolated before
                1986, the agencies opt to keep the fixed effects included constant for
                all body styles. Thus, the projections are built from model year 1986
                to model year 1996 fixed effects. Table VI-173 below, shows the linear
                regressions shown as the line on the left side of Figure VI-70 through
                Figure VI-72. The durability cap represents the last model year where
                the durability trend is assumed to persist. The agencies cap the
                durability impacts at model year 2000, as data beyond this point does
                not exist for enough ages to determine if durability has continued to
                increase since this point. The implication of this cap, is that model
                years after 2000 are assumed to have the same initial durability as
                model year 2000 vehicles. Since there is a limit to the potential
                durability of vehicles, this acts as a bound on this portion of the
                scrappage model.
                [[Page 24670]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.351
                 The durability projections enter the scrappage equation in the CAFE
                modelling in accordance to the following equation:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.352
                 The intercept enters as a constant added to the predicted logistic
                of the instantaneous scrappage rate. The model year slope enters as the
                model year for all model years older than 2000 and enters as 2000 for
                all model years 2000 and newer.
                 Once the predicted logistic scrappage rate is calculated in the
                CAFE model (including the projections of the fixed effect portion of
                the equation), the future population of model year cohorts can be
                predicted. The instantaneous scrappage can be calculated directly from
                S. It identifies the share of remaining vehicles in each calendar year
                that are scrapped in the next year. The population of vehicles in the
                next calendar year can be calculated as follows:
                Populations MY,CY +1 = Population MY,CY *(1 -SMY,CY).
                 This process is iteratively calculated at the end of the CAFE model
                simulation to determine the projected population of each model year in
                each future calendar year. This allows the calculation of vehicle miles
                travelled, fuel usage, pollutant and CO2 emissions, and
                associated costs and benefits. The CAFE model documentation released
                with this final rule further details how the scrappage model is
                projected within the simulations.
                (d) Updates to the Decay Function
                 The scrappage models described above fit the historical data of car
                and truck scrappage well, but when used to project the scrappage of
                future model years they over-predict the remaining cars and trucks for
                ages greater than 30 in an unrealistic manner. Nearly six percent of
                the MY2015 van/SUV fleet and eight percent of the pickup fleet is
                projected to persist until age 40. This is unrealistic, and likely due
                to the fact that the agencies do not observe enough model years for
                those ages and over-predict the impact of durability increases for
                those ages. For this reason, the agencies are using the curves with an
                accelerated decay function to predict instantaneous scrappage beyond
                age 30 for pickups and SUVs/vans. The implementation and parameter
                stricture of the decay function have not changed since the NPRM model.
                Table VI-174, below, shows the inputs used for the final rule analysis.
                [[Page 24671]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.353
                 The final survival rate has not changed since the NPRM, but the
                input Decay age has changed. In the NPRM, the decay function was
                specified to begin after age 20, while the decay function begins after
                age 30 in the final rule analysis. This input change was possible
                because the scrappage model for the final rule predicts shares
                remaining in line with observed historical trends through age 30,
                rather than through age 20. This improvement in the model fits for
                older ages is driven both by the shift of the modelling of the
                durability effect discussed in Section VI.D.1.b)(3)(a)(g) and the
                increase in available data on the scrappage rates of older vehicles
                discussed in Section VI.C.1.b)(3)(c)(ii)(d). Overall, this outcome
                suggests that the final rule model predicts the scrappage rates of
                older vehicle better than the NPRM model.
                 As in the NPRM, the decay function is implemented in the model
                using the following conditions:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.354
                Where:
                t = (age + 1 - b15
                And:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.355
                 Here, the population for ages beyond the start age of the decay
                function depends on the population of the cohort at that start age and
                the final share expected for that body style at age 40. The rate of
                decay necessary to make the final population count equal that observed
                in the historical data is applied.
                (4) The Rebound Effect in the NPRM
                 The fuel economy rebound effect--a specific example of the well-
                documented energy efficiency rebound effect for energy-consuming
                capital goods--refers to the tendency of motor vehicles' use (as
                measured by vehicle-miles traveled, or VMT) to increase when their fuel
                economy is improved and, as a result, the cost per mile (CPM) of
                driving declines. Amending and establishing CAFE and CO2
                standards at a lower degree of stringency than the baseline level will
                lead to comparatively lower fuel economy for new cars and light trucks,
                thus increasing the amount of fuel consumed to travel each mile. The
                resulting increase in CPM will lead to a reduction in VMT over the
                lifetime of new vehicles, an example of the rebound effect working in
                reverse. In the NPRM, the agencies assumed a fuel rebound effect of 20
                percent, meaning that a 5 percent decrease in fuel economy would result
                in a one percent decrease in the annual number of miles driven at each
                age over a vehicle's lifetime.
                 Many of the comments received on different components of the CAFE
                model can be traced back to the agencies' rebound selection. The
                agencies recognize that the value selected for the rebound effect
                influences overall costs and benefits associated with the regulatory
                alternatives under consideration as well as the estimates of lives
                saved under various regulatory alternatives, and that the rebound
                estimate, along with fuel prices, technology costs, and other
                analytical inputs, is part of the body of information that agency
                decision-makers have considered in determining the final levels of the
                CAFE and CO2 standards. The agencies also note that the
                rebound effect diminishes the economic and environmental benefits
                associated with increased fuel efficiency.
                 For the analysis supporting the NPRM, the agencies conducted a
                thorough re-examination of the basis for the estimate of the fuel
                economy rebound effect used to analyze the impacts of CAFE and
                CO2 emission standards for model years 2012-16 and 2017-21.
                This was prompted by three developments. First, more recent updates of
                the 2007 study by Small and Van Dender that had provided the basis for
                assuming the 10 percent rebound effect used in those previous analyses
                reported larger values. Second, projected growth in the income measure
                used in those authors' 2007 study, which was anticipated to reduce the
                magnitude of the rebound effect over the future period spanned by those
                analyses, did not occur during the decade following the 2007 study's
                publication. Finally, extensive new research on the rebound effect had
                become available since those previous
                [[Page 24672]]
                analyses were conducted, and while its findings were mixed, many of
                those more recent studies reported values significantly above the
                agencies' previous 10 percent estimate.
                 In the NPRM, the agencies first summarized estimates of the fuel
                economy rebound effect for light-duty vehicles in the U.S. from studies
                conducted through 2011, when the agencies originally surveyed research
                on this subject. As the accompanying discussion in the proposal
                indicated, the research available through 2011 collectively suggested
                that the rebound effect was likely to fall in the range from 20 percent
                to 25 percent, although the then-recent study by Small and Van Dender
                (2007) pointed to smaller values, particularly for future years. The
                agencies then identified 16 additional studies of the rebound effect
                that had been conducted since their original survey, and the NPRM
                discussed the various approaches they used to measure the magnitude of
                the rebound effect, their data sources and estimation procedures,
                reported findings, and strengths and weaknesses of each study.
                 Based on this re-examination, the agencies concluded that currently
                available evidence did not appear to support the 10 percent estimate
                relied upon in previous rules, and identified a value of 20 percent as
                more representative of the totality of evidence, including both the
                research covered by the earlier and more recent studies examined in the
                NPRM. While acknowledging the wide range of estimates reported in more
                recent research--which extended from zero to more than 80 percent--the
                agencies noted that the central tendency of recent estimates appeared
                to lie in the same 20-25 percent range suggested by their extensive
                review of earlier research. The agencies also recognized that a 20
                percent estimate differed markedly from the 10 percent estimate used in
                the regulatory analyses for the 2010 and 2012 final rules, but noted
                that it represented a return to the value NHTSA originally used to
                analyze the impacts of CAFE standards for model years prior to 2011.
                (a) Comments on the Rebound Effect Used in the NPRM
                 The agencies received numerous comments on the decision to revise
                their previous estimate of the rebound effect, virtually all of which
                echoed a few common arguments. First, commenters generally agreed that
                the most appropriate measure for the agencies to rely on is the current
                long-run fuel economy rebound effect for U.S., although a few suggested
                that using an estimate of its short-run value might be
                preferable.\1748\ However, many commenters argued that some of the more
                recent studies the agencies relied upon to support the revised 20
                percent estimate may have limited relevance to the appropriate measure
                for analyzing the current rule, and that the agencies should place more
                emphasis on those that commenters asserted were more appropriate to
                rely upon.
                ---------------------------------------------------------------------------
                 \1748\ See, e.g., RFF, Comments, NHTSA-2018-0067-11789, at 30.
                For an thorough example of the arguments made for a short- to
                medium-term rebound effect, see generally IPI, Appendix, NHTSA-2018-
                0067-12213, at 61.
                ---------------------------------------------------------------------------
                 To identify the most relevant research, some commenters proposed
                applying various selection criteria to choose which studies were most
                appropriate to rely on when estimating the value of the rebound effect
                to use in this analysis. While commenters proposed using certain
                criteria as ``filters''--that is, to eliminate any studies that did not
                meet those criteria--they also suggested applying other criteria to
                emphasize studies with particular features they argued made them more
                relevant to identifying the current value of the rebound effect for the
                U.S.\1749\ Among these suggested criteria were the following:
                ---------------------------------------------------------------------------
                 \1749\ See, e.g., IPI, Appendix, NHTSA-2018-0067-12213, at 58-
                64; EDF, Analysis of the Value and Application of the Rebound
                Effect, NHTSA-2017-0069-0574, at 16-19; California Office of the
                Attorney General et al., Attachment 1, NHTSA-2017-0069-0625, at 8;
                States and Cities, Attachment 1, Docket No. NHTSA-2018-0067-11735,
                at 78; RFF, Comment, NHTSA-2018-0067-11789, at 3; CARB, Detailed
                Comments, NHTSA-2018-0067-11873, at 120; Aluminum Association,
                Comments, NHTSA-2018-0067-11952, at 5; NCAT, Appendix A, NHTSA-2018-
                0067-11969, at 34; and North Carolina Department of Environmental
                Quality, Comments, NHTSA-2018-0067-12025, at 12; among others. EPA's
                Science Advisory Board shared similar policy opinions. SAB at 26-27.
                ---------------------------------------------------------------------------
                 Exclude estimates based upon data from outside the U.S.;
                 Include only estimates based upon ``more recent'' data,
                usually taken to mean those published within approximately the last
                decade;
                 View estimates based on the U.S. 2009 National Household
                Travel Survey skeptically, or exclude them from consideration
                completely;
                 Emphasize estimates derived from vehicle use and fuel
                economy data spanning multiple years (such as aggregate time-series or
                panel data), while according less weight to those based on a single-
                year cross section (such as most household survey data);
                 Emphasize estimates of the rebound effect that measure the
                response of vehicle use to variations in fuel efficiency, rather than
                in fuel cost per mile driven or fuel price per gallon;
                 Emphasize estimates that rely on identification strategies
                that account for potential endogeneity in fuel economy (as would
                result, for example, if households with high levels of demand for
                travel purchase vehicles with higher fuel economy);
                 Emphasize estimates based on measures of vehicle use
                obtained from odometer readings; and
                 Emphasize estimates that explicitly control for purchase
                prices of new vehicles in order to account for changes in new vehicle
                prices due to CAFE standards.
                 A few commenters illustrated how applying these criteria could
                reduce the large number of published studies of the rebound effect to a
                limited subset that suggested a smaller value than 20 percent.\1750\
                Using multiple criteria to exclude or de-emphasize studies that did not
                meet all of those applied, these commenters argued that the most
                appropriate value for this analysis was closer to (or possibly even
                below) the 10-percent estimate the agencies used for the previous
                rulemaking.\1751\ However, one commenter noted that applying these
                criteria individually to exclude any estimates not meeting them had
                almost no effect on formal measures of the central tendency (the mean
                and median values) of the remaining estimates.\1752\ This commenter
                suggested that only by applying two or more of these criteria jointly
                and excluding any studies that did not meet all of those applied could
                the universe of research on the rebound effect be reduced to a subset
                supporting a lower value than the 20 percent figure the agencies used
                to analyze the NPRM.
                ---------------------------------------------------------------------------
                 \1750\ See, e.g., Gillingham, Nera-Trinity Responses, NHTSA-
                2018-0067-12403, at 16-30.
                 \1751\ See supra note 1749.
                 \1752\ Alliance of Automobile Manufacturers, Attachment 3,
                NHTSA-2018-0067-12386, at 15-17.
                ---------------------------------------------------------------------------
                 Commenters also identified several additional recent studies that
                were not included in the agencies' review of recent evidence for the
                NPRM, and suggested revised interpretations of the empirical estimates
                reported in two studies that had been included (the agencies also
                clarified a third). Commenters represented these additional studies as
                generally supporting lower values than the agencies' revised 20 percent
                estimate, although this appeared to be a selective interpretation of
                some of the results they reported.\1753\ Other commenters asserted
                [[Page 24673]]
                that the two most commonly-demonstrated features of the rebound effect
                are that it varies directly with fuel prices and declines in response
                to rising income over time, and argued that the latter suggests that a
                declining value is likely to be more appropriate for analyzing the
                longer-term impacts of this final rule.\1754\
                ---------------------------------------------------------------------------
                 \1753\ For example, some commenters (e.g., Gillingham, Nera-
                Trinity Responses, NHTSA-2018-0067-12403, Table 2, at 24)
                represented the recent analysis of vehicle use data from Texas by
                Wenzel and Fujita as reporting a rebound effect of 8-15 percent,
                which appears to be based on those authors' estimates of the
                response of vehicle use to changes over time in fuel prices alone.
                This range appears to ignore those same authors' estimates of the
                sensitivity of vehicle use to variation in fuel costs per mile,
                which provides a more direct measure of the fuel economy rebound
                effect because it incorporates fuel economy as well as fuel prices.
                Those estimates range from 7-40 percent, with most falling in the
                interval from 15-25 percent; see generally, Wenzel and Fujita
                (2018), Table 4-12, at 38.
                 \1754\ See particularly Small, NHTSA-2018-0067-7789, at 3.
                ---------------------------------------------------------------------------
                 Some commenters suggested that the rebound effect is asymmetrical,
                meaning that drivers are more responsive to price increases than price
                decreases. These commenters asserted that the asymmetrical nature of
                the rebound effect favors a lower estimate.\1755\ Similarly, other
                commenters suggested that the rebound effect had to be lower than 20
                percent because congestion would limit additional driving.\1756\
                ---------------------------------------------------------------------------
                 \1755\ EDF, Analysis of the Value and Application of the Rebound
                Effect, NHTSA-2017-0069-0574, Comment, 37-38.
                 \1756\ For example, the South Coast Air Quality Management
                District argued that, logistically, rebound cannot exist in Southern
                California because ``any rebound effect will only worsen congestion
                in Southern California, such a result cannot be predicted.'' NHTSA-
                2018-0067-11813 at 45.
                ---------------------------------------------------------------------------
                (b) Agencies' Response to Comments on the NPRM
                 In response to commenters who argued that the agencies' estimate of
                the rebound effect should be reduced, because research that
                incorporates the effects of congestion or allows asymmetrical responses
                to price changes suggests lower values, the agencies note that, for the
                final rule's analysis, those factors would be difficult and perhaps
                even inappropriate to incorporate in their analysis. In the case of
                congestion, the agencies note that their estimate of the rebound
                effect--like research on the rebound effect in general--represents a
                change in aggregate VMT, and has no clear implication about how that
                change in travel is likely to be distributed over times of the day or
                geographic locations.\1757\
                ---------------------------------------------------------------------------
                 \1757\ The agencies' estimate of increased congestion costs
                associated with additional driving due to the rebound effect
                implicitly assumes that increased driving will be distributed
                according to current travel patterns, producing similar proportional
                increases at various hours of the day and geographic locations. Such
                an assumption is made out of necessity to model congestion and
                noise; the agencies acknowledge that the rebound effect is unlikely
                to affect vehicle use in such a uniform fashion.
                ---------------------------------------------------------------------------
                 As for possible asymmetry in the response of vehicle use to changes
                in driving costs, the CAFE model applies a single estimate of the
                rebound effect for all changes in cost-per-mile, and cannot accommodate
                a rebound effect that varies with the magnitude or direction of changes
                in driving costs, which would be necessary to capture asymmetrical or
                non-linear responses to cost changes. The agencies also remind
                commenters that this rule will result in an increase in driving costs,
                for which the research they cite generally suggests a larger value of
                the rebound effect is appropriate. In any case, using a different
                estimate of the rebound effect to analyze impacts of raising and
                lowering standards would not promote consistency or replicability, both
                desirable characteristics of regulatory analysis.
                 The agencies decided to include the previously omitted studies
                raised by commenters in their rebound analysis supporting the final
                rule, but do not feel that they suggest a value different from that
                used to analyze the proposal. Adding these studies to the list of
                recent research discussed in the NPRM, deleting one unpublished
                analysis, and revising the entries for selected studies to reflect more
                accurately the values reported by their authors produces a more
                extensive catalog of recent research, which is summarized in Table VI-
                175 below.
                [[Page 24674]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.356
                 As evidenced in Table VI-175, studies continue to have a wide range
                of estimates, but collectively the research looks remarkably similar to
                the historical estimates. The newer studies suggest that a plausible
                range for the rebound effect is 10-50 percent. The central tendency of
                this range appears to be roughly 30 percent.
                 In response to comments proposing the application of specific
                criteria to eliminate or reduce the consideration accorded to studies
                without certain features thought to increase the relevance of their
                findings, the agencies note that measuring the rebound effect is both
                conceptually and technically challenging, and that analysts have used
                many different approaches in an attempt to surmount these challenges.
                The agencies' view is that each of the studies included in its previous
                survey and in Table VI-175 above provides some useful evidence on the
                likely value of the rebound effect, and while all have some conceptual
                or theoretical weaknesses, each nevertheless provides some useful
                insights into the appropriate magnitude of the rebound effect for the
                current analysis.
                 As a general approach to estimating parameters that are uncertain,
                the agencies prefer to rely on the totality of empirical evidence,
                rather than restricting the available evidence by categorically
                excluding or according less weight to that do not meet selection
                criteria that may not be widely agreed upon. From this perspective,
                analyses that rely on different measurement approaches, data sources,
                and estimation procedures all have the potential to provide valuable
                information for choosing the most representative value. The agencies
                also view sound measurement strategies and careful empirical analysis
                using reliable data as equally important features when compared to a
                study's vintage or geographic scope. Examining the widest possible
                range of research also enables useful comparisons and ``cross-checks''
                on the estimates that individual studies report.
                 Notwithstanding this more inclusive perspective, the agencies
                endorse certain of the characteristics preferred by commenters,
                although the agencies view them as indicators of a strong study, rather
                than a bright-line test of whether to accord it any weight rather than
                discarding it from consideration. Specifically, the agencies agree with
                many commenters that both the extended time span encompassed by their
                analysis of the impacts of CAFE and CO2 standards and the
                long expected lifetimes of vehicles subject to this final rule means
                that estimates of the long-run rebound effect are most relevant for
                purposes of the final rule
                [[Page 24675]]
                analysis.\1758\ The agencies also agree with commenters that estimates
                based upon more recent data are generally preferable, but nevertheless
                note that older studies that combine careful analysis with unusually
                reliable or novel data can offer evidence that remains useful.\1759\
                The agencies also concur with some commenters' argument that estimates
                of the rebound effect that are derived from the relationship of vehicle
                use to fuel efficiency, rather than to fuel cost per mile or gasoline
                prices, are likely to provide more direct measures of the fuel economy
                rebound effect itself, which is the desired parameter for the purposes
                of this analysis. Finally, the agencies generally view identification
                strategies and econometric methods that account or control for
                potential endogeneity in fuel economy as likely to provide more
                reliable estimates.
                ---------------------------------------------------------------------------
                 \1758\ Most of the vehicles affected by today's standards will
                remain on the roads for at least a decade, with a significant
                fraction surviving considerably longer. As such, long-run estimates
                are more likely to reflect the lifetime mileage accumulation of the
                new fleet than either short-run or medium-run estimates.
                Furthermore, a long-run rebound estimate better reflects the
                cumulative impact of successive CAFE and CO2 standards
                such as those adopted by the agencies beginning as early as 2010.
                 \1759\ One example is the study by Greene et al. (1999), which
                used advanced econometric analysis of unusually detailed and
                reliable data on household demographic and economic characteristics,
                household members' use of individual vehicles, and fuel purchases to
                estimate the response of households' use of individual vehicles to
                their actual on-road fuel economy, and its implications for total
                household driving.
                ---------------------------------------------------------------------------
                 In contrast, the agencies view other criteria proposed by
                commenters as unnecessarily restrictive, particularly when they are
                used to disqualify otherwise informative research from consideration.
                For instance, categorically excluding from consideration non-U.S.
                studies--which the agencies agree should be treated cautiously--seems
                likely to exclude useful evidence, particularly recognizing some of
                those studies' access to unusually reliable data on vehicle use and
                fuel economy and use of sophisticated econometric analysis. In
                addition, many foreign studies have been conducted in nations with
                income levels comparable to the U.S., and in some cases levels of auto
                ownership that are beginning to approach U.S. levels. Furthermore,
                driving habits throughout the U.S. are not homogenous. In fact, some
                regions in the U.S. may exhibit driving habits that more closely
                resemble those in some foreign nations than driving patterns in other
                regions of the U.S.\1760\
                ---------------------------------------------------------------------------
                 \1760\ For example, drivers in Manhattan, Kansas likely respond
                to changes in fuel prices and fuel economy differently than drivers
                in Manhattan, New York.
                ---------------------------------------------------------------------------
                 In response to some commenters' recommendation that the agencies
                more heavily weigh studies using data spanning multiple years than
                those relying on data for a single year, the agencies note that
                household surveys, the most common form of data for a single year,
                provide cross-sectional variation in vehicle use and other
                characteristics that is helpful for identifying the desired long-run
                measure of the rebound effect. Household surveys are also an important
                source of information that enable analysts to measure the response of
                individual vehicles' use to variation in their fuel economy, while also
                controlling adequately for household characteristics that affect travel
                patterns and vehicle use. Household survey data can also enable
                analysts to identify the vehicle substitution patterns within multiple-
                vehicle households that are increasingly responsible for producing the
                rebound effect, while even modest-scale household surveys include many
                more observations than are typically available in aggregate time-series
                or panel data.
                 These strengths of course need to be balanced against the potential
                drawbacks of relying on a one-time snapshot of households' behavior
                during a single time period. Surveys also frequently rely on owner-
                reported estimates of vehicle use and usually require analysts to
                impute vehicles' fuel economy ratings from limited and sometimes
                incomplete information on the specific vehicle models and vintages that
                households report owning. One result is that estimates of the rebound
                effect derived from household survey data may be based on inaccurate
                estimates of vehicles' use and fuel economy. Assuming the errors in
                measuring these variables are random, the errors would increase the
                uncertainty surrounding the estimates of the rebound effect, but would
                not bias the estimate.
                 In contrast, studies using nationwide aggregate or average measures
                of vehicle use and fuel economy or fuel cost rarely provide adequate
                independent variation to support reliable estimates of the response of
                vehicle use to variation in fuel economy, even where extended time
                series are available, while State-level measures of these variables are
                subject to potentially extreme measurement error that can compromise
                estimates of these relationships.\1761\ Moreover, controlling for the
                many other demographic and economic factors likely to affect vehicle
                use using national or even State-level aggregate data presents
                difficult challenges.
                ---------------------------------------------------------------------------
                 \1761\ For example, State-level estimates of travel by
                individual vehicle classes such as cars and light-duty trucks often
                exhibit implausible year-to-year variability due to the measurement
                procedures states employ and the difficulty of distinguishing among
                different types of vehicles. At the same time, the potential
                geographic ``mismatch'' between State-level vehicle use and fuel
                sales complicates any effort to measure fuel efficiency or fuel
                costs at the State level.
                ---------------------------------------------------------------------------
                 Finally, the agencies note that no single selection criterion
                proposed by commenters noticeably reduces the central tendency
                displayed by the universe of estimates of the rebound effect, and
                multiple criteria must be applied simultaneously to restrict the
                universe to a subset of studies that points toward a significantly
                lower value than the 20 percent estimate the agencies used to analyze
                the proposal. Applying multiple criteria drastically reduces the number
                of studies that remain available to guide the agencies, while at the
                same time discarding potentially valuable information provided by
                research those criteria exclude from consideration.\1762\ Doing so
                would thereby necessarily reduce the confidence that the agencies can
                have in the resulting estimate.
                ---------------------------------------------------------------------------
                 \1762\ As an illustration, excluding non-U.S. studies reduces
                the number of recent analyses surveyed in the proposal from 15 to 8,
                while eliminating those that rely on the 2009 National Household
                Travel Survey (NHTS) discards another 5, leaving only 3.
                ---------------------------------------------------------------------------
                 Regarding some commenters' assertion that the rebound effect is
                known to decline in response to rising income, and that this
                observation warrants using a lower value for long-term future
                evaluation of the standards' effects, the agencies note that some
                evidence based on household and vehicle use surveys suggests that the
                rebound effect increases with the level of household vehicle ownership,
                which is itself highly correlated with income. Together with forecasts
                of limited future growth in most measures of U.S. household income,
                this finding casts some doubt on whether the rebound effect is likely
                to decline over the time period spanned by the agencies'
                analysis.\1763\
                ---------------------------------------------------------------------------
                 \1763\ For example, the widely cited IHS Markit Long-Term
                Macroeconomic Outlook for Spring 2019 projects that per Capita
                disposable personal income in the U.S. will grow at 1.6 percent
                annually over the next 30 years; see Federal Highway Administration,
                Forecasts of Vehicle Miles Traveled (VMT): Spring 2019, Table 2,
                available at https://www.fhwa.dot.gov/policyinformation/tables/vmt/vmt_forecast_sum.cfm.
                ---------------------------------------------------------------------------
                 The agencies also note that one of the studies cited in Table VI-
                175 above (DeBorger et al., 2016) finds that the decline in the fuel
                economy rebound effect with income reported in the
                [[Page 24676]]
                earlier analysis by Small and Van Dender (2007)--on which the agencies
                relied in reducing their original estimate of the rebound effect to 10
                percent--results entirely from a reduction in drivers' sensitivity to
                fuel prices as their incomes rise, rather than from any effect of
                rising income on the sensitivity of vehicle use to fuel economy.\1764\
                This latter measure--which DeBorger et al. find is quite small and has
                not changed significantly as incomes have risen over time--is the most
                direct measure of the fuel economy rebound effect, so their analysis
                calls into question its widely-assumed sensitivity to income.
                ---------------------------------------------------------------------------
                 \1764\ DeBorger, B., Mulalic, I., and Rouwendal, J., ``Measuring
                the rebound effect with micro data: A first difference approach.''
                Journal of Environmental Economics and Management, 79 (2016), at 1-
                17.
                ---------------------------------------------------------------------------
                 Finally, because there is not a clear consensus around a single
                rebound estimate within the literature, the agencies believe it is
                important to benchmark their analysis with other large scale surveys of
                the literature published by neutral observers. In one early survey,
                Greening, Greene, and Difiglio (2000) reviewed studies that estimated
                the rebound effect for light-duty vehicles in the U.S., concluding that
                those relying on aggregate time-series data found it was likely to
                range from 10-30 percent, while those using cross-sectional analysis of
                household vehicle use suggested a larger rebound effect, in the range
                of 25-50 percent.\1765\ Sorrell et al. (2009) found that the magnitude
                of the rebound effect for personal automobile travel is likely to fall
                in the 10-30 percent range, with some evidence suggesting that the
                lower end of that range might be most appropriate.\1766\
                ---------------------------------------------------------------------------
                 \1765\ Greening, L.A., Greene, D.L. and Difiglio, C., ``Energy
                efficiency and consumption--the rebound effect--a survey.'' Energy
                Policy, Vol. 28 (2000), at 389-401.
                 \1766\ Sorrell, Steve, John Dimitropoulos, and Matt Sommerville,
                ``Empirical Estimates of the Direct Rebound Effect: A Review,''
                Energy Policy 37(2009), at 1356-71.
                ---------------------------------------------------------------------------
                 Most recently, a meta-analysis of 74 published studies of the
                rebound effect conducted by Dimitropoulos et al. (2018) estimated that
                the long-run rebound effect ranges from 22-29 percent when measured by
                the response of vehicle use to variation in fuel efficiency (the
                authors' preferred measure), from 21-41 percent when it is measured
                using the variation fuel cost per unit distance, and from 25-39 percent
                using fuel price per gallon.\1767\ The authors concluded that ``the
                magnitude of the rebound effect in road transport can be considered to
                be, on average, in the area of 20%,'' but noted that the long-run
                estimate was about 32 percent.\1768\ A subsequent published study by
                these same authors (Dimitropoulos et al. (2018)) concludes that the
                most likely estimate of the long-run rebound effect is in the range of
                26-29 percent, but could range from as low as 15 percent to as high as
                49 percent at income levels, development densities, and fuel prices
                that are currently representative of the U.S.\1769\
                ---------------------------------------------------------------------------
                 \1767\ Dimitropoulos, Alexandros, Walid Oueslati, and Christina
                Sintek, ``The rebound effect in road transport: a meta-analysis of
                empirical studies,'' Paris, OECD Environment Working Papers, No.
                113; see esat Table 5, at 25 (and accompanying discussion).
                 \1768\ Id. at 28.
                 \1769\ Dimitropoulos, Alexandros, Walid Oueslati, and Christina
                Sintek, ``The Rebound Effect in Road Transport: A Meta-Analysis of
                Empirical Studies,'' Energy Economics 75 (2018), at 163-79; see esat
                Table 4, at 170, Table 5, at 172 (and accompanying discussion), and
                Appendix B, Table B.V., at 177.
                ---------------------------------------------------------------------------
                (c) Selecting a Value of the Rebound Effect for Evaluating the Impacts
                of This Rule
                 After reviewing the evidence on the rebound effect previously
                summarized in the NPRM, comments the agencies received, other recent
                studies of the rebound effect that were not summarized in the NPRM but
                suggested by commenters, and published surveys of literature, a
                reasonable case can be made to support values of the rebound effect at
                least as high as 30 percent. The totality of evidence, without
                categorically excluding studies on grounds that they fail to meet
                certain criteria, and evaluating individual studies based on their
                particular strengths, suggests that a plausible range for the rebound
                effect is 10-50 percent. The central tendency of this range appears to
                be at or slightly above its midpoint, which is 30 percent. Considering
                only those studies that the agencies believe are derived from unusually
                reliable data, employ identification strategies that are likely to
                prove effective at isolating the rebound effect, and apply rigorous
                estimation methods suggests a range of approximately 10-45 percent,
                with most of their estimates falling in the 15-30 percent range.\1770\
                ---------------------------------------------------------------------------
                 \1770\ As indicated previously, these are the selection criteria
                proposed by commenters with which the agencies concur. In
                chronological order, the studies the agencies feel best meet those
                criteria include Greene et al. (1997), Small and Van Dender (2007)
                and subsequent updates by Hymel, Small, and Van Dender (2010, 2015),
                Linn (2016), Anjovic and Haas (2012), Gillingham (2014), and
                DeBorger et al. (2016). Other studies the agencies believe warrant
                serious consideration because they offer some or most of these same
                advantages include those by Liu et al. (2014), Knittel and Sandler
                (2018), and Wenzel and Fujita (2018).
                ---------------------------------------------------------------------------
                 At the same time, the agencies conclude that a reasonable case can
                also be made to support values of the rebound effect falling in the 5-
                15 percent range. This argument relies on using the criteria proposed
                by commenters to restrict the studies considered to include recently
                published analyses using U.S. data, and to accord the most weight to
                research that relies on measures of vehicle use derived from odometer
                readings, controls for the potential endogeneity of fuel economy, and
                estimates the response of vehicle use to variation in fuel economy
                itself, rather than to fuel cost per distance driven or fuel prices.
                This approach suggests that the rebound effect is likely in the range
                from 5-15 percent, and is more likely to lie toward the lower end of
                that range. The agencies note that estimates of very low or no rebound
                effect cited by some commenters are either misinterpretations of the
                findings reported by their authors, or do not represent measures of the
                fuel economy rebound effect.\1771\
                ---------------------------------------------------------------------------
                 \1771\ For example, some commenters misinterpret Greene's (2012)
                inability to identify a statistically significant estimate of the
                response of vehicle use to variation in fuel economy as evidence
                that its true value is zero. Similarly, some commenters misinterpret
                the result reported by West et al. (2017) that buyers of more fuel-
                efficient vehicles did not increase their driving as evidence that
                fuel economy itself has no effect on vehicle use, when--as the
                study's authors and some commenters acknowledge--it reveals instead
                that buyers regarded those vehicles as providing inferior
                transportation service and drove them less as a consequence. Because
                the agencies repeatedly insist that vehicle attributes other than
                fuel economy will not change as a consequence of this rule, those
                authors' finding is of limited or no relevance to the analysis
                supporting this rule.
                ---------------------------------------------------------------------------
                 Finally, the agencies note that surveys of evidence on the rebound
                effect have consistently found that the most appropriate estimate falls
                in the range of 10-40 percent. These findings have remained
                surprisingly consistent over time, despite a rapidly expanding universe
                of empirical evidence that includes estimates drawn from more diverse
                settings, and reflects continuing improvements in the data they rely
                upon, an expanding range of strategies for identifying the rebound
                effect and distinguishing it from other influences on vehicle use, and
                advances in the econometric procedures analysts use to estimate its
                magnitude.
                 For the aforementioned reasons, the agencies have elected to retain
                the 20 percent rebound effect used to analyze the effects of the NPRM
                on vehicle use and fuel consumption for analyzing the comparable
                effects of this final rule. As explained above and in the NPRM, older
                research suggests a rebound of 20 to 25 percent. The new research in
                Table VI-175 supports a similar--or even larger--range. Extensive
                survey studies support
                [[Page 24677]]
                a rebound at or above 20 percent. As such, the agencies feel 20 percent
                is a reasonable--and probably even conservative--estimate of the
                totality of the evidence. While a lower estimate may be reasonable
                under certain circumstances, the agencies are uncomfortable making the
                requisite assumptions regarding which specific criteria should be used
                to identify relevant studies and relying on a subset of the literature
                for the central analysis. However, recognizing the uncertainty
                surrounding the rebound value, the agencies also examine the
                sensitivity of those estimated impacts to values of the rebound ranging
                from 10 percent to 30 percent, both in isolation and in conjunction
                with plausible variation in other key parameters.
                (5) Vehicle Miles Traveled (VMT)
                 VMT directly influences many of the various effects of fuel economy
                and CO2 standards that decision-makers consider in
                determining what levels of standards to set. For example, fuel savings
                is a function of a vehicle's efficiency, miles driven, and fuel price.
                Similarly, factors like criteria pollutant emissions and fatalities are
                direct functions of VMT. In the CAFE model, VMT is the product of
                average usage per vehicle in the fleet and fleet composition, which is
                itself a function of new vehicle sales and vehicle retirement
                decisions, otherwise known as scrappage. These three components--
                average vehicle usage, new vehicle sales, and older vehicle scrappage--
                jointly determine total VMT projections for each alternative.
                 As the following discussion explains, today's VMT analysis provides
                aggregate results comparable to other well-regarded VMT estimates.
                However, because the agencies' analysis looks at the incremental costs
                and benefits across alternatives (see Section VII), it is more
                important that the analysis capture the variation of VMT across
                alternatives than accurately to predict total VMT within a scenario. As
                such, the agencies note that today's VMT estimates are logical,
                consistent, and precise across alternatives. Furthermore, as will be
                described in further detail below, while the agencies, in response to
                comments, have decided to modify their approach to calculating VMT and
                to use different VMT estimates than those used in the NPRM, the general
                trends between alternatives are comparable.
                 Commenters addressed a number of topics related to the total amount
                of estimated VMT, the incremental differences in estimated VMT between
                regulatory alternatives, and per-vehicle VMT estimates in the NPRM
                analysis. In general, commenters felt that the NPRM's VMT numbers were
                inaccurate and should not be relied on for the analysis.\1772\ Some
                commenters were more specific and argued that the total amount of
                estimated VMT projected in the NPRM started at too low a level, and
                increased too much over the years simulated. Similarly, some commenters
                argued that the agencies' estimates were too different from other
                recognized estimates and suggested that the agencies benchmark VMT
                projections to other sources to ensure both a consistent starting point
                and comparable VMT throughout the calendar years analyzed.
                ---------------------------------------------------------------------------
                 \1772\ See, e.g., Securing America's Energy Future, NHTSA-2018-
                0067-11981 at 37-38.
                ---------------------------------------------------------------------------
                 A few commenters objected to the underlying mileage accumulation
                schedules, which form the basis for per-vehicle VMT estimates in CAFE
                Model simulations. Such commenters speculated that revisions to these
                schedules undertaken in 2016 might be the reason for discrepancies in
                total VMT. Other commenters were less concerned about how VMT was
                computed within each scenario but were apprehensive about differences
                in VMT estimates across regulatory alternatives. For instance, Honda
                argued that, ``[a]ssuming all other parameters are held constant--and
                excluding the rebound effect--it is not obvious why one scenario should
                have different total VMT than another.'' \1773\ While commenters
                generally provided few specific recommendations about the level to
                which VMT estimates should be constrained among alternatives, several
                commenters argued that VMT projections would benefit from consideration
                of travel demand modeling.
                ---------------------------------------------------------------------------
                 \1773\ Honda, NHTSA-2018-0067-11818, at 17.
                ---------------------------------------------------------------------------
                 Additionally, some commenters (RFF, IPI, NRDC) argued that a
                superior, and perhaps even necessary, approach would be to incorporate
                a model that considers jointly the decision to buy, use, and retire
                vehicles at the household level. As RFF posited ``a household makes
                decisions about its vehicle ownership and use jointly: people don't buy
                new vehicles or get rid of existing ones without considering how these
                actions will affect the use of their vehicles.'' \1774\ IPI further
                argued that ``[i]n sum, VMT is influenced by vehicle choice and vehicle
                choice is influenced by VMT. And a `unified model of vehicle choice and
                usage' is necessary.'' \1775\ While the agencies agree that a joint
                household consumer choice model--if one could be developed adequately
                and reliably to capture the myriad circumstances under which families
                and individuals make decisions relating to vehicle purchase, use and
                disposal--would reflect decisions that are made at the household level,
                the agencies do not agree that it is necessary, or necessarily
                appropriate, to model the national program at that scale in order to
                produce meaningful results that can be used to inform policy decisions.
                The most useful information for policymakers relates to national
                impacts of potential policy choices. No other element of this analysis
                occurs at the household level, and the error associated with allocating
                specific vehicles to specific households over the course of three
                decades would easily dwarf any error associated with the estimation of
                these effects in aggregate. The agencies have attempted to incorporate
                estimates of changes to the new and used vehicle markets at the highest
                practical levels of aggregation, and worked to ensure that these
                effects produce fleetwide VMT estimates that are consistent with the
                best, current projections given our economic assumptions. While future
                work will always continue to explore approaches to improve the realism
                of CAFE/CO2 simulation, there are important differences
                between small-scale econometric studies and the kind of flexibility
                that is required to assess the impacts of a broad range of regulatory
                alternatives over multiple decades. The agencies have read and
                evaluated the comments on the NPRM, incorporating many suggestions that
                improve the fidelity of this analysis--taking particular care to be
                conservative with the analysis. The modifications the agencies have
                made in response to these comments are described below (and in the
                RIA).
                ---------------------------------------------------------------------------
                 \1774\ RFF, NHTSA-2018-0067-11789, at 5.
                 \1775\ IPI, Appendix, NHTSA-2018-0067-12213, at 80 (internal
                citation omitted).
                ---------------------------------------------------------------------------
                 The agencies carefully assessed all comments. To address them, the
                agencies have revised their calculation of estimated VMT in two,
                significant respects. First, in response to comments regarding the
                mileage accumulation schedules, the agencies have revised the schedules
                using panel data. Second, to deal with commenters' concerns with the
                fluctuation of estimated ``non-rebound'' VMT across regulatory
                alternatives, the agencies have created a method that constrains ``non-
                rebound'' VMT across regulatory alternatives. The agencies believe
                these two changes collectively resolve the substantive issues raised by
                commenters. The total VMT for the final rulemaking (FRM) analysis now
                aligns with estimates of the Federal Highway Administration
                [[Page 24678]]
                (FHWA) and the only differences in VMT between alternatives is
                attributable to changes in the fleet's fuel economy. The following
                sections discuss these changes in detail.
                (a) Mileage Accumulation Schedule
                 To account properly for the average value of consumer and societal
                costs and benefits associated with vehicle usage under various CAFE and
                CO2 alternatives, it is necessary to estimate the portion of
                these costs and benefits that will occur each calendar year for each
                model year cohort. Doing so requires some estimate of how many miles
                the average vehicle of each body type is expected to drive at each age.
                The agencies call these ``mileage accumulation schedules.'' For this
                final rule, the agencies are modifying the mileage accumulation
                schedules, largely in response to comments.
                (i) Data
                 As mentioned in previous sections, NHTSA purchased a data set
                containing 70 million vehicle odometer readings from Polk in part to
                create the vehicle mileage accumulation schedules used in the NPRM. In
                the proposal, the agencies explained that Polk data was newer and
                believed to be qualitatively superior to the 2001 and 2009 National
                Household Travel Survey (NHTS) data used in prior rules.\1776\
                Consistent with previous analyses,\1777\ the agencies used a cross-
                sectional sample of the Polk data for the NPRM. Cross-sectional data is
                like a ``snapshot'' in time. Rather than tracking vehicles over a
                period, the sample contained a single odometer reading from each
                vehicle sampled. In other words, the sample contained observations of
                the total lifetime accumulation of miles (represented by its odometer
                reading) through CY2015 of all MYs still present on the road. The
                cross-sectional sample was limited in the number of vintages included
                in the sample. While the sample was suitable to capture the heaviest
                usage ages (age zero to 15 years), it contained no observations for
                vehicles older than 16 years. This required the agencies to rely on
                mileage accumulation schedules developed from other data sources to
                produce annual VMT rates for older vehicles. Furthermore, in order to
                develop a schedule of mileage accumulation by age, it was necessary to
                assume that each vehicle traveled the same number of miles each year to
                reach its odometer reading, e.g. if a MY 2007 vehicle had an odometer
                reading of 88,000 in CY2015, the analysis assumed the vehicle drove
                11,000 miles each year from CY2007 to CY2015.
                ---------------------------------------------------------------------------
                 \1776\ See, e.g., 83 FR at 43089-90 (Aug. 24, 2018).
                 \1777\ Previous rules were based on odometer data from the 2001
                National Household Travel Survey (NHTS). S. Lu, ``Vehicle
                Survivability and Travel Mileage Schedules,'' Report Number: DOT HS
                809 952 (January 2006).
                ---------------------------------------------------------------------------
                 The agencies acknowledged that this approach missed some of the
                nuances of car ownership.\1778\ For example, vehicles are commonly part
                of multi-vehicle household fleets and their usage changes over time as
                households buy new vehicles and replace older ones. Similarly, most
                vehicles belong to multiple owners over the course of their useful
                lives, each of whom may have different patterns of usage. The most
                significant limitation of using cross-sectional data is the presence of
                an attrition bias. As a cohort ages, vehicles that have been used more
                heavily are more likely to be retired at each age than vehicles that
                are driven less. As the most heavily-driven vehicles drop out of the
                fleet, the remaining vehicles, which likely have been driven less at
                each age throughout their lives, will have lower odometer readings.
                Making the common, but necessary assumption that each vehicle is driven
                uniformly at each age results in lower miles-per-age estimates because
                of this attrition bias. In the schedules used for the NPRM, the effect
                of this bias occurred during the ages where each model year cohort
                typically scraps at the highest rates--9 to 15 years. These limitations
                led to lower estimates, which led commenters such as EDF to state
                ``[g]iven that the Volpe Model VMT falls far short of confident
                measurements of gasoline consumption, these mileage accumulation
                schedules need to be increased.'' \1779\ The agencies note that many of
                these data limitations were present in previous CAFE and CO2
                analyses.\1780\
                ---------------------------------------------------------------------------
                 \1778\ See 83 FR at 43092 (Aug. 24, 2018).
                 \1779\ EDF, Appendix B (Rykowski comments), NHTSA-2018-0067-
                12108, at 46.
                 \1780\ See, e.g., NHTSA Final Regulatory Impact Analysis:
                Corporate Average Fuel Economy for MY 2012-MY 2016 Passenger Cars
                and Light Trucks, NHTSA-2010-0131, at 372-79.
                ---------------------------------------------------------------------------
                 Several commenters noted the agencies' reliance on cross-sectional
                data, and urged the use of panel data instead to develop mileage
                accumulation schedules. For example, API argued that cross-sectional
                data cannot accurately capture mileage accrual and suggested ``the
                agencies re-consider the use of the [Polk] data for developing revised
                mileage accumulation schedules unless the data can capture mileage
                accumulation rates versus age on an individual-vehicle basis.'' \1781\
                The NPRM discussed the possible use of panel data in the future and the
                benefits that doing so could provide.\1782\
                ---------------------------------------------------------------------------
                 \1781\ API, EPA-HQ-OAR-2018-0283-4548, at 10.
                 \1782\ See 83 FR at 43092 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 In response to these comments, the agencies created new mileage
                accumulation schedules based on panel data for this final rule. Unlike
                cross-sectional data, panel data includes a temporal element, which
                resolves the limitations imposed by cross-sectional data. The data
                source used for the final rule contains sequential readings of
                individual vehicles over time, and the vehicles are tracked at the VIN
                level. Polk accumulates readings about individual vehicles through
                state inspection programs, title changes, and maintenance events, among
                other sources. The Polk data includes observations of a specific
                vehicle's odometer readings over the course of many years, capturing
                the accumulated lifetime mileage at multiple ages. By using the
                observation date and accumulated miles (represented by the odometer
                reading), the agencies can compute the rate of driving (miles per year,
                or month) between observations for each vehicle. This is a superior
                method to assuming that the rate of accumulation, over all ages, is
                simply the ratio of odometer to age, as commenters noted. In
                particular, calculating the rates of mileage accumulation using
                successive observations of the same vehicle explicitly resolves the
                attrition bias and matches the approach to estimating driving rates
                with panel data in other studies.\1783\
                ---------------------------------------------------------------------------
                 \1783\ See, e.g., Kenneth Gillingham, Alan Jenn, and In[ecirc]s
                M.L. Azevedo (2015), ``Heterogeneity in the Response to Gasoline
                Prices: Evidence from Pennsylvania and Implications for the Rebound
                Effect, Energy Economics,'' Volume 52, Supplement 1, 2015, Pages
                S41-S52, ISSN 0140-9883, available at https://doi.org/10.1016/j.eneco.2015.08.011.
                ---------------------------------------------------------------------------
                 The panel dataset has another advantage over other sources: Because
                it tracks individual vehicles over time, the agencies have more precise
                information about each vehicle's useful age. In previous analyses, the
                agencies were forced to assume that ``age'' was simply equal to the
                calendar year minus the model year in which the vehicle was produced.
                For example, a MY2010 vehicle was assumed to be five years old in 2015.
                This created, as API stated, a ``discontinuity in the values between
                year 1 and year 2'' within the schedules.\1784\ It is common for
                vehicles produced in a given model year to be sold and registered over
                the course of multiple calendar years. Thus, a MY2010 vehicle assumed
                to be five years old in 2015, could have been
                [[Page 24679]]
                registered for the first time in CY2012 and might have a real driving
                age of three years, rather than five, simply because it sat on a
                dealership lot for a couple of years before being purchased. The Polk
                data allows us to identify the first registration date of each vehicle
                in the sample and compute its true driving age at each point in time.
                This not only improves the precision of the mileage accumulation rate
                in the first year, but in subsequent years as well. The odometer data
                used in the NPRM had another limitation: Odometer readings were grouped
                into cohorts by nameplate, for which only distributional information
                was available. It was necessary to use the mean odometer reading for
                each cohort at each age, but in cases where the distribution was
                skewed, the mean could be misleading. Making the same assumption about
                registration date, as each cohort contained information about the
                average registration date, further compounded the potential for
                distortion.
                ---------------------------------------------------------------------------
                 \1784\ API, EPA-HQ-OAR-2018-0283-5458, at 9-10.
                ---------------------------------------------------------------------------
                 To the extent that commenters objected to the NPRM's use of Polk
                data on the basis of it being proprietary, the agencies note that using
                proprietary data is common in rulemakings, and, specifically, Polk data
                has been used for CAFE and CO2 analyses on multiple
                occasions previously. For the 2016 final medium- and heavy-duty rule
                and Draft TAR, the agencies used Polk odometer data to develop the
                vehicle mileage accumulation schedules.\1785\ Further, the specific
                data set was cited and is available for acquisition through Polk.
                ---------------------------------------------------------------------------
                 \1785\ See, e.g., 81 FR 73478, 73746 (Oct. 25, 2016); see also
                81 FR 49217 (Jul. 27, 2016).
                ---------------------------------------------------------------------------
                 Recently, the 2017 National Household Travel Survey has become
                available as a possible data source to develop mileage accumulation
                schedules. While attractive from the standpoint of transparency, it
                suffers from the same flaws as data sources used to develop previous
                schedules. In particular, it represents a cross section of odometer
                readings at a single point in time, requiring the assumption that the
                rate of usage is simply reported odometer divided by vehicle/age, or an
                extrapolation of respondents' daily travel behavior into representative
                annual schedules, which commenters suggested was a poor assumption.
                Additionally, all of the odometers in the newest NHTS are self-
                reported, leading to questionable reliability of the individual data
                points (and notably round numbers in many cases). Finally, the NHTS is
                intended to be a representative sample of households, but not a
                representative sample of vehicles. Research has found that creating a
                representative sample of households can represent a significant
                challenge, as past iterations of the NHTS have systematically
                oversampled high income households. The nature of the sample also
                explicitly excludes vehicles used for commercial purposes, which
                nonetheless compose a meaningful portion of the new vehicle market,
                accumulate miles of travel, and consume fuel. The data set on which the
                mileage accumulation schedule used for this final rule is based
                contains at least two readings (and frequently several) for over 70
                percent of the registered light duty vehicle population in 2016.
                (ii) Methodology
                 The data used to construct the schedules initially included between
                two and fifty odometer readings from each of over 251 million unique
                vehicles. While most of the readings had plausible reading dates,
                odometer counts, and implied usage rates, some of the readings appeared
                unrealistic and received additional scrutiny. The agencies used a set
                of criteria to identify and remove readings that were likely record
                errors. For example, odometer readings predating the commercial release
                of the vehicle, showing negative VMT accumulation over time, or taken
                too closely together to provide meaningful insight into annual vehicle
                usage were removed from the analysis.\1786\ Such sanitization of real
                datasets is typically necessary, and each step in the process was
                recorded and described in conformity with standard econometric
                practice.\1787\
                ---------------------------------------------------------------------------
                 \1786\ Refer to Section VI.D.1.(5).(b) of the FRIA for a full
                accounting of the process used to clean the Polk odometer data.
                 \1787\ See, e.g., Osborne, Jason W., Best Practices in Data
                Cleaning, SAGE Publications, Inc., January 2012.
                ---------------------------------------------------------------------------
                 Similar to the NPRM, the remaining readings were sorted into five
                categories: Cars, SUV's/vans, pickups, MDHD pickups/vans, and chassis.
                The car, SUVs/vans and pickup categories match the definitions used to
                build the VMT schedules used in the NPRM, as well as those used to
                build the scrappage model. Table VI-176 shows the number of VINs,
                reading pairs, and average readings per VIN by body style.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.357
                 *Not used in this final rule analysis, in part in response to
                comments.
                 Once the dataset was cleaned, the agencies created a sample of one
                million reading pairs, where each pair represented an initial odometer/
                date reading and a subsequent odometer/date reading from the same
                vehicle. Analysis
                [[Page 24680]]
                of the entire dataset was too computationally demanding and
                statistically unnecessary. Two conditions were created for sampling.
                The first controlled for Polk's censoring in the odometer readings
                recorded in the dataset (described below), and the second ensured the
                usage data was not biased by survival and that it represented usage
                rates over a relatively short period of time compatible with the
                beginning of the FRM analysis. Further analysis suggests that shorter
                periods between readings is still correlated with higher usage rates so
                that further filtering of the data sample was considered in the
                regression analysis. Once these filters were applied, the agencies
                considered several polynomial fits to the average odometer readings.
                These fits inform the final usage rates by age and body style used in
                this FRM analysis. The details are further described below.
                 One element of the usage data (mentioned above as the first
                condition control) required the agencies to filter the dataset. The
                odometer readings recorded are censored at 250k miles.\1788\ For this
                reason, the agencies exclude readings recorded exactly as 250k miles.
                The censoring could bias estimates of usage rates if odometer readings
                and future usage rates are correlated, which they likely are. While the
                agencies hope to reconcile this limitation of the dataset in future
                work, the benefits of observing actual usage data through 30 years
                (rather than average odometer readings by model through 15 years) far
                outweigh the limitation. Still, the agencies filtered out these
                censored data points, since the actual odometer readings for such
                vehicles are likely higher than reported.
                ---------------------------------------------------------------------------
                 \1788\ Polk codes any vehicle whose odometer exceeds 250K miles
                as 250K miles exactly, regardless of the actual odometer reading.
                ---------------------------------------------------------------------------
                 The Polk dataset is conditional on survival so that it represents
                the usage of vehicles on the road at the time of the sample (the end of
                the first quarter of 2017). In this way, it captures the actual
                observed usage rates of vehicles surviving to their current age in the
                dataset. An issue with this is that all readings of a vehicle are
                included in the sample. If usage rates from earlier ages and survival
                are correlated, which they likely are, then including the readings for
                a 30-year-old vehicle when it was 10 years old will bias the estimated
                usage rates of 10-year-old vehicles downward because vehicles that
                survive to advanced ages tend to be used less than vehicles that are
                retired at earlier ages for the same model year. As noted above, the
                NHTS data used in the NPRM suffered from the same problem. To mitigate
                this issue, the agencies applied a second filter when sampling the data
                set: The agencies only included readings where the reading date of the
                second reading in the pair is January 2015 or later. This reduces the
                potential bias from the joint probability of usage and survival to only
                those vehicles scrapped between January 2015 and the first quarter of
                2017. This balances losing information for older, less represented ages
                by excluding too much data on these vehicles and severely biasing the
                estimates of usage by age.
                 For estimates within the CAFE model the average usage is the
                relevant measure. Table VI-177 shows the average usage rates for cars
                by age as well as linear, quadratic, and cubic polynomial fits on these
                points.\1789\ The average usage rates follow a relatively smooth
                pattern, but appear to decline at an accelerating rate for the oldest
                ages. The linear equation captures this trend for older vehicles, but
                underestimates early ages. The quadratic fit shows a diminishing
                decrease in the usage of older vehicles which may overestimate their
                use. The cubic fit captures the early age usage trends and the
                accelerating decrease in the usage of older ages. For this reason, the
                agencies used the cubic curve as the basis for the new VMT schedules by
                age.
                ---------------------------------------------------------------------------
                 \1789\ In general, the objective of a polynomial regression is
                to capture the nonlinear relationship between two variables. While
                the fit produces a nonlinear curve, it is linear in the
                coefficients. Choosing the lowest degree of the polynomial function
                that captures the inflection points in the data preserved degrees of
                freedom and ensures that applying the polynomial function to
                observations outside the range of data (as done here for ages beyond
                30) is well behaved.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.358
                [[Page 24681]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.359
                 Table VI-178 shows the observed and predicted average usage rates
                by age for SUVs/vans. All the polynomial fits predict the observed
                average usage rates reasonably well. However, the linear fit under
                predicts the usage of the oldest vehicles, and the cubic fit predicts
                higher usage rates for vehicle ages beyond age 30. The quadratic fit
                predicts reasonable usage rates for all observed and out-of-sample ages
                through age 40. For this reason, the quadratic fit was used as the
                basis for the SUV mileage schedule.
                [[Page 24682]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.360
                [[Page 24683]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.361
                 Table VI-179 shows the observed and predicted average usage rates
                for pickups by age. The observed rates initially decline at an
                increasing rate, the decline diminishes and appears to accelerate again
                for the oldest ages. The linear fit underestimates the usage rates for
                the youngest and oldest ages and overestimates middle-aged vehicles.
                The quadratic fit reasonably predicts the observed average usage rates
                but predicts an increase in usage rates for the oldest ages out of the
                observed sample. The cubic fit reasonably predicts the observed
                averages and appears to capture the diminishing decline of usage for
                the oldest ages observed in the in-sample averages. For this reason,
                the agencies used the cubic fit as the basis for the pickup VMT
                schedules.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.362
                [[Page 24684]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.363
                 As in the NPRM, the current schedule differs by body-style to
                represent different usage profiles that the agencies observed in the
                data. While more stratification is possible, it is unlikely to provide
                much additional value. Table VI-180 shows the annual miles driven at
                each age for passenger cars, SUVs (and CUVs and minivans), and pickup
                trucks at each age of their useful life, conditional upon surviving to
                that age.
                [[Page 24685]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.364
                (b) Benchmarking Total VMT
                 In order to assess the fuel consumption and environmental impacts
                of regulatory alternatives, it is desirable to have a representation of
                aggregate travel and fuel consumption that is both reasonable and
                internally consistent. Some commenters suggested that the aggregate
                totals presented in the NPRM deviated from other published estimates,
                and argued that the entire analysis was therefore an unreliable source
                of information for decision-makers to consider. For example, EDF
                stated, ``the NHTSA model `projects' aggregate, nationwide VMT levels
                for 2016 and 2017 that are about 20 percent lower than formal
                government estimates by EIA and FHWA.'' \1790\ EDF further stated,
                ``[b]etween 2017 and 2025, fleetwide VMT grows by 3.1% per year in the
                Volpe Model, while it only grows 0.5% per year in the 2018 Annual
                Energy Outlook.'' \1791\ EDF also suggested, ``[o]ne obvious way to
                assess the accuracy of the schedules is to compare the projections of
                the Volpe Model of total fleetwide fuel consumption in a recent
                calendar year with actual gasoline sales.'' \1792\
                ---------------------------------------------------------------------------
                 \1790\ EDF, Appendix A, NHTSA-2018-0067-12108, at 59.
                 \1791\ EDF, Appendix B (Rykowski comments), NHTSA-2018-0067-
                12108, at 44.
                 \1792\ Id. at 43.
                ---------------------------------------------------------------------------
                 The Federal Highway Administration (FHWA) publishes annual VMT
                estimates for the light-duty vehicle fleet, the most recent of which is
                calendar year 2017. The NPRM estimate of total light-duty VMT was 2.22
                trillion miles in calendar year 2016. The FHWA estimate for light duty
                VMT in 2016 was 2.85 trillion miles.\1793\ While the definitions of
                light-duty are not identical in the two cases (where FHWA excludes
                trucks with 10,000 lbs. GVW, the agencies' analysis excludes trucks
                with GVW greater than 8,500 lbs. from its light duty definition), that
                definitional discrepancy is not significant enough to account for the
                difference in the total VMT. While some commenters suggested that the
                agencies compare simulated fuel consumption to published estimates from
                EIA to determine the validity of our VMT assumptions, such a comparison
                requires accurate assumptions about the true on-road fuel efficiency of
                registered vehicles over forty model years in addition to their annual
                usage. Comparing simulated VMT directly to FHWA measurements requires
                fewer assumptions and is a more meaningful comparison.
                ---------------------------------------------------------------------------
                 \1793\ See Highway Statistics 2017, Table VM-1, available at
                https://www.fhwa.dot.gov/policyinformation/statistics/2017/vm1.cfm.
                ---------------------------------------------------------------------------
                 Substituting the updated mileage accumulation schedules for the
                NPRM schedules, and using the calendar year 2016 fleet from the NPRM,
                produces an estimate of total light duty VMT in 2016 that is about 2.85
                trillion miles--nearly identical to the FHWA estimate for 2016, despite
                the use of different estimation methods and data sources. FHWA's
                estimate of total light-duty VMT in 2017 is 2.88 trillion miles,\1794\
                while the estimate produced by the simple product of the mileage
                accumulation schedule on the estimated on-road fleet is 2.94 trillion
                miles, a difference of about two percent. While not as close as the
                estimate for calendar year 2016, the discrepancy is still small
                considering that the estimates are obtained through entirely different
                methods. One important source of
                [[Page 24686]]
                discrepancy with FHWA's 2017 VMT estimate is the fact that the CAFE
                model simulation assumes all of the vehicles produced in a given model
                year are driven for the entire calendar year matching the
                vintage.\1795\ This means, for calendar year 2017, the initial year of
                the simulation used to support this rule, MY2017 vehicles are assumed
                to have been both registered and driven for the entirety of CY2017. As
                a result, it naturally overestimates the true VMT for calendar year
                2017. The analysis accounts for this discrepancy by adjusting calendar
                2017 total VMT downward by one percent, and the discrepancy in total
                VMT caused by conflating model years and calendar years dissipates over
                time.
                ---------------------------------------------------------------------------
                 \1794\ Id.
                 \1795\ The CAFE model uses an annual timestep, meaning that each
                time period represents one year. Because calendar years are
                (obviously) years, and all of the other inputs (discounting and
                inflation, macroeconomic variables, fuel prices, VMT, etc.)
                represent annual values, the timestep in the CAFE model is a
                calendar year. However, model years start prior to the calendar year
                for which they are named, and new model year sales continue (albeit
                only slightly) after their calendar year ends. In order to account
                for model year sales on their true timing relative to calendar
                years, the model would need to be restructured to use a quarterly
                timestep. While this would improve the fidelity between calendar
                year and model year for sales, obtaining quarterly projections of
                nearly every other variable in the analysis would be complicated (if
                not impossible). For this reason, the model conflates ``model year''
                and ``calendar year'' for the analysis, even though it is a
                simplification.
                ---------------------------------------------------------------------------
                 While the agencies have established that the years for which they
                have data are sufficiently similar to published VMT estimates, the
                question of projection still remains. FHWA, in its forecasts of VMT
                (Spring 2019),\1796\ forecasts a compound annual growth rate of 0.8
                percent for light-duty vehicles between 2017 and 2047 in its baseline
                economic outlook. However, that projection uses a different set of
                macroeconomic conditions and fleet assumptions than this analysis. To
                compare CAFE model simulations of total VMT to the FHWA projections,
                the agencies ran the FHWA model with a comparable set of assumptions to
                the greatest extent possible.\1797\ \1798\ Using similar economic
                growth assumptions, our reference case total light-duty VMT grows at a
                compound rate of 0.63 percent per year between 2017 and 2050. Using
                comparable assumptions in the FHWA model produce an annual growth rate
                of 0.66 percent. Again, these differences are remarkably low for models
                created with different methods, and lead to trivial variances, for the
                purposes of our analysis, in total VMT. The relevant annual projections
                for the baseline scenario appear in Table VI-181.
                BILLING CODE 4910-59-P
                ---------------------------------------------------------------------------
                 \1796\ See ``FHWA Forecasts of Vehicle Miles Traveled (VMT):
                Spring 2019,'' Office of Highway Policy Information, available at
                https://www.fhwa.dot.gov/policyinformation/tables/vmt/vmt_forecast_sum.pdf.
                 \1797\ See ``FHWA Travel Analysis Framework: Development of VMT
                Forecasting Models for Use by the Federal Highway Administration,''
                Volpe, available at https://www.fhwa.dot.gov/policyinformation/tables/vmt/vmt_model_dev.pdf.
                 \1798\ In particular, we ran the FHWA VMT forecasting model with
                the same: Personal disposable income, population, fuel prices (all
                of which come from AEO2019), and simulated on-road fleet fuel
                economy in the baseline.
                ---------------------------------------------------------------------------
                [[Page 24687]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.365
                [[Page 24688]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.366
                BILLING CODE 4910-59-C
                (c) Preserving Total VMT Across Regulatory Alternatives
                 In the NPRM, the combined effect of the sales and scrappage
                responses created small percentage differences in total VMT across the
                range of regulatory alternatives.\1799\ However, as the Environmental
                Group Coalition noted, even a 0.4 percent difference can result in
                ``692 billion additional VMT under the CAFE standards and 894 billion
                under the CO2 program.'' \1800\ Since VMT is related to many
                of the costs and benefits of the program, VMT of this magnitude can
                have meaningful impacts on the incremental net benefit analysis. This
                point was made by a number of commenters who were concerned about the
                magnitude and direction of differences in VMT between regulatory
                alternatives (IPI, EDF, CBD, CARB, EPA's Science Advisory Board).\1801\
                ---------------------------------------------------------------------------
                 \1799\ The agencies explained in the NPRM that some amount of
                this difference was due to the rebound effect, and that ``non-
                rebound'' VMT between alternatives differed by as much as 0.4
                percent. See 83 FR at 43099 (Aug. 24, 2018).
                 \1800\ Environmental Group Coalition, Appendix A, NHTSA-2018-
                0067-12000, at 180.
                 \1801\ See, e.g., id.; EDF, Appendix B (Rykowski comments),
                NHTSA-2018-0067-12108, at 42-46; IPI, Appendix, NHTSA-2018-0067-
                12213; at 79; CARB, Detailed Comments, NHTSA-2018-0067-11873, at
                237-242.
                ---------------------------------------------------------------------------
                 More generally, commenters argued that non-rebound VMT should be
                held constant across regulatory alternatives, regardless of the number
                of new vehicles sold and registered vehicles scrapped. For example, CBD
                commented that the ``total number of VMT should be determined based on
                demand for travel, not arbitrarily driven by fleet size.'' CARB added
                that fleet size can change across the alternatives ``as long as the VMT
                schedules are adjusted to account for overall travel activity that is
                distributed over a larger number of vehicles.'' \1802\ NCAT, Global,
                Auto Alliance, EDF, IPI, and Honda made similar arguments.\1803\
                ---------------------------------------------------------------------------
                 \1802\ CARB, Detailed Comments, NHTSA-2018-0067-11873, at 238
                (internal citation omitted).
                 \1803\ See, e.g., Global, Attachment A, NHTSA-2018-0067-12032,
                at A-26-A-30; NCAT, Comments, NHTSA-2018-0067-11969, at 28-32; EDF,
                Appendix A, NHTSA-2018-0067-12108, at 30; IPI, Appendix, NHTSA-2018-
                0067-12213, at 80-85; Honda, NHTSA-2018-0067-12111.
                ---------------------------------------------------------------------------
                 While commenters generally provided few specific recommendations
                about the level to which VMT should be constrained among alternatives,
                several of them argued that VMT projections would benefit from
                consideration of travel demand modeling. UCS, CBD, NCAT, and others
                suggested that the overall level of light-duty VMT in a given year
                should reflect the broader economic context in which travel
                occurs.\1804\ For example, Honda stated, ``[i]ncreasing VMT is closely
                associated with increased economic activity.'' \1805\
                ---------------------------------------------------------------------------
                 \1804\ See, e.g., NCAT, Comments, NHTSA-2018-0067-11969, at 31-
                32; Environmental Group Coalition, Appendix A, NHTSA-2018-0067-
                12000, at 175-76; and, UCS, Technical Appendix, NHTSA-2018-0067-
                12039, at 60-61.
                 \1805\ Honda, Supplemental Analysis, NHTSA-2018-0067-1211, at 4.
                ---------------------------------------------------------------------------
                 The agencies agree that the total demand for VMT should not vary
                excessively across alternatives and stated as much in the NPRM.\1806\
                That said, it is reasonable to assume that fleets with differing age
                distributions and inherent cost of operation will have slightly
                different annual VMT, absent VMT associated with rebound miles;
                however, the difference could conceivably be small. To address these
                comments and to take an intentionally conservative approach, the
                agencies decided to constrain ``non-rebound'' VMT (defined more
                explicitly below) to be identical across regulatory alternatives in
                this analysis using the FHWA VMT demand model to determine the
                constraint; therefore, the only difference in total VMT between
                regulatory alternatives is the rebound miles attributable to
                differences in fuel economy resulting from the regulatory alternatives.
                Nevertheless, as explained in the NPRM and revealed in the extensive
                quantitative results published with the NPRM, setting aside the rebound
                effect, aggregate VMT as estimated in the NPRM was roughly constant
                across alternatives. Although differences may have appeared large in
                absolute terms, especially when aggregated across many calendar years
                and ignoring the underlying annual total quantities, the differences
                were nevertheless very small in relative terms--small enough to be well
                within the range of measurement or estimation error for virtually any
                of the other inputs to, or outputs of, the agencies' analysis. It is
                unclear whether a 0.4 percent change in highway travel can be measured
                with any degree of confidence.
                ---------------------------------------------------------------------------
                 \1806\ See 83 FR at 43099 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 To constrain non-rebound VMT, the agencies needed to create a
                definition of non-rebound VMT and a method for calculating it. The
                agencies used the FHWA VMT forecasting model to produce a forecast of
                non-rebound VMT, to which total non-rebound VMT in every regulatory
                alternative is constrained in each year, regardless of the fleet size
                or distribution of ages in the fleet. In calendar years where total
                non-rebound VMT determined by the size of the fleet and assumed usage
                of each vehicle is lower than the constraint produced from the FHWA
                model, VMT is added to that total and allocated across vehicles to
                match the non-rebound forecast (preserving the constraint). These
                additional miles are then carried throughout the analysis as vehicles
                accrue costs and benefits. Because non-rebound VMT is being held
                constant for the FRM analysis across the set of regulatory alternatives
                in each calendar year, the only difference in VMT among the
                [[Page 24689]]
                alternatives in any calendar year results from differences in fuel
                economy improvement relative to MY2016 that occur as a result of the
                standards. Finally, in Section VII, the agencies calculate the changes
                in total VMT attributable to fuel economy, otherwise known as the
                rebound VMT.
                (i) Defining Non-Rebound VMT
                 In order to constrain non-rebound VMT, it is first necessary to
                define ``non-rebound VMT'' more precisely. The NPRM defined the rebound
                effect as the overall elasticity of travel with respect to changes in
                the cost per mile (CPM). CPM has two components. The first component of
                CPM is fuel prices--the agencies expect vehicles to be driven less if
                fuel prices go up, all else equal. The second component of CPM is fuel
                economy. Therefore, the NPRM defined the percentage change in CPM, for
                a given scenario, model year, and calendar year, as: \1807\
                ---------------------------------------------------------------------------
                 \1807\ See 83 FR at 43091 (Aug. 24, 2018).
                Equation VI-7--Full change in cost per mile of travel
                [GRAPHIC] [TIFF OMITTED] TR30AP20.367
                Where FP is fuel price, FE is fuel economy, and REF refers to the
                reference FE value of a given age (in particular, FE
                2016-(CY-MY), which is the FE of the MY cohort that was
                age CY-MY in CY 2016). In the equation above, FESN,MY,CY
                refers to the observed fuel economy of the MY cohort (typically
                applied at the vehicle level) for a given scenario (SN) in calendar
                year CY.
                 The CAFE model uses one value, the value specified as the rebound
                effect, to measure CPM elasticity. Naturally, the CAFE model produces
                the same magnitude of change in travel for equivalent changes in fuel
                prices and fuel economy. Constructing such a projection of future VMT
                (from 2017 to 2050) that sets aside the rebound effect required
                constructing inputs that were consistent with that perspective. In
                particular, it was necessary to separate the price response associated
                with the change in fuel prices relative to the year on which the
                agencies based the mileage accumulation schedule (end of CY2016), and
                the change in VMT associated with only the improvements in fuel
                economy, relative to MY2016, that occur for future model years at the
                forecasted fuel price.
                 As vehicles age, the agencies expect their VMT to decrease in the
                presence of a non-zero rebound effect if rising fuel prices over time
                increase the per-mile cost of travel, and the rebound effect represents
                the degree to which their travel is reduced for a percentage change
                increase in operating cost. It is intuitive that, as the cost of fuel
                rises over time, a vehicle with a fixed fuel economy would be driven
                less if gasoline costs $3.50/gallon than it would be if gasoline costs
                $2.50/gallon. Such a response is also consistent with economic
                principles (and literature),\1808\ and so it is included in the ``non-
                rebound'' VMT that the agencies constrain across alternatives in each
                calendar year.
                ---------------------------------------------------------------------------
                 \1808\ See, e.g., Goodwin, P., J. Dargay, and M. Hanly.
                Elasticities of road traffic and fuel consumption with respect to
                price and income: A review. Transport Reviews, 24:275-292, 2004.
                ---------------------------------------------------------------------------
                 Similarly, the annual mileage accumulation of cohorts in the
                inherited fleet is clearly affected by fuel price, but also by
                evolution. Setting aside any fuel economy improvements in vehicles sold
                and entering the on-road fleet between 2017 and 2050, the average fuel
                economy of each age cohort is going to improve over that period. The
                travel behavior of the on-road fleet was last observed through calendar
                year 2016 in the Polk data (discussed in (a)(ii)), when a 20-year-old
                car was part of the model year 1997 cohort, and had an average fuel
                economy of 23.4 MPG. However, the fleet continually turns over. In
                2035, the 20-year-old car will be a member of the model year 2016
                cohort, and have an average fuel economy of 29.2 MPG (assumed to be the
                average fuel economy of MY2016 vehicles when they were new).\1809\ If
                fuel prices persist at 2016 levels (in real dollars), then that 25
                percent improvement in fuel economy would reduce the cost per mile of
                travel for 20-year-old vehicles relative to the observed values in
                calendar year 2016, and lead to an increase in travel demand for
                vehicles of that age. Importantly, this transition to more efficient
                age cohorts occurs in all of the regulatory alternatives. Considering
                only the fuel economy levels of vehicles that exist prior to the first
                year of simulation (2017), a secular improvement in the fuel economy of
                the on-road fleet would occur with no further improvements in fuel
                economy from new vehicles in model years 2017 to 2050. As the fleet
                turns over, its fuel efficiency will gradually resemble that of the
                model year 2016 cohort, up to the point at which each age cohort is as
                efficient as the model year 2016 cohort.\1810\
                ---------------------------------------------------------------------------
                 \1809\ In practice, vehicles will scrap at different rates over
                time, even within a body-style. Some nameplates and manufacturers
                have reputations for longevity and individual vehicle models with
                different fuel economies may seem like better candidates for repairs
                under particular fuel price scenarios. In light of this, the fuel
                economy for a given body-style will likely not continue to be the
                sales-weighted average fuel economy when the cohort was new, even
                without accounting for degradation and changes to the on-road gap
                over time. The agencies make this assumption here out of necessity.
                 \1810\ Vehicles scrap at different rates over time, and there
                are important differences by body style for both scrappage rates and
                mileage accumulation. This discussion is intended to provide
                intuition, without all of the computational nuance that exists in
                the model's implementation.
                ---------------------------------------------------------------------------
                 The notion of ``non-rebound'' VMT is a construct necessary to
                support this regulatory analysis by controlling for VMT attributable to
                reasons other than rebound driving, but present only in theory. Using
                our symmetrical definition of rebound to represent the expected
                response to changes in CPM, regardless of whether those changes occur
                as a result of changes in fuel price or fuel economy, it is well
                established that demand for VMT responds to the cost of travel. To
                isolate the change in VMT for which the regulatory alternatives are
                responsible, the agencies have also included the VMT attributable to
                secular fleet turnover (through MY2016) in the total ``non-rebound''
                VMT projection. In particular, this means that the conventional rebound
                definition used in previous analyses, is replaced in the ``non-
                rebound'' VMT estimation with a more limited definition:
                Equation VI-8--Fuel price and secular improvement component of
                elasticity
                [[Page 24690]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.368
                Where FP is fuel price, FE is fuel economy, and REF refers to the
                reference FE value of a given age (in particular, FEREF =
                FE 2016-(CY-MY), which is the average FE of the MY cohort
                that was age (CY-MY) in CY 2016). In Equation VI-8,
                FEMIN(2016,MY)
                refers to the observed fuel economy of the model year being evaluated
                up to and including the 2016MY cohort. This construction explicitly
                accounts for the improvement in fuel economy between MY2016 and all the
                historical ages (through MY1977) with respect to the change in (real)
                fuel price relative to calendar year 2016. Thus, the VMT associated
                with the rebound effect in this analysis only accounts for changes to
                CPM that result from the amount of fuel economy improvement that occurs
                relative to MY2016. The full elasticity definition (in Equation VI-7)
                differs from that in Equation VI-8 in only one way; the fuel economy in
                the denominator of the first term is the fuel economy of the model year
                being evaluated, rather than being the minimum of the actual model year
                and model year 2016.
                 Combining this demand elasticity with the endogenously estimated
                vehicle population and the mileage accumulation schedule provides an
                initial estimate of non-rebound VMT, as in Equation VI-9.
                Equation VI-9--Unadjusted total non-rebound VMT in a calendar year
                [GRAPHIC] [TIFF OMITTED] TR30AP20.369
                 In Equation VI-9, VMT represents the non-rebound mileage
                accumulation schedule (by age, A, and body style, S), Population is the
                on-road vehicle population simulated by the CAFE Model (in calendar
                year CY, for each age, A, and body style, S), [egr] is the elasticity
                of demand for travel (the rebound effect, assumed to be -0.2 in this
                analysis).
                 However, there are factors beyond the CPM that affect light-duty
                demand for VMT. The FHWA VMT forecasting model includes additional
                parameters that can mitigate or increase the magnitude of the effect of
                fuel price changes on demand for VMT. In particular, the model accounts
                for changes to per-capita personal disposable income (and U.S.
                population) over time. This means that even if fuel prices are
                increasing over the study period (as they are in the central case), and
                fleetwide fuel economy improves only through fleet turnover (as it does
                in the simulated ``non-rebound'' case), total demand for VMT can still
                grow as a result of increases in these other relevant factors. Not only
                does the forecast of non-rebound VMT continue to grow in the non-
                rebound case, it does so at a faster rate than Equation VI-9 produces.
                Thus, in order to preserve non-rebound VMT in a way that represents
                expected VMT demand, the agencies must constrain non-rebound VMT in
                each alternative to match the forecast produced by the FHWA model using
                the fuel price series from the central analysis, AEO2019 Reference case
                assumptions for per-capita personal disposable income, and fleetwide
                fuel economy values produced by simulating the effect of fleet turnover
                (only) in the CAFE model.\1811\
                ---------------------------------------------------------------------------
                 \1811\ Non_rebound_VMT_forecasting.xls in Docket No. NHTSA-2018-
                0067.
                ---------------------------------------------------------------------------
                Constraining Non-Rebound VMT
                 For this final rule, total `non-rebound' VMT is calculated for each
                calendar year and reported in Section VI.D.1.b)(5)(d). In any future
                calendar year, ``non-rebound'' VMT is calculated as a product of the
                initial CY2017 total and a series of compound growth rates:
                Equation VI-10--Total non-rebound VMT constraint in each calendar year
                [GRAPHIC] [TIFF OMITTED] TR30AP20.370
                Where CY is calendar year, r is the compound annual growth rate
                (unique to each CY), and TotalVMT is the calendar year total light-
                duty VMT estimated by the CAFE Model using the annual VMT for each
                body style and age in the mileage accumulation schedule (defined in
                Table VI-180), the population of each age/style cohort in CY2017,
                and the initial difference between operating costs in 2016 and 2017.
                The compound annual growth rates, rCY, in Equation VI-10
                are derived from the inter-annual differences in the forecast of
                total non-rebound VMT that the agencies created using the FHWA
                model.
                 The agencies used the FHWA forecasting model to produce two
                distinct VMT forecasts (both of which appear in Table VI-182). The
                first of these is identical to the forecast of total VMT reported in
                Table VI-181, and represents the AEO2019 Reference case assumptions
                with the exception of average on-road fuel economy, which was simulated
                using the CAFE model to simulate new vehicle fuel economy, new vehicle
                sales, and vehicle retirement under the baseline standards. The
                forecast in the second column of Table VI-182 is identical to the
                first, except that the average on-road fuel economy accounts for only
                the effect of fleet turnover on fuel economy
                [[Page 24691]]
                improvements (new vehicles are assumed to be only as fuel efficient as
                the MY2016 cohort, discussed above).
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.371
                [[Page 24692]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.372
                BILLING CODE 4910-59-C
                 The third column is the non-rebound VMT constraint produced by the
                CAFE model, to which non-rebound VMT is constrained to in every
                regulatory alternative (under central analysis assumptions regarding
                fuel prices and economic growth). The non-rebound VMT constraint is
                produced endogenously by the model in each run based on the estimated
                VMT for calendar year 2017 and a series of growth rates intended to
                reproduce the general growth trend in light-duty VMT under the set of
                ``non-rebound'' assumptions in the FHWA model (Equation VI-10).\1812\
                It differs from the ``non-rebound'' forecast produced by the FHWA model
                by one to three percent in any year. This adjustment was both an
                attempt to match the FHWA model's projection of total VMT (including
                rebound) in the baseline, and an acknowledgment that differing levels
                of modeling resolution and construction are likely to produce slightly
                different projections. In general, the one to three percent difference
                in non-rebound VMT is within the range of projections based on the
                confidence intervals of the coefficients that define the FHWA
                forecasting model.
                ---------------------------------------------------------------------------
                 \1812\ This ensures internal consistency with the set of
                assumptions provided by the user, but can lead to differences
                between the non-rebound VMT constraint in the central analysis and
                one that is generated under a different set of assumptions (as in
                the sensitivity analysis, for example).
                ---------------------------------------------------------------------------
                [[Page 24693]]
                 The fourth column in Table VI-182 represents the unadjusted ``non-
                rebound'' VMT produced by the CAFE Model using Equation VI-9. The
                reader will observe that in every calendar year, this total is lower
                than the non-rebound VMT constraint. This occurs because the projected
                fuel prices in the central analysis increase much faster than the
                fleetwide fuel economy (in the non-rebound case). This increases CPM
                and, as a consequence, reduces demand for VMT based on the price
                elasticity of demand for travel (rebound effect). However, the FHWA
                model accounts for additional variables that recognize the economic
                context in which this fuel price projection occurs. In particular, the
                model accounts for changes in the U.S. (human) population and changes
                to personal disposable income over the same period. These factors act
                to attenuate the demand response to rising fuel prices, producing a
                rising demand for VMT even as the CPM rises for several years.
                 In order to constrain non-rebound VMT to be identical in each year
                across regulatory alternatives, it is necessary to add VMT to the
                unadjusted total, endogenously calculated by the CAFE Model in each
                calendar year. These additional miles, denoted [Delta]miles for this
                discussion, represent the simple difference between the annual VMT
                constraint (column 3 of Table VI-182) and the unadjusted VMT defined in
                Equation VI-9 (above) in each calendar year.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.373
                 Because each regulatory scenario produces a unique on-road fleet
                (in terms of the number of vehicles, the distribution of ages among
                them, and the resulting distribution of fuel economies), the total
                unadjusted VMT in each calendar year (given by Equation VI-9) will be
                unique to each regulatory scenario. As a corollary,
                [Delta]milescy will also be unique to each regulatory
                scenario. By distributing [Delta]milescy across the vehicle
                fleet in each calendar year, the CAFE Model scales up the unadjusted
                non-rebound VMT to equal the non-rebound VMT constraint in each
                calendar year, for each regulatory alternative. While there are a
                number of ways to reallocate [Delta]milescy across the on-
                road fleet in order to match the non-rebound VMT constraint, the fact
                that unadjusted VMT is always lower suggests an obvious approach.
                 The primary goal of reallocation is to adjust total non-rebound VMT
                so that it is identically equal to the VMT constraint in every calendar
                year for each regulatory alternative, while conserving the general
                trends of the mileage accumulation schedule--which represents a good
                estimate of observed usage at the start of the simulation. In
                particular, the reallocation approach should preserve the basic ideas
                that annual mileage decreases with vehicle age because newer (and more
                efficient) vehicles are more likely to be driven additional miles than
                their older counterparts, and mileage accumulation varies by body
                style. To accomplish the reallocation, the CAFE Model computes a ratio
                that varies by body style, calendar year, and regulatory alternative.
                The ratio captures the share of additional VMT that can be absorbed by
                the registered vehicle population of each body style based on their
                relative representation in the fleet, so that per-vehicle totals across
                ages remain sensible (even if the distribution of body styles should
                change over time as the new vehicle market evolves). Then this quantity
                is further scaled by the total VMT for a given body style in the
                calendar year for which [Delta]miles has been computed. The resulting
                ratio is then used to scale the unadjusted miles from Equation VI-9, so
                that the new sum of annual (non-rebound) VMT across all of the vehicles
                in the on-road fleet equals the constraint. For a single calendar year,
                CY, and a single body style, S, the scaling ratio, R, is computed as:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.374
                 In Equation VI-12, Population, refers to the on-road vehicle
                population for a given age and body style (summed over the full range
                of ages in the simulation, where vehicles are modeled to survive for,
                at most, forty years). The fraction in the numerator calculates the
                fleet composition by body type.\1813\ As long as the unadjusted non-
                rebound VMT produced by the CAFE Model is smaller than the VMT
                constraint for all years and regulatory alternatives (and it is), this
                scaling ratio allows the CAFE Model to add miles to the annual total in
                a way that preserves the basic ideas of the mileage accumulation
                schedule and achieves equality with the constraint. In particular, the
                total adjusted non-rebound VMT is then calculated as:
                ---------------------------------------------------------------------------
                 \1813\ We also considered basing this ratio on each body style's
                share of total VMT in that calendar year. However, that approach has
                the potential to result in allocations that add (or remove) too many
                miles per vehicle, depending on the age distribution and size of
                each body style cohort. While that approach better preserves the age
                distribution of VMT within a style, capturing the differences in age
                distribution of the population in each scenario is an objective of
                the VMT accounting. In testing, the differences in approach were
                small (about 0.1 percent difference).
                ---------------------------------------------------------------------------
                [[Page 24694]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.375
                 To make each alternative match the VMT constraint, Equation VI-13
                allocates miles (in this case, adds) to each vehicle in a calendar year
                by multiplying the product of the mileage accumulation schedule (for
                that style vehicle, at that age), the %[Delta]NrbdCPM (described in
                Equation VI-8), and the elasticity (the rebound effect of -0.2) with
                the appropriate scaling ratio (defined in Equation VI-12). The
                ``Allocated Miles'' in Table VI-176 are the result of this calculation
                for a passenger car in CY2020.
                 Unlike some of the accounting, which focuses on the impacts to a
                model year cohort of vehicles over the course of its useful life, the
                rebound constraint and reallocation are calendar year concepts. The
                constraint represents demand for VMT absent ``rebound miles'' (defined
                more explicitly above) in a specific calendar year. Thus, this
                reallocation occurs in every calendar year, and a vehicle of a model
                year cohort will likely experience many of these reallocation events
                during its simulated useful life. The resulting survival weighted
                mileage accumulation is discussed in detail in the discussion of VMT
                Resulting From Simulation found in Section (d), but an example of the
                annual reallocation is provided here.
                 In the baseline alternative, the non-rebound VMT constraint in
                CY2020 is about 3.068T miles, but the endogenously computed ``non-
                rebound'' VMT is only 2.955T miles. This creates a difference,
                [Delta]miles2020, of 112.6B miles that must be added to the
                total unadjusted non-rebound VMT in calendar year 2020 and allocated
                across the on-road fleet in that year to preserve total non-rebound
                VMT. Over time, this discrepancy between the FHWA model's projection
                and the unadjusted total non-rebound VMT grows to about 230 billion
                miles. While the other classes operate identically, this example uses
                the reallocation that occurs to passenger cars to illustrate the
                mechanics of reallocation. Rising fuel prices depressing non-rebound
                VMT (relative to the mileage schedule) over time is a general trend
                that emerges for all body styles, as shown for passenger cars in Table
                VI-183.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.376
                [[Page 24695]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.377
                BILLING CODE 4910-59-C
                 The number of miles added to each age vehicle is generally less
                than the difference between the unadjusted non-rebound VMT (for a given
                age) and the mileage schedule. Thus, adding the requisite miles to each
                age does not distort either the shape of the schedule with age, nor
                does it create annual usage estimates that are out of line with
                observed usage. The example shown here uses the baseline alternative to
                illustrate the reallocation of VMT in 2020, but this reallocation
                differs by alternative. In less stringent regulatory alternatives, new
                vehicles are less expensive; this increases new vehicle sales and
                accelerates the retirement of older vehicles (relative to the
                baseline). In those cases, the unadjusted non-rebound VMT is higher,
                [Delta]miles smaller, and corresponding allocation of [Delta]miles
                smaller--though still consistently positive.
                 Commenters encouraged us to use a demand model to avoid creating
                unrealistic VMT projections that failed to account for factors that
                exogenously influence total demand for VMT, which the agencies have
                done here.\1814\ Had baseline case been used instead, regardless of
                whether it happens to be the most or least stringent alternative, as
                the non-rebound VMT constraint, both the non-rebound VMT and VMT with
                rebound would have differed meaningfully from both other government
                forecasts and from the projections produced by the demand models
                underlying those forecasts. By producing and enforcing a non-rebound
                constraint based on results from a travel demand model, the agencies
                ensure realism in the projections of total VMT under each regulatory
                alternative and ensure that the costs and benefits associated with
                rebound VMT result only from fuel economy improvements in the
                regulatory alternatives considered.
                ---------------------------------------------------------------------------
                 \1814\ See, e.g., NCAT, Comments, NHTSA-2018-0067-11969, at 31-
                32; Environmental Group Coalition, Appendix A, NHTSA-2018-0067-
                12000, at 175-76; UCS, Technical Appendix, NHTSA-2018-0067-12039, at
                59; Honda, Supplemental Analysis, NHTSA-2018-0067-1211, at 4.
                ---------------------------------------------------------------------------
                (d) VMT Resulting From Simulation
                 This section has already demonstrated that total VMT projections
                from the simulation are consistent with FHWA projections of total light
                duty VMT using the same set of economic assumptions. Lifetime mileage
                accumulation is now a function of the sales model, scrappage model,
                mileage accumulation schedules (described in Table VI-180), and the
                redistribution of VMT across the age distribution of registered
                vehicles in each calendar year to preserve the non-rebound VMT
                constraint.
                 The definition of ``non-rebound'' VMT in this analysis determines
                the
                [[Page 24696]]
                additional miles associated with secular fleet turnover and fuel price
                changes. Conversely, rebound miles measure the VMT difference due to
                fuel economy improvements relative to MY2016 (independent of changes in
                fuel price, or secular fleetwide fuel economy improvement resulting
                from the continued retirement of older vehicles and their replacement
                with newer ones). In order to calculate total VMT with rebound, the
                agencies apply the rebound elasticity to the full change in CPM and the
                initial VMT schedule, but apply the rebound elasticity to the
                incremental percentage change in CPM between the non-rebound and full
                CPM calculations to the miles applied to each vehicle during the
                reallocation step that ensured adjusted non-rebound VMT matched the
                non-rebound VMT constraint.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.378
                Where VMTA,S is the initial VMT schedule by age and body-style,
                %[Delta]NonReboundCPM and %[Delta]CPM are defined in Equation VI-8
                and Equation VI-7, respectively, and [Delta]MilesA,S,CY is the per-
                vehicle miles added by the reallocation described in Equation VI-13.
                The additional miles that are added to each vehicle in the
                reallocation step ([Delta]MilesA,S,CY) are multiplied by the
                difference between the percentage changes in CPM (full and non-
                rebound, respectively) because the %[Delta]NonRbdCPM was used to
                derive the allocated miles and using the full CPM change to scale
                the allocated miles would count that change twice. Taking the
                difference avoids overestimating the total mileage in the presence
                of the rebound effect. The ``rebound miles'' will be the difference
                between Equation VI-14 and Equation VI-10 for each alternative. To
                the extent that regulatory scenarios produce comparable numbers of
                rebound miles in early calendar years, the impacts associated with
                those miles net out across the alternatives in the benefit cost
                analysis.
                BILLING CODE 4910-59-P
                 Table VI-184 displays the annual survival-weighted VMT at each age
                of a MY2025 vehicle, by regulatory class including and reallocation
                needed to preserve the VMT constraint and all rebound miles (using a 20
                percent rebound effect).\1815\
                ---------------------------------------------------------------------------
                 \1815\ Annual survival-weighted VMT is calculated by dividing
                the annual VMT of a MY cohort by the total population of the cohort
                purchased. As such, Table VI-183 and Table VI-184 report different
                types of values.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.379
                [[Page 24697]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.380
                BILLING CODE 4910-59-C
                 As earlier portions of this section have shown, the second decade
                of useful life now shows significantly higher utilization than the NPRM
                analysis for both passenger cars and light trucks. While the current
                lifetime accumulation is similar to the values produced in the 2012
                final rule, those values were simulated to occur under fuel prices that
                were consistently 40 percent higher than the prices in this analysis
                (when adjusted for inflation).\1816\ Under comparable prices, lifetime
                mileage accumulation would have been considerably higher.
                ---------------------------------------------------------------------------
                 \1816\ The 2012 final rule also assumed a 10 percent rebound
                effect, which would have further affected lifetime mileage
                accumulation.
                ---------------------------------------------------------------------------
                (e) Sales, Scrappage and VMT Integration
                 The VMT construct described above, while an improvement over the
                version presented in the NPRM for the reasons explained, does not
                represent the fully integrated model of ownership, usage, and
                retirement decisions that some commenters argued would be preferred or
                even required to assess properly the impacts of CAFE/CO2
                standards. In particular, RFF commented that integrating sales,
                scrappage and VMT would ``make the analysis internally consistent and
                will account for the fact that households do not make scrappage and
                vehicle use decisions in isolation.'' \1817\ IPI concurred and expanded
                in their comment, stating `` `a
                [[Page 24698]]
                unified model of vehicle choice and usage' is necessary.'' \1818\
                ---------------------------------------------------------------------------
                 \1817\ RFF, Comments, NHTSA-2018-0067-11789 at 14.
                 \1818\ IPI, Appendix, NHTSA-2018-0067-12213, at 80 (internal
                citation omitted).
                ---------------------------------------------------------------------------
                 The implication of such commenters is that the agencies have
                ignored important benefits of more stringent standards by not
                explicitly considering household decisions at the level of household
                vehicle fleet management. However, the opposite may be true. A recent
                National Bureau of Economic Research (``NBER'') paper finds that
                households engage in attribute substitution while managing the set of
                attributes in their vehicle portfolios.\1819\ In particular, the
                authors argue that attribute substitution within a household's vehicle
                portfolio may erode up to 60 percent of the intended fuel economy
                benefits of the footprint-based CAFE/CO2 standards, as the
                higher fuel economy of owned vehicles reduces demand for efficiency in
                the next bought vehicle, all else equal. This suggests that examining
                effects at the household level may not be as beneficial, or as
                meaningful, as some commenters might hope.
                ---------------------------------------------------------------------------
                 \1819\ Archsmith, J., Gillingham, K., Knittel, C., Rapson, D.
                (Sept. 2017), Attribute Substitution in Household Vehicle
                Portfolios. NBER Working Paper No. NBER Working Paper No. 23856.
                Available at https://www.nber.org/papers/w23856 (last accessed Feb.
                4, 2020).
                ---------------------------------------------------------------------------
                 While commenters have suggested ambitious models of dynamic
                relationships at the household level, moreover, it is not clear that
                such a model is currently possible. Capturing the heterogeneous
                preferences of households across purchase, usage, and retirement
                decisions at the same level of detail required to produce meaningful
                estimates of regulatory compliance costs is beyond the current scope of
                this analysis. While the agencies agree that expected usage influences
                the household decision of which vehicle to purchase, how long to hold
                it, and how to manage the usage and retirement of other vehicles within
                a household fleet, the agencies do not agree that such a detailed model
                is a necessary prerequisite to assess the impacts of CAFE and tailpipe
                CO2 emissions standards, nor that it is necessarily
                appropriate to do so given that the agencies are examining aggregate
                national fleetwide effects of such standards. Furthermore, in the most
                recent peer review of the CAFE Model, one reviewer remarked that while
                the sales and VMT would benefit from a household choice model, ``the
                decision to scrap a vehicle (remove it from the national in-use fleet)
                and the decision to purchase a new vehicle often are not made by the
                same household. No U.S. national-level transportation demand models
                (that this reviewer is aware of) tackle the issue with this level of
                complexity.'' \1820\
                ---------------------------------------------------------------------------
                 \1820\ CAFE Model Peer Review, DOT HS 812 590, Revised (July
                2019), pp. B19-B29, available at https://www.regulations.gov/contentStreamer?documentId=NHTSA-2018-0067-0055&attachmentNumber=2&contentType=pdf.
                ---------------------------------------------------------------------------
                 Each iteration of these regulatory analyses has endeavored to
                improve the accuracy and breadth of modeling to capture better the
                relevant dynamics of the markets affected by these policies. The
                agencies intend to address current limitations in future rulemakings,
                and meanwhile believe that the scope of the current analysis is
                reasonable and appropriate for informing decision-makers as to the
                effects of different levels of CAFE and tailpipe CO2
                emissions stringency.
                (6) What is the mobility benefit that accrues to vehicle owners?
                (s) Mobility Benefits in the NPRM Analysis
                 As the proposal noted, the increase in travel associated with the
                rebound effect provides benefits that reflect the value to drivers and
                other vehicle occupants of the added--or more desirable--social and
                economic opportunities that become accessible with additional travel.
                The fact that drivers and their passengers elect to make more frequent
                or longer trips to gain access to these opportunities when the cost of
                driving declines demonstrates that the benefits they gain by doing so
                exceed the costs they incur, including the economic value of their
                travel time, fuel and other vehicle operating costs, and the economic
                cost of safety risks drivers assume. The amount by which the benefits
                of this additional travel exceeds its economic costs measures the net
                benefits drivers and their passengers experience, usually referred to
                as increased consumer surplus.
                 Under the proposal, the fuel cost of driving each mile would have
                increased as a consequence of the lower fuel economy levels it
                permitted, thus reducing the number of miles that buyers of new cars
                and light trucks would drive as the well-documented fuel economy
                rebound effect operates in reverse.\1821\ The agencies' analysis of the
                proposed rule described the resulting loss in consumer surplus, and
                calculated its annual value using the conventional approximation, which
                is one half of the product of the increase in vehicle operating costs
                per vehicle-mile and the resulting decrease in the annual number of
                miles driven. Because the value of this loss depends on the extent of
                the change in fuel economy, it varied by model year, and also differed
                among the alternative standards that the NPRM considered.
                ---------------------------------------------------------------------------
                 \1821\ Normally, the fuel economy rebound effect refers to an
                increase in vehicle use that results when increased fuel economy
                reduces the fuel cost for driving each mile.
                ---------------------------------------------------------------------------
                 The agencies' analysis specifically recognized that the economic
                value of any additional travel prompted by the fuel economy rebound
                effect must exceed the additional fuel costs drivers incur, plus the
                economic cost of safety risks they and their passengers assume.\1822\
                Thus, when vehicle use was projected to decline in response to lower
                fuel economy, the agencies noted that the resulting loss in benefits
                must have more than offset both the savings in fuel costs and the value
                of drivers' and passengers' reduced exposure to safety risks. In the
                accounting of benefits and costs for the preferred alternative, the
                loss of benefits associated with reduced mobility was recognized by
                reporting losses in travel benefits that exactly offset the value of
                reduced risks of being involved in both fatal and non-fatal crashes.
                ---------------------------------------------------------------------------
                 \1822\ Although it did not attempt to estimate operating costs
                other than those for fuel or the value of drivers' and passengers'
                travel time, the benefits from any additional travel that occurs
                voluntarily must also at least compensate for these costs.
                ---------------------------------------------------------------------------
                 In addition, the accounting reported a loss in mobility benefits
                from reduced use of new cars and light trucks, which included a
                component that exactly offset the fuel savings from reduced driving,
                together with the loss in consumer surplus that foregone travel would
                otherwise have provided. Including this first component was necessary
                to offset the fact that the savings in fuel costs had already been
                recognized elsewhere in the accounting, by deducting those savings from
                the increase in fuel costs resulting from lower fuel economy to arrive
                at the reported net increase in fuel costs. Thus, the resulting value
                of the net loss in travel benefits was exactly equal to the loss in
                consumer surplus that any travel foregone in response to higher fuel
                costs would otherwise have provided.
                (b) Comments on the Agencies' Treatment of Mobility Benefits in the
                NPRM
                 The agencies received only two comments referring to their
                treatment of mobility benefits in the analysis supporting the proposed
                CAFE and CO2 standards. The California Air Resources Board
                (CARB) noted that the accounting of benefits and costs resulting from
                the proposal included losses in mobility benefits that offset the
                reduction in fatality costs related to the decline in
                [[Page 24699]]
                new vehicle use from the fuel economy rebound effect. While CARB did
                not comment on the agencies' inclusion of losses in mobility benefits
                in their accounting, it did object to the fact that the agencies also
                reported the numerical change in fatalities that could be ascribed to
                the rebound effect, and considered the improvement in safety it
                reflected when selecting their proposed alternative.\1823\ Similarly,
                the Institute for Policy Integrity (IPI) termed the agencies' reliance
                on the estimated change in the number of fatalities as partial
                justification for selecting their preferred alternative as arbitrary,
                while at the same time arguing that the reduction in driving due to the
                rebound effect had no net welfare impact.\1824\
                ---------------------------------------------------------------------------
                 \1823\ California Air Resources Board (CARB), NHTSA-2018-0067-
                11873, at pp. 121.
                 \1824\ Institute for Policy Integrity (IPI), NHTSA-2018-0067-
                12213, at pp. 11. In fact, the agencies did not treat the reduction
                in driving as having no net impact on welfare, since as explained
                immediately above, the loss in consumer surplus benefits on the
                foregone driving was not accompanied by any offsetting cost savings.
                Therefore, the decline in driving in response to the rebound effect
                resulted in a net loss in welfare.
                ---------------------------------------------------------------------------
                 In response to these comments, the agencies observe that
                considering changes in the actual number of fatalities as well as the
                welfare effects of changes in drivers' and passengers' exposure and
                valuation of the risks of being involved in fatal crashes represents a
                sound approach to assessing the impacts of proposed CAFE and
                CO2 standards. The safety implications of alternative future
                standards are clearly a legitimate and highly visible consequence for
                the agencies to consider when evaluating their relative merits, as are
                the implications of changes in the safety risks for the economic
                welfare of car and light truck users. Thus the agencies see no
                inconsistency or duplication in separately considering both factors as
                part of their assessment of alternative future standards.
                (c) Mobility Benefits in the Final Rule
                 The analysis supporting this final rule continues to treat losses
                in mobility benefits in the same manner the agencies previously did
                when analyzing the alternatives considered for the proposed rule.
                Because there are several subtleties in this treatment, Figure VI-75 is
                included below to clarify its details. In the figure, the demand curve
                shows the relationship of annual use of new cars (and light trucks),
                which can be thought of as their total or average annual vehicle-miles
                driven, to the cost per mile of driving.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.381
                 The initial cost per mile OC0 consists of the per mile
                economic costs of the risks of being involved in fatal and non-fatal
                crashes, shown by the heights of Og and gd on the vertical axis,
                together with per-mile fuel costs at the baseline level of fuel
                economy, the height of segment dC0.\1825\ Annual miles
                driven at this initial per-mile cost are shown by the distance
                OM0 on the horizontal axis in Figure VI-75. When fuel
                economy declines from its baseline level under one of the regulatory
                alternatives considered, fuel costs per mile increase from
                dC0 to dC1, but the per-mile economic costs of
                crash risks (both fatal and non-fatal) are unaffected, so total costs
                per mile driven rise to OC1. In response to this increase in
                the per-mile fuel and total cost of driving, annual use declines to
                OM1.
                ---------------------------------------------------------------------------
                 \1825\ Per-mile fuel costs are equal to the dollar price of fuel
                per gallon, divided by fuel economy in miles per gallon. For
                simplicity, this figure omits non-fuel operating costs, vehicle
                maintenance and depreciation, and the value of occupants' travel
                time. Including them would not change the analysis.
                ---------------------------------------------------------------------------
                 The resulting loss in total benefits when vehicle use declines from
                OM0 to OM1 is the trapezoidal area
                M1acM0, but most of this loss is offset by cost
                savings from reduced driving, so the net welfare loss is considerably
                smaller. Specifically, the rectangle M1hiM0
                represents a reduction in the total economic costs of the risk that
                drivers and passengers will be involved in fatal crashes when the
                decline in driving
                [[Page 24700]]
                reduces their exposure to that risk. The dollar value of this area thus
                appears in the agencies' accounting of costs and benefits as both a
                benefit from that reduction in risk and an exactly offsetting loss in
                benefits from reduced mobility. The same is true of the rectangle hefi,
                the dollar value of which corresponds to both the reduction in the
                economic cost of non-fatal crash risks and an identical loss in
                mobility benefits.
                 Total fuel costs for driving OM0 miles are initially the
                rectangular area dC0cf, and the decline in driving to
                OM1 that results as per-mile fuel and total driving costs
                rise changes total fuel costs to the rectangle dC1ae.
                Because these two areas share rectangle dC0be, the net
                change in fuel costs reported in the agencies' accounting consists of
                the dollar value of rectangle C0C1ab, minus that
                of rectangle ebcf. The economic value of the loss in mobility benefits
                the agencies report in their accounting is the trapezoid eacf, but part
                of that area consists of rectangle ebcf, and is thus exactly equal to
                the savings in fuel costs from reduced driving. Since this savings has
                been already incorporated in the reported change in total fuel costs,
                and it offsets part of the reported loss in mobility benefits, leaving
                only the loss in consumer surplus that travelers would otherwise have
                experienced on foregone reduced driving, the value of triangle bac, as
                the net loss in mobility benefits.\1826\
                ---------------------------------------------------------------------------
                 \1826\ Thus the change in driving is not welfare-neutral, as IPI
                asserted in the comment cited previously; instead, it results in a
                net loss in welfare.
                ---------------------------------------------------------------------------
                 This discussion assumes that drivers correctly estimate and
                consider--or ``internalize''--the risks of being involved in both fatal
                and non-fatal crashes that are associated with their additional
                driving. However, as is noted in the discussion of the potential
                effects of the rule on the mass of vehicles and its resulting impact on
                safety, consumers may value safety risks imperfectly. This possibility
                is accounted for in the final rule analysis by assuming the portion of
                the added safety risk that consumers internalize to be 90 percent. In
                Figure VI-75 above, this would be reflected by including a total social
                cost per mile that is higher than the C0 and C1
                values for the baseline and reduced MPG cases shown in the graphic by
                10 percent of the combined cost of fatal and non-fatal crash risks (the
                distance Od on the figure's vertical axis), while reducing the costs of
                safety risks that drivers do consider to 90 percent of the values
                shown. The higher social costs would offset a portion of the consumer
                surplus associated with additional mobility (in each case), and result
                in a small ``deadweight loss'' over the region where the social cost of
                driving exceeds the demand curve. These impacts are also fully
                accounted for in the final rule analysis.
                (7) What is the sales surplus that accrues to vehicle owners?
                 Buyers who would not have purchased new models with the baseline
                standards in effect but decide to do so in response to the changes in
                new vehicles' prices with less demanding standards in place will also
                experience increased welfare. Collective benefits to these ``new''
                buyers are measured by the consumer surplus they receive from their
                increased purchases.
                 At the proposed rule stage, the agencies elected to exclude the
                consumer surplus associated with new vehicle purchases because ``it is
                not entirely certain that sales of new cars and light trucks [would]
                increase in response to [the] proposed action.'' \1827\ Consumer
                surplus is a fundamental economic concept and represents the net value
                (or net benefit) a good or service provides to consumers. It is
                measured as the difference between what a consumer is willing to pay
                for a good or service and the market price. OMB circular A-4 explicitly
                identifies consumer surplus as a benefit that should be accounted for
                in cost-benefit analysis. For instance, OMB Circular A-4 states the
                ``net reduction in total surplus (consumer plus producer) is a real
                cost to society,'' and elsewhere elaborates that consumer surplus
                values be monetized ``when they are significant.'' \1828\
                ---------------------------------------------------------------------------
                 \1827\ See PRIA at 954.
                 \1828\ OMB Circular A-4, at 37-38.
                ---------------------------------------------------------------------------
                 The decision to exclude consumer surplus for new vehicles at the
                proposed rule stage was an error and inconsistent with OMB's guidance
                on regulatory analysis. The agencies are confident that lower vehicle
                prices, holding all else equal, should stimulate new vehicle sales and
                by extension produce additional consumer surplus. That preliminary
                decision was also inconsistent with other parts of the agencies'
                analysis. For instance, the agencies calculate the lost consumer
                surplus associated with reductions in driving owing to the increase in
                the cost per mile in less stringent regulatory cases, as discussed in
                Section VI.D.3. The surpluses associated with sales and additional
                mobility are inextricably linked as they capture the direct costs and
                benefits accrued by purchasers of new vehicles. The sales surplus
                captures the savings to consumers when they purchase cheaper vehicles
                and the additional mobility measures the cost of higher operating
                expenses. It would be inappropriate to include one without the other.
                 The shaded area in Figure VI-76 reflects the consumer surplus
                calculated for new vehicle sales. Line C0 reflects the baseline vehicle
                cost. The final rule is expected to reduce the cost of light duty
                vehicles, as represented by dotted line C '. Consistent with other
                sections of the analysis, the agencies assume that consumers value 30
                months of fuel savings. Under the final rule, consumers are expected to
                experience higher fuel costs than they would under the baseline
                scenario, shifting costs from line C ' to line C1. The consumer surplus
                is equal to the area under the curve between Q0 and Q1.\1829\
                ---------------------------------------------------------------------------
                 \1829\ The exact calculation is 0.5 * the increase in sales *
                the reduction in the cost of light duty vehicles net of the
                increased fuel cost.
                ---------------------------------------------------------------------------
                [[Page 24701]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.382
                (8) Implicit Opportunity Cost
                 The agencies' central analysis assumes the selling price for new
                vehicles will be reduced to fully reflect manufacturers' savings in
                technology costs for complying with less stringent CAFE and
                CO2 emission standards. Specifically, new car and light
                truck prices are assumed to decline by the average savings in
                technology costs per vehicle that manufacturers would realize from
                complying with the standards this rule establishes, instead of with the
                more demanding baseline standards. The agencies' analysis assumes that
                under these final standards, attributes of new cars and light trucks
                other than fuel economy would remain identical to those under the
                baseline standards, so that changes in sales prices and fuel economy
                would be the only sources of benefits or costs to new car and light
                truck buyers. Furthermore, the agencies recognize that buyers may have
                time preferences that cause them to discount the future at higher rates
                than the agencies are directed to consider in their regulatory
                evaluations. In either case, the agencies' central analysis may
                overstate both the net private and social benefits from adopting more
                stringent fuel economy and CO2 emissions standards. For
                instance, Table VII-93 (Combined LDV Societal Net Benefits for MYs
                1975-2029, CAFE Program, 7 percent Discount Rate) shows that the CAFE
                final rule would generate $16.1 billion in total social net benefits
                using a 7 percent discount rate, but without the large net private loss
                of $26.1 billion, the net social benefits would equal the external net
                benefits, or $42.4 billion. Therefore, given that government action
                cannot improve net social benefits absent a market failure, if no
                market failure exists to motivate the $26.1 billion in private losses
                to consumers, the net benefits of these final standards are $42.2
                billion.
                 As indicated earlier, EPA's Science Advisory Board urged the
                agencies to account for ``consumer preferences for performance and
                other vehicle attributes'' in their analysis.\1830\ To explore further
                the possibility that the central analysis is incomplete regarding the
                consumer benefits of other vehicle attributes, the agencies conducted a
                sensitivity analysis using a conservative estimate of this value. In
                the proposal, the agencies considered the lost value of other vehicle
                attributes in two sensitivity cases that reduced the total consumer
                benefit.\1831\ The agencies received several comments suggesting that
                the analysis of other vehicle attributes lost could be improved. For
                example, CARB commented that the ``analyses do not adequately model how
                vehicle values will change in response to improving fuel economy, or
                the competing effects of other attributes.'' \1832\ In response to
                commenters, the agencies have revised their sensitivity analyses to
                model better the impact of the standards on other vehicle attributes.
                ---------------------------------------------------------------------------
                 \1830\ SAB at 10.
                 \1831\ See PRIA at 954. See also, PRIA at 1539.
                 \1832\ CARB, Detailed Comments, NHTSA-2018-0067-11873 at 189.
                ---------------------------------------------------------------------------
                 The agencies considered, such as they did in the proposal,
                offsetting the net private costs associated with enabling more choices
                in fuel-saving technologies in a manner similar to rebound driving.
                However, the agencies believe that this approach is unnecessary, as
                such an analysis would produce nearly identical net benefits to the
                external net benefits--which the primary analysis already generates.
                Furthermore, given that consumers are free to choose more fuel-
                efficient vehicles absent more stringent regulations, consumers who
                prefer certain vehicle attributes instead of fuel economy necessarily
                value those attributes more than the fuel efficiency technologies they
                voluntarily forgo. As such, a sensitivity analysis including a value
                for other vehicle attributes should more than offset the net private
                costs to consumers from the primary analysis.
                 For the final rule, instead of keeping the same approach as the
                preliminary analysis, the agencies have elected to estimate consumer
                benefits of other vehicle attributes in a sensitivity case using
                similar logic to that used for the sales and scrappage models. In those
                models, the agencies assume that consumers value thirty months of
                undiscounted fuel savings. Given this assumption, it would be
                reasonable for the agencies then to assume that the value of other
                vehicle attributes must be greater than the fuel savings for the
                remaining term of the useful life of the vehicle--as these are fuel
                economy savings that consumers are clearly
                [[Page 24702]]
                willing to forgo. The agencies acknowledge that vehicles are typically
                sold more than once, but evidence suggests that fuel savings are
                capitalized into sales prices in the used car market.\1833\ If this is
                the case, new car purchasers would internalize the additional value on
                resale owing to fuel efficiency technologies, and the fuel savings over
                the remaining useful life less thirty months would be an appropriate
                value to use for the value of other vehicle attributes. Nevertheless,
                the agencies have elected to be conservative and, instead, opted to use
                the fuel savings over the first seventy-two months (less the first
                thirty months), which approximates the amount of time the first owner
                typically holds a new vehicle.\1834\ This value is referred to as the
                ``implicit opportunity cost'' of forgoing other vehicle attributes in
                favor of increased fuel economy (or using their scarce financial
                resources to invest in savings or the purchase of other goods that they
                prefer more than fuel economy),\1835\ showing a cost savings for less
                stringent alternatives.\1836\ Unlike the sales surplus, which measures
                the consumer surplus of new vehicle buyers entering the market, the
                implicit opportunity cost contained in this sensitivity case represents
                the forgone benefits to consumers the model assumes would have
                purchased a vehicle regardless of the standards (but would prefer to
                take the upfront cost of fuel economy technologies and invest that
                money elsewhere, whether it be on different vehicle attributes or
                different goods altogether). These results are shown in Table VII-91
                through Table VII-95 (Combined LDV Societal Net Benefits (Accounting
                for Implicit Opportunity Cost) for MYs 1975-2029 CAFE Program, 3
                percent Discount Rate and 7 percent Discount Rate, as well as the C02
                Program, 3 percent Discount Rate and 7 percent Discount Rate).
                ---------------------------------------------------------------------------
                 \1833\ For further discussion of the evidence, see section
                VI.D.2 of the preamble.
                 \1834\ There are several reasons why 72 months is an appropriate
                approximation. According to a report from the Federal Reserve bank
                of Chicago the average new vehicle is owned for over 77 months as of
                2015. From the same report, the average new car financing term was
                over 67 months in 2016. (https://www.chicagofed.org/publications/working-papers/2019/2019-04; accessed: December 23, 2019). Data from
                R.L. Polk suggest that the average new car is held for 71.4 months
                (as cited in https://www.autotrader.com/car-shopping/buying-car-how-long-can-you-expect-car-last-240725). State Comptrollers and
                Treasurers referred to an IHS Markit report that the average length
                of time a consumer keeps a new car is approximately 6.6 years (78
                months). EPA-HQ-OAR-2018-0283-4153, at 2. CFA commented that new
                vehicle leases are running, on average, 68 months and new vehicles
                are being held, on average, longer than 60 months. Comments, NHTSA-
                2018-0067-12005, at 76. The agencies selection of 72 months is
                comfortably within the range of these estimates, but errs towards
                the lower-end and therefore provides a conservative estimate.
                 \1835\ These vehicle attributes may include any that consumers
                may value and are not explicitly modeled to be neutral across
                regulatory alternatives. For instance, trim levels, entertainment
                systems, crash avoidance technologies, etc. may be sacrificed to pay
                for higher fuel economy technology levels.
                 \1836\ The implicit opportunity cost must be considered a value
                that consumers place on other vehicle attributes that is net of the
                cost of those attributes. This is the forgone consumer surplus of
                other vehicle attributes. As such it is appropriately additive to
                the technology cost/savings estimated in the primary analysis.
                ---------------------------------------------------------------------------
                 The agencies note that the central analysis of the final rule
                features a conservative treatment of private benefits and costs that
                may bias the results in the favor of more stringent regulatory
                alternatives. This bias arises from the agencies' treatment of rebound
                driving. The agencies assume that drivers make a rational decision when
                electing to drive additional miles, which considers not only the risks
                the additional driving poses to their own lives and property, but also
                most of the risks their behavior poses to their passengers as well as
                the person and property of other road users. In such a case, drivers
                ``internalize'' most of these risks, and it can be assumed that
                benefits to drivers must be more valuable to them than the risks they
                considered when deciding whether to undertake the additional driving.
                Therefore, the agencies have appropriately offset the loss in safety
                benefits, which are associated with the increased cost of driving in
                the final rule, with commensurate lost benefits of additional driving.
                 In contrast, the agencies can be assured the private benefits and
                costs of fuel saving technologies (aside from the external
                environmental damages) are internalized--as there is no doubt that the
                owners of the vehicles will accrue the fuel costs/savings. The agencies
                believe it would be entirely contradictory to assert that consumers are
                rational, informed, and considerate enough to internalize the risks of
                additional driving to themselves, their passengers, as well as other
                drivers and passengers; but are not similarly rational and informed
                enough to consider the additional fuel costs of purchasing a vehicle
                without a particular fuel-saving technology. After all, existing
                regulations require that the estimated annual fuel costs of a vehicle
                are disclosed on the new vehicle a consumer intends to purchase--and no
                such disclosure exists for the risks associated with driving a rebound
                mile. The agencies' decision to offset rebound miles, but not net
                private costs stemming from enabling more choices in fuel-saving
                technologies, significantly favors more stringent alternatives.
                 Another possibility, however, is that manufacturers could redirect
                some or all of their savings in technology costs to instead improve
                other attributes of cars and light trucks--passenger comfort, safety,
                carrying and towing capacity, or performance--that potential buyers
                value. For example, they could redeploy the energy efficiency
                improvements from some technologies that would otherwise have been used
                to increase fuel economy to instead improve vehicles' performance, or
                redirect spending on fuel economy technology to improve safety or
                interior comfort. Producers could also offer combinations of price
                reductions and more limited improvements in these other attributes on
                some of their models, while continuing to offer high levels of fuel
                economy on other models, and channeling their entire cost savings into
                price reductions on yet other vehicles. Individual manufacturers would
                presumably select different combinations of these strategies, each in
                an effort to realize maximum additional sales and profits.
                 The agencies' analysis does not quantify specific improvements in
                other attributes manufacturers could make, or identify potential
                combinations of lower prices and improvements in other attributes they
                might offer when they face less demanding fuel economy and
                CO2 standards. Nevertheless, there is ample empirical
                evidence that tradeoffs among fuel economy and other attributes that
                buyers value are important considerations in vehicle design and
                marketing strategy, and that manufacturers commonly offer combinations
                of both higher fuel economy and improvements in other attributes when
                standards do not require them to focus exclusively on improving fuel
                economy.
                 Table VI-185 summarizes empirical estimates of the tradeoffs among
                fuel economy, horsepower (for cars) or torque (for light trucks), and
                weight derived from different authors' econometric estimates of the
                ``curvature'' of technology frontiers for cars and light trucks. Such
                frontiers describe the combinations of fuel economy and other
                attributes that manufacturers can provide with different levels of
                spending on vehicle design and technology, accounting for the gradual
                improvements in technology and energy efficiency that occur over time.
                The entries in the table show different authors' estimates of the
                percent increases in horsepower, torque, and weight that car and light
                truck manufacturers could instead achieve if
                [[Page 24703]]
                they reduced fuel economy by one percent. (Although increased weight is
                not desirable in and of itself, it is associated with features such as
                a vehicle's passenger- and cargo-carrying capacity, interior volume,
                comfort, and safety, which potential buyers do value.). It is important
                to note that these tradeoffs apply to the overall average values of
                each attribute for cars and light trucks produced during recent model
                years, rather than to the features of specific individual models.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.383
                 For example, Table VI-185 shows that Klier & Linn estimate reducing
                the average fuel economy of cars by one percent would enable producers
                to increase their average horsepower by 0.24 percent, and Knittel's
                estimate of that tradeoff is very similar (0.26 percent). Similarly,
                those two studies estimate that reducing the average fuel economy of
                cars and light trucks by one percent would enable their weight to be
                increased by 0.34-0.39 percent, which would in turn enable
                manufacturers to make modest improvements in their passenger- and
                cargo-carrying capacity, interior volume, comfort, or safety. (Note
                that reducing average fuel economy by one percent would permit either
                power or weight to increase as indicated in the table, but not both at
                the same time.).
                 The tradeoffs summarized in Table VI-185 provide some indication of
                changes in attributes other than fuel economy that manufacturers are
                likely to offer under the less demanding CAFE and CO2
                standards. For example, the agencies estimate that the baseline CAFE
                standards would have required increases in fuel economy approximately 5
                percent annually over model years 2020-26 for cars, while this rule
                reduces the required rate of increase to 1.5 percent annually. This
                less demanding standard would thus enable producers to accompany higher
                fuel economy with significant improvements in other features that new
                car buyers also value, as an alternative to simply reducing prices to
                reflect their savings in technology costs. As noted previously, they
                would do so only if they thought such a strategy would be more
                attractive to buyers, so the agencies' estimates of benefits to new car
                and light truck buyers represents the minimum improvement in utility
                they would realize.
                 The historical evolution of car and light truck characteristics
                under CAFE standards may also provide some indication about how
                manufacturers are likely to respond to the less aggressive standards
                this rule establishes. Figure VI-77 and Figure VI-78 show that during
                the period when CAFE standards remained unchanged or increased slowly--
                approximately 1985-2010--manufacturers gradually improved cars' and
                light trucks' average fuel economy as well as their power (or torque)
                and weight, while only modestly increasing the average interior volume
                of cars.
                BILLING CODE 4910-59-P
                [[Page 24704]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.384
                [[Page 24705]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.385
                BILLING CODE 4910-59-C
                 Table VI-186 summarizes the rates of change in fuel economy and
                other attributes of cars and light trucks over that period. As it
                shows, most advances in cars' drive train technology were used to
                increase power and fuel economy, while most of the improvement in light
                trucks' energy efficiency was channeled into higher torque and weight,
                with relatively little used to improve fuel economy.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.386
                 The last column of Table VI-186 combines the actual historical
                rates of increase in attributes other than fuel economy with the
                tradeoffs between fuel economy and other attributes shown previously in
                Table VI-185 to estimate the annual rates of increase in fuel economy
                that could have been achieved if all technological progress had been
                channeled into improving fuel economy. As it indicates, manufacturers
                could have increased the fuel economy of both cars and light trucks
                over the period spanned by Table VI-186 at almost exactly the 1.5
                percent annual rate this rule requires, if they had believed that
                sacrificing other improvements in the interest of achieving higher fuel
                economy was the most effective strategy to meet potential customers'
                demands.
                [[Page 24706]]
                 While this result should be regarded as illustrative, it appears to
                show that meeting even these relaxed standards may require
                manufacturers to focus on improving fuel economy instead of other
                vehicle attributes. It also suggests that meeting the more demanding
                baseline standards may have required manufacturers to make significant
                sacrifices in other attributes, rather than simply holding those other
                features at or near their current levels. Viewed from this perspective,
                while this rule might not enable manufacturers to improve other
                desirable features of cars and light trucks at the same time as they
                provide the improvements in fuel economy it requires, it may
                nevertheless prevent them from having to sacrifice other improvements
                that buyers regard as valuable in order to focus solely on complying
                with more demanding CAFE and CO2 standards.
                (9) Additional Consumer Purchase Costs
                 Some costs of purchasing and operating new and used vehicles scale
                with the value of the vehicle. When fuel economy standards increase the
                price of new vehicles, both taxes and registration fees increase, too,
                because they are calculated as a percentage of vehicle price.
                Increasing the price of new vehicles also affects the average amount
                paid on interest for financed vehicles and the insurance premiums for
                similar reasons. The agencies compute these additional costs as scalar
                multipliers on the MSRP of new vehicles. These costs are included in
                the consumer per-vehicle cost-benefit analysis, but, for the reasons
                described below, are not included in the societal cost-benefit
                analysis.
                 It is worth noting that these costs are not included in the sales
                and scrappage models, discussed above. The agencies do not expect that
                the omission of these costs affects the sales and scrappage models
                because of how these additional costs are calculated in the modeling.
                These costs are assumed to be a fixed scalar on the average MSRP of new
                vehicles, so that their inclusion would simply scale the coefficients
                in the sales and scrappage models. While these costs have not stayed
                constant over time (particularly not over the times series from 1970 to
                today), the agencies do not have a time series dataset to accurately
                estimate these costs.
                 The agencies hope to reconsider including sales taxes, registration
                fees, additional interest payments and insurance costs in the sales and
                scrappage models in future research.
                (a) Sales Taxes and Registration Fees
                 In the analysis, sales taxes and registration fees are considered
                transfer payments between consumers and the government and are
                therefore not considered a cost from the societal perspective. However,
                these costs do represent an additional cost to consumers and are
                accounted for in the private consumer perspective. To estimate the
                sales tax for the analysis, the agencies weighted the auto sales tax of
                each state by its population--using Census population data--to
                calculate a national weighted-average sales tax of 5.46%.\1837\
                ---------------------------------------------------------------------------
                 \1837\ See Car Tax by State, FactoryWarrantyList.com, http://www.factorywarrantylist.com/car-tax-by-state.html (last visited June
                22, 2018). Note: County, city, and other municipality-specific taxes
                were excluded from weighted averages, as the variation in locality
                taxes within states, lack of accessible documentation of locality
                rates, and lack of availability of weights to apply to locality
                taxes complicate the ability to reliably analyze the subject at this
                level of detail. Localities with relatively high automobile sales
                taxes may have relatively fewer auto dealerships, as consumers would
                endeavor to purchase vehicles in areas with lower locality taxes,
                therefore reducing the effect of the exclusion of municipality-
                specific taxes from this analysis.
                ---------------------------------------------------------------------------
                 The agencies recognize that weighting state sales tax by new
                vehicle purchases within a state would likely produce a better estimate
                since new vehicle purchasers represent a small subset of the population
                and may differ between states. The agencies explored using Polk
                registration data to approximate new vehicle sales by state by
                examining the change in new vehicle registrations across several recent
                years. The results derived from this examination resulted in a national
                weighted-average sales tax rate slightly above 5.5%, which is almost
                identical to the rate calculated using population instead. The agencies
                opted to utilize the population estimate, rather than the registration-
                based proxy of new vehicle sales, because the results were negligibly
                different and the analytical approach involving new vehicle
                registrations has not been as thoroughly reviewed.
                (b) Financing Costs
                 Consumers who purchase new vehicles with financing options incur an
                additional cost above the new vehicle price--interest. Based off an
                Experian data, \1838\ the analysis assumes 85% of automobiles are
                purchased through financing options. The analysis used data from Wards
                Automotive and JD Power on the average transaction price of new vehicle
                purchases, average principle of new auto loans, and the average OEM-
                offered incentive as a percent of MSRP to compute the ratio of the
                average financed new auto principal to the average new vehicle MSRP for
                calendar years 2011-2016. Table VI-187 shows that the average financed
                auto principal was between 82% and 84% of the average new vehicle MSRP.
                Applying the assumption that 85% of new vehicle purchases involve some
                financing, the average share of the MSRP financed for all vehicles
                purchased, including non-financed transactions, was computed. Table-II-
                34 shows that the average percentage of MSRP financed ranges between
                70% and 72%. From this, the agencies chose to assume that 70% of the
                value of all vehicles' MSRP is financed. It is likely that the share
                financed is correlated with the MSRP of the new vehicle purchased, but
                for simplification purposes, it is assumed that 70% of all vehicle
                costs are financed, regardless of the MSRP of the vehicle. The agencies
                note that this simplification does not impact the accuracy of the
                calculation of the average cost to consumers, but concede that it
                obfuscates which consumers bear the additional financing burden when
                vehicle prices increase (selection of specific vehicles is likely not
                independent of consumer characteristics). For sake of simplicity, the
                model also assumes that increasing the cost of new vehicles will not
                change the share of new vehicle MSRP that is financed; the relatively
                constant share from 2011-2016 when the average MSRP of a vehicle
                increased 10% supports this assumption. The agencies recognize that
                this is not indicative of average individual consumer transactions but
                provides a useful tool to analyze the aggregate marketplace.
                ---------------------------------------------------------------------------
                 \1838\ A report by Experian found that 85.2% of 2016 new
                vehicles were financed, as were 85.9% of 2015 new vehicle purchases.
                Zabritski, M. State of the Automotive Finance Market: A look at
                loans and leases in Q4 2016, Experian, https://www.experian.com/assets/automotive/quarterly-webinars/2016-Q4-SAFM-revised.pdf (last
                visited June 22, 2018).
                ---------------------------------------------------------------------------
                [[Page 24707]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.387
                 From Wards Auto data, the average 48- and 60-month new auto
                interest rates were 4.25% in 2016, and the average finance term length
                for new autos was 68 months. The agencies recognize that longer
                financing terms generally include higher interest rates. The share
                financed, interest rate, and finance term length are added as inputs in
                the parameters file so that they are easier to update in the future.
                 Using these inputs the model computes the stream of additional
                costs associated with financing options paid for the average financed
                purchases as follows: \1839\
                ---------------------------------------------------------------------------
                 \1839\ As alluded to above, the principle portion of repayments
                do not represent an additional cost to consumers since it represents
                the sales price.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.600
                 0Note: The above assumes the interest is distributed evenly over
                the period, when in reality more of the interest is paid during the
                beginning of the term. However, the incremental amount calculated as
                attributable to the standard will represent the difference in the
                annual payments at the time that they are paid, assuming that a
                consumer does not repay early. This will represent the expected
                ---------------------------------------------------------------------------
                change in the stream of financing payments at the time of financing.
                 The above stream does not equate to the average amount paid to
                finance the purchase of a new vehicle. In order to compute this amount,
                the share of financed transactions at each interest rate and term
                combination would have to be known. Without having projections of the
                full distribution of the auto finance market into the future, the above
                methodology reasonably accounts for the increased amount of financing
                costs due to the purchase of a more expensive vehicle, on an average
                basis taking into account non-financed transactions. Financing payments
                are also assumed to be an intertemporal transfer of wealth for a
                consumer; for this reason, it is not included in the societal cost and
                benefit analysis. However, because it is an additional cost paid by the
                consumer, it is calculated as a part of the private consumer welfare
                analysis.
                 It is recognized that increased financing terms, combined with
                rising interest rates, lead to longer periods before a consumer will
                have positive equity in the vehicle to trade in toward the purchase of
                a newer vehicle. This has impacts in terms of consumers either trading
                vehicles with negative equity (thereby increasing the amount financed
                and potentially subjecting the consumer to higher interest rates and/or
                rendering the consumer unable to obtaining financing) or delaying the
                replacement of the vehicle until they achieve suitably positive equity
                to allow for a trade.
                (c) Insurance Costs
                 More expensive vehicles will require more expensive collision and
                comprehensive (e.g., fire and theft) car insurance. Actuarially fair
                insurance premiums for these components of value-based insurance will
                be the amount an insurance company will pay out in the case of an
                incident type weighted by the risk of that type of incident occurring.
                For simplicity of this calculation, the agencies assume that the
                vehicle has the same exposure to harm throughout its lifetime. However,
                the value of vehicles will decline at some depreciation rate so that
                the absolute amount paid in value-related insurance will decline as the
                vehicle depreciates. This is represented in the model as the following
                stream of expected collision and comprehensive insurance payments:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.388
                [[Page 24708]]
                 To utilize the above framework, estimates of the share of MSRP paid
                on collision and comprehensive insurance and of annual vehicle
                depreciations are needed to implement the above equation. Wards has
                data on the average annual amount paid by model year for new light
                trucks and passenger cars on collision, comprehensive and damage and
                liability insurance for model years 1992-2003; for model years 2004-
                2016, they only offer the total amount paid for insurance premiums. The
                share of total insurance premiums paid for collision and comprehensive
                coverage was computed for 1979-2003. For cars the share ranges from 49
                to 55%, with the share tending to be largest towards the end of the
                series. For trucks the share ranges from 43 to 61%, again, with the
                share increasing towards the end of the series. It is assumed that for
                model years 2004-2016, 60% of insurance premiums for trucks, and 55%
                for cars, is paid for collision and comprehensive. Using these shares
                the absolute amount paid for collision and comprehensive coverage for
                cars and trucks is computed. Then each regulatory class in the fleet is
                weighted by share to estimate the overall average amount paid for
                collision and comprehensive insurance by model year as shown in Table
                VI-188. The average share of the initial MSRP paid in collision and
                comprehensive insurance by model year is then computed. The average
                share paid for model years 2010-2016 is 1.83% of the initial MSRP. This
                is used as the share of the value of a new vehicle paid for collision
                and comprehensive in the future.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.389
                 2017 data from Fitch Black Book was used as a source for vehicle
                depreciation rates; two- to six-year-old vehicles in 2016 had an
                average annual depreciation rate of 17.3%.\1840\ It is assumed that
                future depreciation rates will be like recent depreciation, and the
                analysis used the same assumed depreciation. Table VI-189 shows the
                cumulative share of the initial MSRP of a vehicle assumed to be paid in
                collision and comprehensive insurance in five-year age increments under
                this depreciation assumption, conditional on a vehicle surviving to
                that age--that is, the expected insurance payments at the time of
                purchase will be weighted by the probability of surviving to that age.
                If a vehicle lives to 10 years, 9.9% of the initial MSRP is expected to
                be paid in collision and comprehensive payments; by 20 years 11.9% of
                the initial MSRP; finally, if a vehicle lives to age 40, 12.4% of the
                initial MSRP.
                ---------------------------------------------------------------------------
                 \1840\ Fitch Ratings Vehicle Depreciation Report February 2017,
                Black Book, http://www.blackbook.com/wp-content/uploads/2017/02/Final-February-Fitch-Report.pdf (last visited June 22, 2018).
                ---------------------------------------------------------------------------
                [[Page 24709]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.390
                 The increase in insurance premiums resulting from an increase in
                the average value of a vehicle is a result of an increase in the
                expected amount insurance companies will have to pay out in the case of
                damage occurring to the driver's vehicle. In this way, it is a cost to
                the private consumer, attributable to the CAFE standard that caused the
                price increase.
                (10) Measuring Fuel Consumption
                 The procedure the agencies use to estimate fuel consumption assumes
                that all vehicle models of the same body type--cars, SUVs and vans, and
                light trucks--and age are driven identical amounts each year. Under
                this assumption, the agencies' estimates of fuel consumption from
                increasing the fuel economy of each individual model depend only on how
                much its fuel economy is increased, and do not reflect whether its
                actual use differs from other models of the same body type. Neither do
                the agencies' estimates of fuel consumption account for variation in
                how much vehicles of the same body type and age are driven each year,
                which appears to be significant.
                 This assumption may cause the agencies' estimates of fuel
                consumption from imposing stricter CAFE and CO2 standards to
                be too large. Because the distribution of annual driving is wide using
                its mean value to estimate fuel savings for individual car or light
                truck models may overstate the fuel consumption likely to result from
                tighter standards, even when the fuel economy of different models are
                correctly averaged.\1841\ This will be the case even when increases in
                fuel economy can be estimated reliably for individual models, as the
                agencies' analysis does, because the reduction in a specific model's
                fuel consumption depends on how much it is actually driven as well as
                on the increase that stricter standards require.
                ---------------------------------------------------------------------------
                 \1841\ The correct average fuel economy of vehicles whose
                individual fuel economy differs is the harmonic average of their
                individual values, weighted by their respective use; for two
                vehicles with fuel economy levels MPG1 and
                MPG2 that are assumed to be driven identical amounts (as
                in the agencies' analysis), their harmonic average fuel economy is
                equal to 2/(1/MPG1 + 1/MPG2).
                ---------------------------------------------------------------------------
                 To illustrate, the agencies estimate that new automobiles are
                driven about 17,000 miles on average during their first year. If the
                17,000 mile figure represents the average of two different models that
                are driven 14,000 and 20,000 miles annually, and the two initially
                achieve, respectively, 30 and 40 miles per gallon--thus averaging 35
                miles per gallon--they will consume a total of 967 gallons
                annually.\1842\ Improving the fuel economy of each model by 5 miles per
                gallon will reduce their total fuel use to 844 gallons, thus saving 123
                gallons annually.\1843\ In contrast, the agencies' would estimate total
                fuel consumption for the two vehicles using the 17,000 mile average
                figure for both, thus yielding estimated fuel savings of 128 gallons
                per year, about 5% above the correct value.\1844\
                ---------------------------------------------------------------------------
                 \1842\ Calculated as 14,000 miles/30 miles per gallon + 20,000
                miles/40 miles per gallon = 467 gallons + 500 gallons = 967 gallons
                (all figures in this calculation are rounded to whole gallons).
                 \1843\ Calculated as 14,000 miles/35 miles per gallon + 20,000
                miles/45 miles per gallon = 400 gallons + 444 gallons = 844 gallons
                (again, all figures in this calculation are rounded to whole
                gallons).
                 \1844\ The agencies estimate of their combined initial fuel
                consumption would be 17,000 miles/30 miles per gallon + 17,000
                miles/40 miles per gallon, or 567 gallons + 425 gallons = 992
                gallons. After the 5 mile per gallon improvement in fuel economy for
                each vehicle, the agencies' estimate would decline to 17,000 miles/
                35 miles per gallon + 17,000 miles/45 miles per gallon = 486 + 378 =
                863 gallons, yielding an estimated fuel savings of 992 gallons--863
                gallons = 128 gallons (as previously, all figures in this
                calculation are rounded to whole gallons).
                ---------------------------------------------------------------------------
                 The magnitude of this potential overestimation of fuel savings
                increases with any association between annual driving and fuel economy,
                which seems likely to be strong. Acting in their own economic interest,
                car and light truck buyers who anticipate driving more should be more
                likely choose models offering higher fuel economy, because the number
                of miles driven directly affects their fuel costs and thus the savings
                from driving a model that features higher fuel economy.\1845\
                Conversely, buyers who anticipate driving less are likely to purchase
                models with lower fuel economy. Such behavior--whereby buyers who
                expect to drive more extensively are likely to select models offering
                higher fuel economy--cannot be fully accounted for in today's analysis,
                because that analysis is necessarily based on
                [[Page 24710]]
                empirical estimates of average vehicle use. To the extent it occurs,
                the agencies are likely to consistently overstate actual fuel savings
                from requiring higher fuel economy, as well as to overstate increases
                in fuel consumption resulting from lower standards. Thus, the agencies'
                central analysis is likely to overestimate the final rule's impact on
                consumer benefits such as reduced fuel consumption and increased
                refueling time, as well as on the resulting environmental impacts of
                fuel production and use.
                ---------------------------------------------------------------------------
                 \1845\ For example, some businesses, rental car firms, taxi
                operators, and ride sharing drivers are likely to anticipate using
                their vehicles significantly more than the average new car or light
                truck buyer. Furthermore, their choices among competing models are
                likely to be more heavily influenced by economics than by the
                preferences for other attributes that motivate many other buyers,
                making them more likely to select vehicles with higher fuel economy
                in order to improve their economic returns.
                ---------------------------------------------------------------------------
                 A similar phenomenon may cause the agencies to overstate the value
                of fuel savings resulting from requiring higher fuel economy as well.
                As with miles driven, the agencies' analysis assumes all vehicle owners
                pay the national average fuel price at any time. However, fuel prices
                vary substantially among different regions of the U.S., and one would
                expect buyers in regions with consistently higher fuel prices to
                purchase vehicles with higher fuel economy, on average. To the extent
                they actually do so, evaluating the savings from requiring higher fuel
                economy identically in all regions using nationwide average fuel prices
                is likely to overstate their actual dollar value; similarly, assessing
                the increased fuel costs likely to result from lower standards using
                national average fuel prices is likely to overstate their true value
                insofar as car and light truck buyers facing above-average fuel prices
                choose higher-mpg models.
                 As an illustration, suppose gasoline averages $3.00 per gallon
                nationwide, but a buyer who expects to drive a new car 17,000 miles
                during its first year (the same value used in the example above) faces
                a local price of $4.00 per gallon, and chooses a model that achieves 40
                mpg. That driver's cost of fuel during the vehicle's first year will
                total $1,700 (calculated at 17,000 miles/40 miles per gallon x $4.00
                per gallon). A buyer who plans to drive the same number of miles but
                faces a lower price of $2.00 per gallon and thus chooses a vehicle that
                offers only 30 mpg will have first-year fuel costs of $1,133
                (calculated as 17,000 miles/30 miles per gallon x $2.00 per gallon), so
                total annual fuel costs for these two vehicles will be $1,700 + $1,133
                = $2,633. If the fuel economy of both vehicles increases by 5 mpg,
                their actual fuel savings will be $189 and $162, or a total savings of
                $351. However, evaluating total fuel savings using the national average
                price of $3.00 per gallon yields savings of $382, thus overstating
                actual savings by about 10%. This same phenomenon would cause the
                agencies to overestimate of costs of increased fuel use when standards
                are relaxed, as with this rule.
                (11) Refueling Benefit
                 Increasing CAFE/CO2 standards, all else being equal,
                affect the amount of time drivers spend refueling their vehicles in
                several ways. First, they increase the fuel economy of ICE vehicles
                produced in the future and, consequentially, decrease the number of
                refueling events for those vehicles. Second, given increased production
                costs, they reduce sales of new vehicles and scrappage of existing
                ones, causing more VMT to be driven by older and less efficient
                vehicles which require more refueling events for the same amount of VMT
                driven. Finally, they may change the number of electric vehicles that
                are produced, and shift refueling to occur at a charging station,
                rather than at the pump--changing per-vehicle lifetime expected
                refueling costs. While there are multiple ways that fuel economy
                standards alter refueling costs, the proposal accounted for only the
                first. Before the inclusion of the sales and scrappage models, which
                first appeared in the NPRM analysis for the first time a CAFE/
                CO2 rulemaking, the agencies did not have the means to
                capture the other two effects. While the agencies modeled sales and
                scrappage effects, they did not extend the results to refueling time.
                This oversight was noted by commenters, and the final rule model now
                includes these additional factors. The basic calculation for all three
                effects is the same: The agencies multiply the additional amount of
                time spent refueling by the value of time of passengers, which is
                assumed to be the same for all three effects.
                (a) Value of Time
                 The calculation of the value of time remains relatively unchanged
                from the proposal and follows the guidance from DOT's 2016 Value of
                Travel Time Savings memorandum (``VTTS Memo'').\1846\ The economic
                value of refueling time savings is calculated by applying valuations
                for travel time savings from the VTTS Memo to estimates of how much
                time is saved across alternatives.\1847\
                ---------------------------------------------------------------------------
                 \1846\ United States Department of Transportation, The Value of
                Travel Time Savings: Departmental Guidance for Conducting Economic
                Evaluations, (2016), available at https://www.transportation.gov/sites/dot.gov/files/docs/2016%20Revised%20V.
                 \1847\ VTTS Memo Tables 1, 3, and 4.
                ---------------------------------------------------------------------------
                 IPI commented that the agencies used old data to calculate the
                refueling benefit in the proposal. Specifically, IPI pointed out that
                the data used in the proposal seemed ``to come from the 2003 version of
                [the VTTS Memo].'' \1848\ For the final rule, the analysis uses the
                most recent VTTS memo along with updated wages. The value of travel
                time depends on average hourly valuations of personal and business
                time, which are functions of annual household income and total hourly
                compensation costs to employers. As designated by the 2016 VTTS memo,
                the nationwide median annual household income, $56,516 in 2015, is
                divided by 2,080 hours to yield an income of $27.20 per hour. Total
                hourly compensation cost to employers, inclusive of benefits, in 2015$
                is $25.40.\1849\ Table VI-190 demonstrates the agency's approach to
                estimating the value of travel time ($/hour) for both urban and rural
                (intercity) driving. This approach relies on the use of DOT-recommended
                weights that assign a lesser valuation to personal travel time than to
                business travel time, as well as weights that adjust for the
                distribution between personal and business travel.\1850\ In accordance
                with DOT guidance, wage valuations are estimated with base year 2015
                dollars and end results are adjusted to 2018 dollars.
                ---------------------------------------------------------------------------
                 \1848\ IPI, Appendix, NHTSA-2018-0067-12213, at 51.
                 \1849\ Ibid at11.
                 \1850\ Business travel is higher than personal travel because an
                employer has additional expenses, e.g. taxes and benefits costs,
                above and beyond an employee's hourly wage. In the proposal, the
                agencies erroneously used the same value for personal and business
                travel, which was inconsistent with the VTTS Memo.
                ---------------------------------------------------------------------------
                [[Page 24711]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.391
                 Estimates of the hourly value of urban and rural travel time
                ($14.14 and $20.40, respectively) shown in Table VI-190, must be
                adjusted to account for the nationwide ratio of urban to rural
                driving.\1851\ This adjustment, which gives an overall estimate of the
                hourly value of travel time--independent of urban or rural status--is
                shown in Table VI-191.
                ---------------------------------------------------------------------------
                 \1851\ Estimate of Urban vs. Rural travel weights from FHWA
                December 2018 Traffic Volume Trends, Monthly Report, Table 2--
                Cumulative Monthly Vehicle-Miles of Travel in Billions. Available at
                https://www.fhwa.dot.gov/policyinformation/travel_monitoring/18dectvt/page3.cfm.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.392
                [[Page 24712]]
                 Note that the calculations above consider the value of travel time
                for only one occupant. To estimate fully the average value of vehicle
                travel time per vehicle, the agencies must account for the presence of
                all additional passengers during refueling trips. The agencies
                estimated average vehicle occupancy using survey data gathered as part
                of our 2010-2011 National Automotive Sampling System's Tire Pressure
                Monitoring System (TPMS) study.\1852\ The study was conducted at
                fueling stations nationwide and researchers made observations regarding
                a variety of characteristics of thousands of individual fueling station
                visits from August, 2010 through April, 2011. Among these
                characteristics of fueling station visits, the total number of
                occupants per vehicle were observed. Average vehicle occupancy was
                calculated and multiplied by the value of travel time per occupant. As
                shown in Table VI-192, this adjustment is performed separately for
                passenger cars and for light trucks, yielding occupancy-adjusted
                valuations of vehicle travel time during refueling trips for each
                fleet. Lastly, the occupancy-adjusted value of vehicle travel time is
                converted to 2018 dollars using the GDP deflator as shown in Table VI-
                193.\1853\
                ---------------------------------------------------------------------------
                 \1852\ Docket for Peer Review of NHTSA/NASS Tire Pressure
                Monitoring System, available at https://www.regulations.gov/docket?D=NHTSA-2012-0001.
                 \1853\ Bureau of Economic Analysis, NIPA Table 1.1.9 Implicit
                Price Deflators for Gross Domestic Product, available at https://apps.bea.gov/iTable/index_nipa.cfm.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.393
                 IPI commented that the exclusion of children from the NPRM's
                refueling time analysis was inconsistent with DOT's 2016 Value of
                Travel Time Savings memorandum (``VTTS Memo''). IPI claimed that the
                VTTS Memo ``consider[ed] whether the value of travel time is different
                for parents versus children, but ultimately conclude[d] that `it must
                be assumed that all travelers' VTTS are independent and additive.' ''
                IPI also quoted language from page 13 of the VTTS Memo that
                ``[a]lthough riders may be a family with a joint VTTS or passengers in
                a car pool or transit vehicle with independent values, these
                circumstances can seldom be distinguished [. . .] therefore, all
                individuals are assumed to have independent values,'' and that it is
                ``inappropriate to use different income levels or sources for different
                categories of traveler.'' \1854\
                ---------------------------------------------------------------------------
                 \1854\ See IPI, Appendix, NHTSA-2018-0067-12213, at 52-53
                (citing United States Department of Transportation (``DOT''), The
                Value of Travel Time Savings: Departmental Guidance for Conducting
                Economic Evaluations, (2016), available at https://www.transportation.gov/sites/dot.gov/files/docs/2016%20Revised%20V).
                ---------------------------------------------------------------------------
                 IPI further asserted that excluding passengers under age 16 from
                the calculation of travel time savings was inconsistent with the best
                practices of benefit-cost analysis. IPI noted that Circular A-4 does
                not distinguish between children and adults except when monetizing
                health effects. IPI then cited Dale Whittington and Duncan MacRae as
                stating ``there is a clear consensus that children should be counted in
                cost-benefit analysis.'' Finally, IPI commented that Congress intended
                that the agencies consider the economic impact to children when setting
                standards.\1855\
                ---------------------------------------------------------------------------
                 \1855\ See IPI, Appendix, NHTSA-2018-0067-12213, at 53-54
                (internal citations omitted).
                ---------------------------------------------------------------------------
                 The agencies point out that the first passage from the VTTS Memo
                cited by IPI does not conclude, or even deliberate, that the VTTS of
                children is the same as adults, but instead states that the VTTS of
                children, parents and other passengers should be independent and
                additive.\1856\ Assuming that the opportunity cost of children's time
                is zero is compatible with this practice. Likewise, IPI concluded from
                the text on page 12 that it was inappropriate to use different incomes
                for children. However, IPI's analysis suffers from two errors.
                ---------------------------------------------------------------------------
                 \1856\ See VTTS Memo at 5.
                ---------------------------------------------------------------------------
                 First, the two quotes from page 12 reside in a section of the VTTS
                Memo
                [[Page 24713]]
                entitled Special Issues, which provides guidance on three distinct
                topics. The first quoted text comes from a paragraph advising how to
                treat vehicles with multiple passengers, while the second is from an
                ensuing topic about passenger incomes. It is baseless to assume that
                the conclusion of the second topic holds true for the first.
                 Second, assuming IPI intended to comment that age is a ``category
                of traveler'' for which ``it is inappropriate to use different income
                levels,'' the agencies note that such an interpretation is tenuous. The
                VTTS Memo clearly recognizes that some categories of travelers should
                have different levels of income,\1857\ and provides two examples.\1858\
                As children are not part of the workforce, they do not have wage
                incomes. Therefore, it is not wild speculation that they do not bear a
                financial opportunity cost associated with their time spent in vehicles
                during refueling.\1859\ As such, excluding children from the
                calculation of the refueling benefit is consistent with DOT's guidance.
                ---------------------------------------------------------------------------
                 \1857\ The full text quoted by IPI reads, ``[e]xcept for
                specific distinctions, we consider it inappropriate to use different
                income levels or sources for different categories of traveler.''
                VTTS Memo at 12 (emphasis added). The VTTS Memo further contemplates
                that it is appropriate to assign different incomes if ``estimates
                [of income are] derived by reliable and focused research [. . .] in
                specific cases.'' Id.
                 \1858\ The VTTS Memo provides specific guidance on how to
                differentiate between personal and business travel, and air or high
                speed rail from other modes of transportation. See VTTS Memo at 12.
                 \1859\ The TMPS study affords the agencies the opportunity to
                distinguish between adults and passengers, a luxury not available in
                every instance. Furthermore, there may be certain instances where it
                is appropriate to value the VTTS of children the same as adults,
                e.g., rules focusing primarily on the VTTS of children.
                ---------------------------------------------------------------------------
                 Turning to IPI's comments on best practices and Congress' intent,
                the agencies agree that the benefit-cost analysis should include
                children when appropriate. The majority of the components of the CAFE
                model (e.g., safety analyses) include children. However, children are
                excluded from the analysis when it is appropriate (e.g., employment).
                For this specific valuation, it is reasonable to assume the value of a
                child's time is not equivalent to an adult's. Nonetheless, the agencies
                have examined the impact of valuing children's time as equal to adults'
                by including them in the average vehicle occupancy rates applied in the
                refueling analysis and using the full VTTS for personal travel. Results
                indicate that the effect of this issue is minor and impacts total
                benefits by about one-quarter percent. The agencies will continue to
                consider this issue in future CAFE and CO2 rulemakings. IPI
                also noted that the only portion of the TPMS publicly available was the
                ``User's Coding Manual.'' Specifically, IPI argued that ``the agencies'
                failure to make available the full data and methodology used to
                calculate these average occupancy figures frustrates any meaningful
                public review.'' The agencies disagree. IPI was able to submit a
                meaningful comment about the agencies' decision to exclude children
                from the occupancy-adjusted value of vehicle travel time. Furthermore,
                commenters knew that the agencies intended to use occupancy estimates
                to calculate the refueling benefit; however, the agencies did not
                receive any alternative estimates or methodologies from commenters.
                Nonetheless, the agencies have provided reference to the docket folder
                containing peer review documents, analysis documentation, and data for
                the 2011 TPMS survey.
                (b) Accounting for Improved Fuel Economy of ICE Vehicles
                 The methodology for calculating the refueling benefits associated
                with improved fuel economy in new vehicles remains unchanged from the
                proposal. The CAFE model calculates the number of refueling events for
                each ICE vehicle in a calendar year. This is calculated as the number
                of miles driven by each vehicle in that calendar year divided by the
                product of that vehicle's on road fuel economy, tank size, and an
                assumption about the average share of the tank refueled at each event,
                as follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.394
                 The model then computes the cost of refueling as the product of the
                number of refueling events, total time of each event and value of the
                time spent on each event (computed as average salary), as below:
                 The event time of a vehicle is calculated by summing a fixed and
                variable component. The fixed component is the number of minutes it is
                assumed each event takes, independent of any assumptions about tank
                size or share refueled at each event (the time it takes to get to and
                from the pump). The variable component is the ratio of the average
                number of gallons refueled for each event (the product of the tank size
                and share refueled) and the rate at which gallons flow from the pump.
                This is shown below:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.600
                 1In order to calculate the refueling time cost, as described above,
                the CAFE model takes the following inputs: The value of time, the fixed
                time component of each refueling event, share of the tank refueled at
                each event, rate of flow of fuel from the pump, and vehicle tank size.
                The first of these is taken from DOT guidance on travel time savings.
                The fixed time component, share refueled, and rate of flow are
                calculated from survey data gathered as part of our 2010-2011 National
                Automotive Sampling System's Tire Pressure Monitoring System (TPMS)
                study.\1860\ Finally, the vehicle fuel tank sizes are taken from
                manufacturer specs for the reference fleet and historical averages are
                calculated from popular models for the existing vehicle fleet, as
                described, below, in discussion of the legacy fleet.
                ---------------------------------------------------------------------------
                 \1860\ Docket for Peer Review of NHTSA/NASS Tire Pressure
                Monitoring System, available at https://www.regulations.gov/docket?D=NHTSA-2012-0001.
                ---------------------------------------------------------------------------
                 The agencies estimated the amount of saved refueling time using
                survey data gathered as part of the aforementioned TPMS study. In this
                nationwide study, researchers gathered information on the total amount
                of time spent pumping and paying for fuel. From a separate sample (also
                part of the TPMS study),
                [[Page 24714]]
                researchers conducted interviews at the pump to gauge the distances
                that drivers travel in transit to and from fueling stations, how long
                that transit takes, and how many gallons of fuel are purchased.
                 The agencies focused on the interview-based responses in which
                respondents indicated the primary reason for the refueling trip was due
                to a low reading on the gas gauge. Such drivers experience a cost due
                to added mileage driven to detour to a filling station, as well as
                added time to refuel and complete the transaction at the filling
                station. The agencies believe that drivers who refuel on a regular
                schedule or incidental to stops they make primarily for other reasons
                (e.g., using restrooms or buying snacks) do not experience the cost
                associated with detouring in order to locate a station or paying for
                the transaction, because the frequency of refueling for these reasons
                is unlikely to be affected by fuel economy improvements. This
                restriction was imposed to exclude distortionary effects of those who
                refuel on a fixed (e.g., weekly) schedule and may be unlikely to alter
                refueling patterns as a result of increased driving range. The relevant
                TPMS survey data on average refueling trip characteristics are
                presented below in Table VI-194.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.395
                 The agencies assume that all of the round-trip time necessary to
                travel to and from the fueling station is a part of the fixed time
                component of each refueling event. However, some portion of the time to
                fill and pay is also a part of the fixed time component. Given the
                information in Table VI-194, the agencies assume that each refueling
                event has a fixed time component of 3.5 minutes. E.g., (for passenger
                cars) the sum of 2.28 minutes round trip time to/from fueling station
                and roughly 1.2 minutes to select and pay for fuel, remove/recap fuel
                tank, remove/replace fuel nozzle, etc. The time to fill the fuel tank
                is the variable time component; e.g., about 2.9 minutes for passenger
                cars (2.28 + 1.2 + 2.9 = 6.38 total minutes). However, the CAFE model
                uses a different methodology to determine the variable time component,
                which is explained below.
                 Cars have average tank sizes of about 15 gallons, SUVs/vans of
                about 18 gallons, and pickups of about 27 gallons (see Table VI-195
                through Table VI-197 in discussion of the legacy fleet). It is a
                reasonable assumption that the average passenger car has a tank of 15
                gallons and the average light truck has a tank of 20 gallons (there are
                more SUVs/vans than pickups in the light truck fleet). From these
                assumptions, it is calculated that the average refueling event fills
                approximately 65 percent of the fuel tank for both passenger cars and
                light trucks. This value is used as an input in the CAFE model for all
                three body styles (cars, SUVs/vans, and pickups).
                 Finally, the rate of the pump flow can be calculated either as the
                total gallons pumped over the assumed variable time component
                (approximately 3 minutes) or as the difference in the average number of
                gallons filled between light trucks and passenger cars over the
                difference in the time to fill and pay between the two classes. The
                first methodology implies a rate between 3 and 4 gallons per minute.
                Although the second methodology implies a rate of 15 gallons per
                minute, there is a legal restriction on the flow of gasoline from pumps
                of 10 gallons per minute.\1861\ Thus, the agencies assume the rate of
                gasoline pumps range between 4 and 10 gallons per minute, and use 7.5
                gallons per minute--a value slightly above the midpoint of that range--
                as the average flow rate in the CAFE model.
                ---------------------------------------------------------------------------
                 \1861\ 40 CFR 80.22(j), Regulation of Fuels and Fuel Additives--
                subpart B. Controls and Prohibitions, available at https://www.law.cornell.edu/cfr/text/40/80.22.
                ---------------------------------------------------------------------------
                 The calculations described above are repeated for each future
                calendar year that light-duty vehicles of each model year affected by
                the CAFE standards considered in this rule would remain in service for
                each regulatory alternative. The resulting cumulative lifetime
                valuations of time savings account for both the reduction over time in
                the number of vehicles of a given model year that remain in service and
                the reduction in the number of miles (VMT) driven by those that stay in
                service. After calculating the absolute value for each regulatory
                alternative using the methodology and inputs described above, the model
                calculates the incremental value relative to the baseline as the
                refueling cost or benefit for that regulatory alternative. More
                efficient vehicles have to be refueled less often and refueling costs
                per vehicle decline. In previous rules this was sufficient to account
                for the majority of any changes in cost of refueling under different
                CAFE standards as the modelling permitted, since the volumes of new
                vehicles and existing vehicles on the road was assumed to be constant
                under all possible standards. However, when sales and scrappage models
                are included the distribution of new and vehicles varies and a
                different number of miles will be driven by new and used vehicles in
                each regulatory alternative.
                 IPI commented that it was inappropriate for the agencies to
                [[Page 24715]]
                exclude benefits from reducing the frequency of refueling events where
                the primary reason for stopping at a fuel station was not to refuel a
                vehicle. IPI argued that fuel efficiency impacts from relaxed standards
                would affect all drivers regardless of their rationale for refueling,
                by requiring either more frequent or marginally longer refueling
                events.\1862\ The agencies note that the language in the NPRM suggested
                that the agencies eliminated 40 percent of the potential benefit from
                fewer refueling stops--where 40 percent represents the fraction of
                refueling stops that were routinely scheduled or otherwise not made in
                response to a low fuel reading--and this appears to have been the
                origin of IPI's concern.\1863\ In fact, the agencies did not apply a 40
                percent discount factor to the refueling benefits; instead, the total
                number of additional refueling events that would result from
                alternative CAFE levels was calculated, and these were valued based on
                an assumption that their characteristics (e.g., vehicle occupancy)
                would match those of drivers who refueled due to a low fuel reading.
                ---------------------------------------------------------------------------
                 \1862\ IPI, Appendix, NHTSA-2018-0067-12213, at 54-55.
                 \1863\ See 83 FR 43088 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 To the extent that lower fuel economy affects those who refuel on a
                routine schedule or incidental to stops made primarily for other
                reasons, the per-event cost would actually be limited to the extra time
                spent pumping a slightly larger volume of fuel. However, the agencies
                note that by assuming that all extra fuel consumed under lower CAFE
                standards results in added refueling trips, the agencies are adopting a
                conservative assumption, in the sense that it maximizes the disbenefits
                of alternatives to the current standards.
                 IPI also expressed concern that the agencies may have excluded the
                fuel costs and added emissions from additional miles driven in the
                course of the more frequent refueling events that would be required
                with more lenient CAFE standards, and correspondingly lower on-road
                fuel economy.\1864\ In the NPRM, the agencies asserted that these added
                costs are reflected in their overall estimates of fuel cost savings,
                while any increase in emissions is also reflected in the reported
                changes in total emissions. However, IPI noted that the agencies did
                not clearly explain how these cost savings and emissions reductions are
                actually accounted for in their methodology.
                ---------------------------------------------------------------------------
                 \1864\ IPI, Appendix, NHTSA-2018-0067-12213, at 55.
                ---------------------------------------------------------------------------
                 The agencies' methodology fully accounts for both of these impacts
                through its calculation of changes in the use of new cars and light
                trucks due to the fuel economy rebound effect, which captures the
                impact on their aggregate use (VMT) that results from changes in the
                fuel cost of driving each mile. Studies that estimate the rebound
                effect analyze the relationship between VMT per time period and fuel
                economy or per-mile fuel costs, using data for individual vehicles,
                fleet-wide average values, or aggregate estimates for an entire fleet.
                Regardless of the level of aggregation they employ, their measures of
                vehicle use invariably include travel for all purposes, including any
                extra miles driven in the course of refueling.
                 Thus, the estimates of the rebound effect--the response of vehicle
                use to changes in fuel economy or per-mile fuel costs--inevitably
                capture any change in the number of miles driven for the purpose of
                refueling that occurs in response to higher or lower fuel economy. This
                change reflects the net effect of more or less frequent refueling trips
                required by their baseline or ``pre-rebound'' level of use, and any
                change in the number of refueling trips associated with increased or
                reduced driving in response to the rebound effect.
                 As a consequence, the agencies' estimates of changes in aggregate
                fuel consumption and fuel costs incorporate--that is, are net of--the
                volume and cost of fuel consumed by changes in vehicle use that result
                from the rebound effect, including any change in driving associated
                with more or less frequent refueling. Similarly, the agencies'
                estimates of changes in emissions resulting from vehicle storage and
                use (referred to as ``tailpipe'' or ``downstream'' emissions) are
                derived by applying per-mile emission factors to changes in aggregate
                vehicle travel, so they necessarily incorporate changes in vehicle use
                for all purposes, including more or less frequent refueling.
                 Furthermore, as the agencies demonstrated in the proposal with a
                practical example, the benefit associated with fewer miles spent
                refueling is less than 23[cent] per year for new vehicles. The
                cumulative impact of this benefit amounts to less than one tenth of
                percent of the costs of the rule.\1865\
                ---------------------------------------------------------------------------
                 \1865\ See 83 FR at 43088. Also, note that the 23 cents estimate
                was derived for a less stringent alternative than today's standards
                and included taxes which would have been removed had the agencies
                calculated this number separately.
                ---------------------------------------------------------------------------
                 Because all of the alternative standards evaluated in this
                rulemaking would permit lower fuel economy levels than under the
                baseline standard, per-mile driving costs would be higher and total
                vehicle use would decline in response. Although some (perhaps most) new
                vehicles would require more frequent refueling, the agencies' estimates
                of the change in aggregate use of new vehicles reflects (i.e., is net
                of) any increase in driving associated with more frequent refueling
                stops. As a result, the agencies' estimates of changes in total fuel
                consumption, aggregate fuel costs, and emissions resulting from the
                lower fuel economy levels that relaxing CAFE standards would permit
                reflect the net reduction in use of new cars and light trucks due to
                the fuel economy rebound effect, after considering any additional miles
                that would be driven in the course of more frequent refueling stops.
                (c) Including the Legacy Fleet
                 Under more stringent regulatory alternatives, more miles will be
                driven by older and less efficient vehicles, and the effect is to
                reduce or eliminate any refueling benefit from increasing the fuel
                efficiency of new vehicles. Failing to include the existing fleet makes
                the costs of refueling artificially lower under more stringent
                standards because new vehicle sales are lower and not only because new
                vehicles are more efficient. This update to the calculation of the
                absolute refueling costs corrects this oversight present in the NPRM
                cost-benefit analysis by calculating fleet-wide absolute refueling
                costs before considering the incremental change relative to the
                baseline.
                 For other portions of the CAFE model, the agencies track the legacy
                vehicles by body style and vintage, using average measures for fuel
                economy, horsepower and curb weight. To estimate refueling costs for
                these vehicles, measures of average fuel tank sizes by body style and
                vintage are needed. The agencies are unaware of any data that directly
                estimates this value, but an estimate can be derived from publicly
                available data on fuel tank sizes of 17 high-volume nameplates with
                long histories. The tank sizes are averaged by body style, and these
                historical values are used as estimates of the average by body style
                and vintage. The vehicles included, their fuel tank sizes, and the
                averages are reported in Table VI-195 through Table VI-197 for cars,
                vans/SUVs, and pickups, respectively. The averages are used to
                represent the fuel tank sizes by vintage and vehicle body style. The
                agencies used the fuel tank sizes from Table VI-195 to Table VI-196 to
                determine the number of refueling events and time spent refueling to
                [[Page 24716]]
                compute refueling costs using the methodology described above.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.396
                [[Page 24717]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.397
                [[Page 24718]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.398
                [[Page 24719]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.399
                [[Page 24720]]
                BILLING CODE 4910-59-C
                (d) Including Electric Vehicle Recharging
                 In addition to adding the refueling costs associated with the
                ``legacy fleet,'' this update adds the cost to recharge electric
                vehicles to the total refueling costs. Excluding the time spent
                recharging ignores a real cost borne by owners of electric vehicles,
                one which was noted by multiple commenters. For example, Ariel Corp.
                and VNG.co LLC commented that, ``EVs require significant changes in
                consumer fueling behavior given the need to park at recharging points
                for long periods of time.'' \1866\
                ---------------------------------------------------------------------------
                 \1866\ Ariel Corp. and VNG.co LLC, Comment, NHTSA-2018-0067-
                7573, at 13.
                ---------------------------------------------------------------------------
                 In order to do so, it is important to first understand how many
                electric vehicle charging events will require the driver to wait and
                for how long. The answer to this question depends on the range of the
                electric vehicle and the length of the trip.\1867\ For trips shorter
                than the range, the driver can recharge the vehicle at times that will
                not require them to be actively waiting and thus there is no recharging
                cost. Only for trips where the vehicle is driven more miles than the
                range will the driver have to stop at mid-trip, a time that is assumed
                to be inconvenient, to recharge the vehicle at least enough to reach
                the intended destination.
                ---------------------------------------------------------------------------
                 \1867\ While the range of EVs is dependent on a number of
                factors, such as that grade, acceleration, and weather, the agencies
                take a conservative approach and assume a best-case scenario.
                ---------------------------------------------------------------------------
                 The agencies use trip data from the National Household
                Transportation Survey (NHTS) to estimate the frequency and expected
                length of trips that exceed the range of the electric vehicle
                technologies in the simulation (200 and 300 mile ranges).
                 The NHTS data is collected from a representative random sample of
                U.S. households. The survey collects data on individual trips by mode
                of transportation. A trip is defined by the starting and ending point
                for any personal travel, so that vehicle trips will capture any time a
                car is driven. The survey includes identification numbers for
                households, individuals, and vehicles, and mode of transportation
                (including the body style of the vehicle for vehicle trips), and the
                date of the trip. Although some trips made in the same day may allow
                for convenient charging in between trips, the agencies assume that
                travel in the same day exceeding the range will involve the driver
                waiting for the vehicle to charge. Thus, the total number of miles
                driven by the same vehicle in a single day is summed, and it is assumed
                that charging stations are not conveniently available to the driver in
                between.
                 Some of the trips in the NHTS have missing information about the
                duration or length of the trip; these trips are excluded from the
                dataset. The agencies subset the dataset into three body styles--cars,
                vans/SUVs, and pickups--consistent groupings with how the VMT schedules
                and scrappage rates are estimated. The agencies exclude data on taxis
                and rental cars as the body style of the vehicle for these trips is not
                specified (they make up only 0.3 percent of the dataset, so their
                exclusion is unlikely to alter the estimate). Table VI-198, below,
                shows the resulting quantiles of the distribution of daily travel for
                all vehicles considered in the final dataset. This will include
                multiple days of travel for the same vehicle if more than one day of
                trip data is recorded in the NHTS.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.400
                [GRAPHIC] [TIFF OMITTED] TR30AP20.401
                 The data in Table VI-198 shows that excluding taxis and rentals may
                be the best choice even if their body styles were known. For taxi
                trips, only the number of trips an individual driver makes in a day is
                known. The number of trips that the taxi cab itself makes in a day is
                unknown. As can be seen, the distribution of ``daily'' travel is to the
                left for taxis because not all trips for those vehicles are reported.
                Thus, including these vehicles would incorrectly skew the daily travel
                rates downwards.
                 The distribution of trip lengths for rental cars, on the other
                hand, is generally to the right of trips taken privately-owned
                vehicles. This is likely because individuals are travelling longer
                distances when they are on vacation or otherwise out-of-town. It seems
                likely that individuals renting cars for longer trips will not choose
                electric vehicles for such temporary travel. Thus,
                [[Page 24721]]
                including these trips in the dataset would likely overestimate the
                number of mid-trip charging events necessary for ordinary travel in a
                way that will not match what actually occurs.
                 From the final body style datasets, the agencies are able to
                calculate two measures that allow for the construction of the value of
                recharging time. First, the expected distance between trips that exceed
                the range of 200-mile and 300-mile BEVs (BEV200 and BEV300,
                respectively) is calculated. This is calculated as the quotient of the
                sum of total miles driven by each individual body style and the total
                number of trips exceeding the range, as shown below:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.402
                 This calculates the expected frequency of enroute recharging
                events, or the amount\1868\ of miles traveled per inconvenient
                recharging event. This is used later used to calculate the total
                expected time to recharge a vehicle.
                ---------------------------------------------------------------------------
                 \1868\ The denominator counts the number of incontinent
                recharging events by body style. It is not a measurment of VMT.
                ---------------------------------------------------------------------------
                 The second measure needed to calculate the total expected
                recharging time is the expected share of miles driven that will be
                charged in the middle of a trip (causing the driver to wait and lose
                the value of time). In order to calculate this measure the difference
                of the trip length and range is summed, conditional on the trip length
                exceeding the range for each body style. This figure is then divided by
                the sum of the length of all trips for that body style. See the
                equation below:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.403
                 The calculated frequency of inconvenient charging events and share
                of miles driven that require the driver to wait for BEV's with 200 and
                300-mile ranges are presented in Table VI-199, below. As the table
                shows, cars are expected to require less frequent inconvenient charges
                and a smaller share of miles driven will require the driver to charge
                the vehicle in the middle of a trip. Pickups and vans/SUVs have fairly
                similar measures, with vans and SUVs requiring slightly more
                inconvenient charging than pickups.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.404
                 The measures presented in Table VI-199, above, can be used to
                calculate the expected time drivers of electric vehicles of a given
                body style and range will spend recharging at a time that will require
                them to wait. First the agencies calculate the expected number of
                refueling events for a vehicle of a given style and range in a given
                calendar year. This is shown below as the expected miles driven by a
                vehicle in a given calendar year divided by the charge frequency of a
                vehicle of that style and range (from Table VI-199).
                [[Page 24722]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.405
                 Next the agencies calculate the number of miles charged for a
                vehicle of a given style\1869\ and range in a specific calendar year.
                This is the product of the number of miles driven by the vehicle and
                the share of miles driven that require an inconvenient charge for a
                vehicle of that style and range (from Table VI-199), as presented
                below:
                ---------------------------------------------------------------------------
                 \1869\ Note that [Sigma]Trip [epsi] Style Trip Length and Miles
                CY,Veh are different values. MilesCY,Veh is the estimated amount of
                VMT predicted by VMT while [Sigma]Trip [epsi] Style Trip Length is
                the sum of trips observed by the NHTS study.
                ---------------------------------------------------------------------------
                 Then, the expected time that a driver of an electric vehicle of a
                given style and range will spend waiting for the vehicle to charge is
                calculated. This is the product of the fixed amount of time it takes to
                get to the charging station and the number of recharging events plus
                the quotient of the expected miles that will require inconvenient
                charging over an input assumption of the rate of which a vehicle of
                that style and range can be charged in a given calendar year (expressed
                in units of miles charged per hour). The fixed amount of time it takes
                to get to a charging station is set equal to the average time it takes
                for an ICE vehicle to get to a gas station for a refueling event, as
                discussed above.\1870\ This is shown below:
                ---------------------------------------------------------------------------
                 \1870\ The agencies note that this is a conservative estimate.
                Gas stations vastly outnumber publicly available recharging stations
                and are often in more convenient locations.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.406
                 The expected time that a driver will wait for their vehicle to
                charge can then be multiplied by the value of time estimate, as is done
                with gasoline, diesel, and E85 vehicles (see description above of the
                current approach to accounting for refueling time costs).
                 It is worth a final note to talk about how plug-in hybrids are
                treated in the modelling (which remains unchanged from the NPRM).
                Presumably, plug-in hybrids that are taken on a trip that exceeds their
                electric range will be driven on gasoline and the driver will recharge
                the battery at a time that is convenient. For this reason, the electric
                portion of travel should be excluded from the refueling time
                calculation. The gasoline portion of travel is treated the same as
                other gasoline vehicles so that when the tank reaches some threshold,
                the vehicles is assumed to be refueled with the same fixed event time
                and the same rate of refueling flow.
                 The NPRM calculation of refueling benefits did not account for the
                impacts of fleet turnover--specifically the impact on ``legacy'' fleet
                vehicles and new electric vehicles. However, when the quantities of
                vehicles on the road varies between scenarios it becomes important to
                calculate the refueling costs for all vehicles since fuel economy and
                tank sizes (and therefore range before refueling) vary with vintage.
                This updated analysis adds these elements to the calculation of the
                refueling time and costs and is thus a more accurate estimation of the
                refueling benefit.
                (12) Energy Security
                 By amending existing standards, the final rule is expected to
                increase domestic consumption of gasoline by a relatively minimal
                amount relative to the baseline standards finalized in 2012, producing
                a correspondingly small increase in the Nation's demand for crude
                petroleum, a commodity that is traded actively in a worldwide market.
                Specifically, the agencies project that this rule will increase
                gasoline consumption by cars and light trucks produced during model
                years 1978 through 2029 by 3.1 percent.\1871\ Although the U.S.
                accounts for a sufficient (albeit diminishing) share of global oil
                consumption that the resulting increase in global petroleum demand will
                exert some upward pressure on worldwide prices, the rule is projected
                to increase global petroleum demand by less than one half of one
                percent from 2017 through 2050, so its effects on global prices is
                likely to be minimal.
                ---------------------------------------------------------------------------
                 \1871\ This includes fuel consumed by cars and light trucks
                produced during model years 1978-2017 that are on the road today
                during their remaining lifetimes, as well as fuel consumed by cars
                and light trucks projected to be manufactured during model years
                2018-2029 over their entire lifetimes.
                ---------------------------------------------------------------------------
                 U.S. consumption and imports of petroleum products has three
                potential effects on the domestic economy that are often referred to
                collectively as ``energy security externalities,'' and increases in
                their magnitude are sometimes cited as possible social costs of
                increased U.S. demand for petroleum.m First, any increase in global
                petroleum prices that results from higher U.S. gasoline demand will
                cause a transfer of revenue to oil producers worldwide from consumers
                of petroleum, because consumers throughout the world are ultimately
                subject to the higher global price that results. Although this transfer
                is simply a shift of resources that produces no change in global
                economic welfare, the financial drain it produces on the U.S. economy
                is sometimes cited as an external cost of increased U.S. petroleum
                consumption, because consumers of petroleum products are unlikely to
                consider it.
                 As the U.S. approaches self-sufficiency in petroleum production
                (the nation is expected to become a net exporter of petroleum by 2020),
                this transfer is increasingly from U.S. consumers of refined petroleum
                products to U.S. petroleum producers, so it not only leaves welfare
                unaffected, but even ceases to be a financial burden on the U.S.
                economy.\1872\ In fact, as the U.S. becomes a net petroleum exporter,
                any transfer from global consumers to petroleum producers would become
                a financial benefit to the U.S. economy. Nevertheless, uncertainty in
                the nation's long-term import-export balance makes it difficult to
                project precisely how these effects might change in response to
                increased consumption.
                ---------------------------------------------------------------------------
                 \1872\ The United States became a net exporter of oil on a
                weekly basis several times in late 2019, and EIA's AEO 2019 projects
                that will do so on a sustained, long-term basis by 2020; see EIA,
                AEO 2019 Reference Case, Table 21 https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=pet&s=wttntus2&f=4.
                ---------------------------------------------------------------------------
                 Higher U.S. petroleum consumption can also increase domestic
                consumers' exposure to oil price shocks and thus
                [[Page 24723]]
                increase potential costs to all U.S. petroleum users (including those
                outside the light duty vehicle sector, whose consumption would be
                unaffected by today's final rule) from possible interruptions in the
                global supply of petroleum or rapid increases in global oil prices.
                Because users of petroleum products are unlikely to consider the effect
                of their increased purchases on these risks, their economic value is
                often cited as an external cost of increased U.S. consumption. Finally,
                some analysts argue that domestic demand for imported petroleum may
                also influence U.S. military spending; because the increased cost of
                military activities would not be reflected in the price paid at the gas
                pump, this is often alleged to represent a third category of external
                costs form increased U.S. petroleum consumption.
                 Each of these three costs could rise incrementally--albeit by a
                very limited magnitude--as a consequence of increases in U.S. petroleum
                consumption--likely to result from the final rule. This section
                describes the extent to which each cost is expected to increase as a
                result of this action, whether it represents a significant economic
                cost (or simply a transfer of resources), and how the agencies have
                measured each cost and incorporated it into their analysis.
                (a) U.S. Petroleum Demand and Its Effect on Global Prices
                 Figure VI-79 illustrates the effect of the increase in U.S. fuel
                and petroleum demand anticipated to result from reducing CAFE and
                CO2 standards on global demand for petroleum and its market
                price. The marginal increase in domestic demand can be represented as
                an outward shift in the U.S. demand curve for petroleum from its
                position at DUS,0 with the baseline standards for future
                model years in effect, to DUS,1 with the final rule
                standards replacing them. Because global demand is simply the sum of
                what each nation would purchase at different prices, the outward shift
                in U.S. demand causes an identical shift in the global demand schedule,
                as the figure shows.\1873\
                ---------------------------------------------------------------------------
                 \1873\ The figure exaggerates the U.S. share of total global
                consumption, which currently stands at 20 percent, for purposes of
                illustration.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.407
                 The global supply curve for petroleum slopes upward, reflecting the
                fact that it is progressively costlier for oil-producing nations to
                explore for, extract, and deliver additional supplies of oil to the
                world market.\1874\ Thus the upward shifts in the U.S. and world demand
                schedules cause an increase in the global price for oil, from
                P0 to P1 in the figure. U.S. purchases of
                petroleum increase from QUS,0 to QUS,1, but the
                resulting increase in global consumption from QG,0 to
                QG,1 will be slightly smaller than the increase in U.S.
                demand and purchases, because the amount of petroleum other nations
                purchase will decline slightly in response to its higher price.
                Spending on petroleum by U.S. buyers who purchase the additional oil
                will increase by the area QUS,0acQUS,1, the
                product of its new, higher price P1 and the increase in U.S.
                consumption, QUS,1-QUS,0, while spending by U.S.
                consumers whose purchases remain unchanged will increase by the product
                of their previous purchases QUS,0 and the price increase
                P1-P0, or the area P1abP0.
                ---------------------------------------------------------------------------
                 \1874\ The figure depicts the relationship between the global
                supply of petroleum and its worldwide price during a single time
                period. The global supply curve for petroleum has been shifting
                outward over time in response to increased investment in
                exploration, the ability of refineries to utilize feedstocks other
                than conventional petroleum, and technological innovations in
                petroleum extraction. The combination of these developments may also
                have reduced its upward slope, meaning that global supply now
                increases by more in response to increases in the world price than
                it once did.
                ---------------------------------------------------------------------------
                 CARB asserted in their comments, that the NPRM analysis was biased
                [[Page 24724]]
                against the baseline standards because the fuel prices in the NPRM were
                based on a unique run of DOE's NEMS model that included the
                baseline.\1875\ They argued that the proposal would have reduced fleet
                average fuel economy, leading to increased demand and subsequently
                higher fuel prices faced by consumers. As a result, the additional fuel
                costs associated with the proposal (relative to the baseline) should
                have been even higher than estimated because the fuel price faced by
                drivers in that scenario would have been higher than in the baseline.
                However, while the difference between the baseline and preferred
                alternative could create differences in fleet fuel economy in a manner
                that could influence prices at the pump, those differences are likely
                to be small. In response to CARB's comments, the agencies conducted
                additional runs with NEMS to compare the fuel price under the baseline
                standards and the fuel price under the proposed standards. Through
                2050, the fuel price difference between the alternatives was never
                higher than two percent. The standards being finalized in this rule are
                considerably closer to the baseline than were those in the proposal.
                ---------------------------------------------------------------------------
                 \1875\ NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 SAFE commented that the United States is a ``price-taker'' in the
                global market and ``must accept the prevailing global oil price since
                it lacks sufficient market power to influence decisively this price.''
                \1876\ This comment, however, is directly at odds with both the
                economics of the world oil market shown in Figure VI-79 above and other
                comments asserting that the increase in U.S. gasoline demand resulting
                from this rule will increase U.S. and global petroleum demand, thus
                increasing world oil prices. In response to the comment from SAFE, the
                agencies utilized a forecast of fuel prices in today's analysis that
                considers the effect of the revised standards on global petroleum
                demand and prices. This assumption slightly increases the cost of
                forgone fuel savings in the preferred alternative, compared to their
                value under the assumption that U.S. demand cannot change global prices
                and the nation acts as a price-taker.
                ---------------------------------------------------------------------------
                 \1876\ NHTSA-2018-0067-11981.
                ---------------------------------------------------------------------------
                 In Figure VI-79, the increase in the price of oil from
                P0 to P1 will mean that global consumers who
                previously purchased the quantity of oil QG,0 at its lower
                price will now pay more for that same amount. Specifically, previous
                purchasers will pay the additional area P1deP0,
                whose value is the increase in price P1-P0
                multiplied by the volume they originally bought, QG,0. Of
                this increase in revenue to oil producers, the rectangular area
                P1abP0--which as indicated above is the product
                of the increase in price P1-P0 and previous U.S.
                purchases QUS,0, and thus measures the increase in spending
                by previous U.S. consumers--is simply transferred from U.S. consumers
                to global oil suppliers.\1877\ The remaining fraction of increased
                payments to producers, the rectangular area adeb, whose value is the
                product of the price increase P1-P0 and previous
                purchases by other nations, which were QG,0-
                QUS,0, is a transfer from consumers outside the U.S. to
                global oil producers.
                ---------------------------------------------------------------------------
                 \1877\ Note that global oil suppliers include domestic as well
                as US-owned foreign suppliers.
                ---------------------------------------------------------------------------
                 The total increase in global spending--including the additional
                spending by U.S. consumers as well as by those in other nations--on the
                amount of oil they previously purchased is simply a transfer of revenue
                from consumers of petroleum products to oil producers. This transfer
                can be described as a ``pecuniary'' externality, since it describes the
                effect of the price increase on wealth allocation, but is considered
                separately from any effects on quantity produced and consumed. Some of
                the increase in payments by U.S. consumers for the petroleum products
                they originally consumed may be made to foreign-owned oil producers,
                and thus represents a financial drain on the U.S. economy, while the
                remainder is received by domestic producers and thus remains within the
                U.S. economy.\1878\
                ---------------------------------------------------------------------------
                 \1878\ Neither transfer, however, has an effect on domestic or
                global economic welfare.
                ---------------------------------------------------------------------------
                 To an increasing extent, however, the additional payments by U.S.
                consumers that result from upward pressure on the world oil price are a
                transfer entirely within the Nation's economy, because a growing
                fraction of domestic petroleum consumption is supplied by U.S.
                producers. The U.S. is projected to become a net exporter of petroleum
                in 2020--and in fact became a net exporter in September 2019--and as
                the Nation moves toward that status, an increasing share of any higher
                costs paid by U.S. consumers of petroleum products becomes a gain to
                U.S. oil producers.\1879\ When the U.S. becomes self-sufficient in
                petroleum supply--which is now anticipated to occur in the year this
                final rule publishes--the entire value of increased payments by U.S.
                petroleum users that results from relaxing CAFE and CO2
                standards will have the same effect as if it were simply a transfer
                within the U.S. economy. As a consequence, the financial burden that
                transfers from U.S. consumers to foreign producers places on the U.S.
                economy will disappear.
                ---------------------------------------------------------------------------
                 \1879\ The U.S. Energy Information Administration EIA estimates
                that the United States exported more total crude oil and petroleum
                products in September and October of 2019, and expects the United
                States to continue to be a net exporter. See Short Term Energy
                Outlook November 2019, available at https://www.eia.gov/outlooks/steo/archives/nov19.pdf.
                ---------------------------------------------------------------------------
                 Over almost the entire time period spanned by the analysis of this
                final rule, any increase in domestic spending for petroleum caused by
                the effect of higher U.S. fuel consumption and petroleum use on world
                oil prices is expected on balance to be a transfer within the U.S.
                economy and thus produce no drain on domestic economic resources. For
                this reason--and because in any case such transfers do not create real
                economic costs or benefits--increased U.S. spending on petroleum
                products that results from increased U.S. fuel demand and any resulting
                upward pressure on petroleum prices stemming from this action is not
                included among the economic costs accounted for in this final rule.
                (b) Macroeconomic Costs of U.S. Petroleum Consumption
                 In addition to influencing global demand and prices, U.S. petroleum
                consumption imposes further costs that are unlikely to be reflected in
                the market price for petroleum, or in the prices paid by consumers of
                refined products such as gasoline.\1880\ Petroleum consumption imposes
                external economic costs by exposing the U.S. economy to increased risks
                of rapid increases in prices triggered by global events that may also
                disrupt the supply of imported oil, and U.S. consumers of petroleum
                products are unlikely to take such costs into account when making their
                decisions about how much to consume.
                ---------------------------------------------------------------------------
                 \1880\ See, e.g., Bohi, D.R. & W. David Montgomery (1982), Oil
                Prices, Energy Security, and Import Policy Washington, DC--Resources
                for the Future, Johns Hopkins University Press; Bohi, D.R., & M.A.
                Toman (1993), ``Energy and Security--Externalities and Policies,''
                Energy Policy 21:1093-1109; and Toman, M.A. (1993). ``The Economics
                of Energy Security--Theory, Evidence, Policy,'' in A. V. Kneese and
                J.L. Sweeney, eds. (1993), Handbook of Natural Resource and Energy
                Economics, Vol. III, Amsterdam--North-Holland, pp. 1167-1218.
                ---------------------------------------------------------------------------
                 Sudden interruptions in oil supply and rapid increases in its price
                can impose significant economic costs, because they raise the costs of
                producing all commodities whose manufacturing and distribution consumes
                petroleum, thus temporarily reducing the level of output that the U.S.
                economy can produce using its available supplies of labor and capital.
                The magnitude of any reduction in
                [[Page 24725]]
                economic output depends on the extent and duration of the increases in
                prices for petroleum products that result from a disruption in global
                oil supplies, as well as on whether and how rapidly prices return to
                their pre-disruption levels--which in turn depends largely on the rest
                of the world's capability to respond to interruptions by increasing
                production elsewhere. Even if prices for oil return completely to their
                original levels, however, economic output will be at least temporarily
                reduced from the level that would have been possible with uninterrupted
                oil supplies and stable prices, so the U.S. economy will bear some
                transient losses it cannot subsequently recover.
                 Supply disruptions and price increases caused by global political
                events tend to occur suddenly and unexpectedly, so they can also force
                businesses and households to adjust their use of petroleum products
                more rapidly than if the same price increase occurred gradually. Rapid
                substitutions between energy derived from oil and other forms of
                energy, as well as between energy and other inputs, and other changes
                such as adjusting production levels and downstream prices, can be
                costly for businesses to make. As with businesses, sudden changes in
                energy prices and use are also difficult for households to adapt to
                quickly or smoothly, and doing so may impose at least temporary costs
                or losses in utility for the various adjustments they make.
                 Interruptions in oil supplies and sudden increases in petroleum
                prices are both uncertain prospects, and the costs of the disruptions
                they can cause must be weighted or adjusted by the probability that
                they will occur, as well as for their uncertain duration. The agencies
                estimate this expected cost of such disruptions by combining the
                probabilities that price increases of different magnitudes and
                durations will occur during the future period spanned by their analysis
                with the costs of reduced U.S. economic output and abrupt adjustments
                to sharply higher petroleum prices. Any change in the probabilistic
                ``expected value'' of such costs that can be traced to higher U.S. fuel
                consumption and petroleum demand stemming from this final rule to
                establish less demanding fuel economy standards is considered to be an
                external cost of the adopting it.
                 A variety of mechanisms exist to ``insure'' against higher
                petroleum prices and reduce their costs for adjusting to sudden price
                increases, including making purchases or sales in oil futures markets,
                adopting energy conservation measures, diversifying the fuel economy
                levels within the set of vehicles owned by the household, locating
                where public transit provides a viable alternative to driving, and
                installing technologies that permit rapid fuel switching. Growing
                reliance on such measures, coupled with continued improvements in
                energy efficiency throughout the economy, has certainly reduced the
                vulnerability of the U.S. economy to the costs of oil shocks in recent
                decades.
                 Thus, there is now considerable debate about the magnitude and
                continued relevance of potential economic damages from sudden increases
                in petroleum prices. The petroleum intensity of the U.S economy has
                declined considerably and global oil prices are dramatically lower than
                when analysts first identified and quantified the risks they create to
                the U.S. economy. Further, not only has the Nation dramatically
                increased its own petroleum supply, but other new global supplies have
                emerged as well, both of which reduce the potential impact of
                disruptions that occur in unstable or vulnerable regions where oil is
                produced.
                 As a consequence, the potential macroeconomic costs of sudden
                increases in oil prices are now likely to be considerably smaller than
                when they were original identified and estimated. Research by the
                National Research Council (2009) argued that non-environmental
                externalities associated with dependence on foreign oil are small, and
                perhaps trivial.\1881\ Research by Nordhaus and by Blanchard and Gali
                have also questioned how harmful to the economy oil price shocks have
                been, noting that the U.S. economy actually expanded immediately after
                the most recent oil price shocks, and that there was little evidence of
                higher energy prices being passed through to higher wages or
                prices.\1882\
                ---------------------------------------------------------------------------
                 \1881\ National Research Council, Hidden Costs of Energy--
                Unpriced Consequences of Energy Production and Use, National Academy
                of Sciences, Washington, DC (2009).
                 \1882\ Nordhaus argues that one reason for limited vulnerability
                to oil price shocks is that monetary policy has become more
                accommodating to the price impacts, while another is that U.S.
                consumers and businesses may determine that such movements are
                temporary and abstain from passing them on as inflationary price
                increases in other parts of the economy. He also notes that changes
                in productivity in response to recent oil price increases are have
                been extremely modest, observing that ``energy-price changes have no
                effect on multifactor productivity and very little effect on labor
                productivity.'' at p. 19. Blanchard and Gali (2010) contend that
                improvements in monetary policy, more flexible labor markets, and
                the declining energy intensity of the U.S. economy (combined with an
                absence of concurrent shocks to the economy from other sources)
                lessened the impact of oil price shocks after 1980. They find that
                ``the effects of oil price shocks have changed over time, with
                steadily smaller effects on prices and wages, as well as on output
                and employment . . . The message . . . is thus optimistic in that it
                suggests a transformation in U.S. institutions has inoculated the
                economy against the responses that we saw in the past.'' at p. 414;
                See William Nordhaus, ``Who's Afraid of a Big Bad Oil Shock?''
                Available at http://aida.econ.yale.edu/~nordhaus/homepage/
                Big_Bad_Oil_Shock_Meeting.pdf; and Blanchard, Olivier and Jordi
                Gali, J., ``The Macroeconomic Effects of Oil price Shocks--Why are
                the 2000s so Different from the 1970s?,'' in Gali, Jordi and Mark
                Gertler, M., eds., The International Dimensions of Monetary Policy,
                University of Chicago Press, February (2010), pp. 373-421, available
                at http://www.nber.org/chapters/c0517.pdf.
                ---------------------------------------------------------------------------
                 Since these studies were issued in 2009 and 2010, the petroleum
                intensity of the U.S. economy has continued to decline while domestic
                energy production has increased in ways and to an extent that experts
                failed to predict, so that the U.S. became the world's largest producer
                in 2018.\1883\ The U.S. shale oil revolution has both established the
                potential for energy independence and placed downward pressure on
                prices. Lower oil prices are also a result of sustained reductions in
                U.S. consumption and global demand resulting from energy efficiency
                measures, many undertaken in response to previously high oil prices.
                ---------------------------------------------------------------------------
                 \1883\ See U.S. Energy Information Administration EIA, Today in
                Energy August 20, 2019, available at https://www.eia.gov/todayinenergy/detail.php?id=40973; Today in Energy September 12,
                2018, available at https://www.eia.gov/todayinenergy/detail.php?id=37053.
                ---------------------------------------------------------------------------
                 Reduced petroleum intensity and higher U.S. production have
                combined to produce a decline in U.S. petroleum imports--to
                approximately 20 percent of domestic consumption in 2017--which permits
                U.S. supply to act as a buffer against artificial or natural
                restrictions on global petroleum supplies due to military conflicts or
                natural disasters. In addition, the speed and relatively low
                incremental cost with which U.S. oil production has increased suggests
                that both the magnitude and (especially) the duration of future oil
                price shocks may be limited, because U.S. production offers the
                potential for a large and relatively swift supply response.
                 And while some risk of price shocks certainly still exists, even
                the potential for a large and swift U.S. production response may be
                playing a role in limiting the extent of price shocks attributable to
                external events. The large-scale attack on Saudi Arabia's Abqaiq
                processing facility--the world's largest crude oil processing and
                stabilization plant--on September 14, 2019 caused ``the largest single-
                day [crude oil] price increase in the past decade,'' of between $7 and
                $8 per
                [[Page 24726]]
                barrel, according to EIA.\1884\ The Abqaiq facility has the capacity to
                process 7 million barrels per day, or about 7 percent of global crude
                oil production capacity. EIA declared, however, that by September 17,
                only three days after the incident:
                ---------------------------------------------------------------------------
                 \1884\ https://www.eia.gov/todayinenergy/detail.php?id=41413.
                 Saudi Aramco reported that Abqaiq was producing 2 million
                barrels per day, and they expected its entire output capacity to be
                fully restored by the end of September. In addition, Saudi Aramco
                stated that crude oil exports to customers will continue by drawing
                on existing inventories and offering additional crude oil production
                from other fields. Tanker loading estimates from third-party data
                sources indicate that loadings at two Saudi Arabian export
                facilities were restored to the pre-attack levels. Likely driven by
                news of the expected return of the lost production capacity, both
                Brent and WTI crude oil prices fell on Tuesday, September 17.\1885\
                ---------------------------------------------------------------------------
                 \1885\ Id.
                 Thus, the largest single-day oil price increase in the past decade
                was largely resolved within a week, and assuming very roughly that
                average crude oil prices were $70/barrel in September 2019 (slightly
                higher than actual), an increase of $7/barrel would represent a 10
                percent increase as a result of the Abqaiq attack. Contrast this with
                the 1973 Arab oil embargo, which lasted for months and raised prices
                350 percent.\1886\ Saudi Arabia could have experienced increased
                revenue resulting from higher prices following the Abqaiq attack, but
                instead moved rapidly to restore production and tap reserves to control
                the risk of resulting price increases. In doing so, the Saudis likely
                recognized that sustained, long-term price increases would reduce their
                ability to control global supply (and thus prices and their own
                revenues) by relying on their lower cost of production.\1887\
                ---------------------------------------------------------------------------
                 \1886\ See Jeanne Whalen, ``Saudi Arabia's oil troubles don't
                rattle the U.S. as they used to,'' Washington Post, September 19,
                2019, available at https://www.washingtonpost.com/business/2019/09/19/saudi-arabias-oil-troubles-dont-rattle-us-like-they-used/.
                 \1887\ See, e.g., ``Dynamic Delivery: America's Evolving Oil and
                Natural Gas Transportation Infrastructure,'' National Petroleum
                Council (2019) at 18, available at: https://dynamicdelivery.npc.org/downloads.php.
                ---------------------------------------------------------------------------
                 Some commenters asserted that U.S. shale oil resources cannot serve
                as ``swing supply'' to provide stability in the face of a sudden,
                significant global supply disruption (Jason Bordoff,
                SAFE).1888 1889 Despite its greater responsiveness to price
                changes, commenters argued that lead time to bring new shale resources
                to market (6-12 months) is inferior to ``true spare capacity'' (like
                Saudi Arabia's large oil fields) because it cannot be deployed quickly
                enough to mitigate the economic consequences resulting from rapidly
                rising oil prices. Bordoff, however, also notes that shale oil
                projects' lead times are still shorter--and possibly much shorter--than
                conventional oil resource development. So, while new U.S. oil resources
                may take some time to respond to supply disruptions, they are
                nevertheless likely to provide a stabilizing influence on supply.
                ---------------------------------------------------------------------------
                 \1888\ NHTSA-2018-0067-11981.
                 \1889\ NHTSA-2018-0067-10718.
                ---------------------------------------------------------------------------
                 This is especially true for price increases that occur more slowly.
                When Beccue and Huntington updated their 2005 estimates of supply
                disruption probabilities in 2016,\1890\ they found that the probability
                distribution was generally flatter--suggesting that supply disruptions
                of most potential magnitudes were less likely to occur under today's
                market conditions than they had estimated previously in 2005. In
                particular, Beccue and Huntington find that supply disruptions of
                between two and four million barrels per day are significantly less
                likely than their previous estimates suggested. Although their recent
                study also estimated that larger supply disruptions (nine or more
                million barrels per day) are now slightly more likely to occur than in
                previous estimates, disruptions of that magnitude are extremely
                unlikely under either set of estimates.
                ---------------------------------------------------------------------------
                 \1890\ Beccue, Phillip, Huntington, Hillard, G., 2016. An
                Updated Assessment of Oil Market Disruption Risks: Final Report.
                Energy Modeling Forum, Stanford University.
                ---------------------------------------------------------------------------
                 Based on this review of the literature, the agencies concede that
                shale resources may not be able to stabilize oil markets fully to
                prevent a price increase associated with a large supply disruption
                elsewhere in the world. However, if supply disruptions are small
                enough, or move slowly enough, U.S. resources may be an adequate
                stabilizer.
                 The agencies reviewed further research that emphasizes the
                continued threat to the U.S. economy posed by the potential for sudden
                increases in global petroleum prices.\1891\ For example, Ramey and Vine
                (2010) note ``remarkable stability in the response of aggregate real
                variables to oil shocks once we account for the extra costs imposed on
                the economy in the 1970s by price controls and a complex system of
                entitlements that led to some rationing and shortages.'' \1892\ In
                contrast, another recent study found that while the likely effects of
                sudden oil price increases have become smaller over time, the declining
                sensitivity of petroleum demand to prices means that any future
                disruptions to oil supplies will have larger effects on petroleum
                prices, so that on balance their economic impact is likely to remain
                significant.\1893\
                ---------------------------------------------------------------------------
                 \1891\ Hamilton (2012) reviewed the empirical literature on oil
                shocks and concluded that its findings are mixed, noting that some
                recent research (e.g., Rasmussen and Roitman, 2011) finds either
                less evidence for significant economic effects of oil price shocks
                or declining effects (Blanchard and Gali 2010), while other research
                finds evidence of their continuing economic importance. See
                Hamilton, J. D., ``Oil Prices, Exhaustible Resources, and Economic
                Growth,'' in Handbook of Energy and Climate Change available at
                http://econweb.ucsd.edu/~jhamilto/handbook_climate.pdfhttp://
                econweb.ucsd.edu/~jhamilto/handbook_climate.pdf.
                 \1892\ Ramey, V. A., & Vine, D. J. ``Oil, Automobiles, and the
                U.S. Economy--How Much have Things Really Changed?'' National Bureau
                of Economic Research Working Paper 16067 (June 2010). Available at
                http://www.nber.org/papers/w16067.pdf.
                 \1893\ Baumeister, C. and G. Peersman (2012), ``The role of
                time-varying price elasticities in accounting for volatility changes
                in the crude oil market,'' Journal of Applied Econometrics 28 no. 7,
                November/December 2013, pp.1087-1109.
                ---------------------------------------------------------------------------
                 Some commenters (SAFE, CARB, Fuel Freedom Foundation, IPI)
                expressed skepticism that the United States could become a net
                petroleum exporter in the future without the continuation of the
                baseline standards. They cautioned that the global oil market is
                inherently uncertain, and Bordoff cautioned that America's shale
                resources may not last as long, or be as easy to develop, as they
                currently appear.\1894\ If the U.S. does not become a net exporter of
                petroleum as anticipated, any wealth effects from a high price of oil
                would continue to accrue to foreign owners of oil reserves. In
                addition, several of these commenters (CARB, SAFE, Bordoff, Zozana)
                argued that, regardless of whether or not the U.S. becomes a net
                petroleum exporter, its levels of petroleum consumption make it still
                vulnerable to price shocks arising in the global oil market.
                ---------------------------------------------------------------------------
                 \1894\ NHTSA-2018-0067-10718.
                ---------------------------------------------------------------------------
                 The agencies believe that the United States lacks the power
                (significantly) to control the global oil price and as a consequence
                remains vulnerable to the effects of oil price spikes, regardless of
                our own oil output. Geopolitical factors influence the global oil
                price--unstable regimes are often unreliable suppliers, large suppliers
                attempt strategically to manage supply to influence price or retain
                market share, and international negotiations around politically
                sensitive topics can influence the production behavior of firms in oil-
                rich nations. All of these factors, as well as wars and natural
                disasters, can influence the
                [[Page 24727]]
                global supply and the market price for oil.
                 In this analysis, any increase in the expected value of potential
                costs from economy-wide disruptions caused by sudden price increases
                that results from higher U.S. fuel and petroleum demand is accounted
                for separately from the direct cost for increased purchases of
                petroleum products. Consumers of petroleum products are unlikely to
                consider their contributions to these costs when deciding how much
                energy to consume, because those costs will be distributed widely
                throughout the economy, falling largely on businesses and households
                other than those whose decisions impose them. Thus, they represent an
                external (or ``social'') cost that users of petroleum energy such as
                transportation fuel are unlikely to internalize fully, and the agencies
                analysis includes the estimated increase in these costs among of the
                social costs stemming from the final rule. While increased U.S.
                petroleum production may impose some limits on their potential
                magnitude, their underlying source continues to be domestic petroleum
                use rather than imports.
                 Although the vulnerability of the U.S. economy to oil price shocks
                depends on aggregate petroleum consumption rather than on the level of
                oil imports, variation in U.S. oil imports may itself have some effect
                on the frequency, size, or duration of sudden oil price increases. The
                expected value of the resulting economic costs would also depend partly
                on the fraction of U.S. petroleum use that is supplied by imports.
                While total U.S. petroleum consumption is the primary determinant of
                potential economic costs to the Nation from rapid increases in oil
                prices, the estimate of these costs that have been relied upon on in
                past regulatory analyses--and in this analysis--is nevertheless
                expressed per unit (barrel) of imported oil. When they are converted to
                a per-gallon basis, they thus apply to fuel that is either imported in
                refined form, or refined domestically from imported crude petroleum.
                 Table VI-200 reports the per-barrel estimates of external costs
                from potential oil price shocks this analysis uses to estimate the
                increase in their total value likely to result from this final rule.
                These values differ from those used in previous analysis of CAFE and
                CO2 standards. In their comments on the NPRM, SAFE pointed
                out recent studies that have updated the estimates of the oil security
                premium since the study--on which the agencies relied upon in the
                NPRM--had been published. They depend in part on projected future oil
                prices, the elasticities of consumption with respect to price, income,
                and U.S. GDP. Since the NPRM values were last updated by the agencies,
                all of these factors have evolved in directions that would reduce the
                magnitude of the oil security premium, so continuing to use the NPRM
                values would have overestimated the increase in expected costs to the
                U.S. economy from potential oil price shocks calculated in this
                analysis, perhaps significantly.\1895\
                ---------------------------------------------------------------------------
                 \1895\ The costs reported in Table VI-188 also depend on the
                probabilities or expected frequencies of supply interruptions or
                sudden price shocks of different sizes and durations. The most
                recent reassessment of the probabilities on which these estimates
                are based (which were originally developed in 2005) was conducted in
                2016; see Beccue, Phillip C. and Hillard G. Huntington, An Updated
                Assessment of Oil Market Disruption Risks--Final Report EMF SR 10,
                Stanford University Energy Modeling Forum (February 5, 2016)
                available at https://emf.stanford.edu/publications/emf-sr-10-updated-assessment-oil-market-disruption-risks.
                ---------------------------------------------------------------------------
                 Specifically, the global petroleum prices projected in EIA's Annual
                Energy Outlook 2018 Reference Case range from 33-57 percent below those
                used to develop the estimates used in the NPRM and reported in Table
                VI-200. U.S. petroleum consumption and imports are now projected to be
                3-8 percent and 20-27 percent lower than the forecast values used to
                construct the NPRM estimates in the table. Finally, total petroleum
                expenditures are now projected to average 1.5-2.4 percent of U.S. GDP,
                in contrast to the 3.8-4.0 percent shares reflected in those values.
                Each of these differences suggests that the values in the NPRM
                overstated the current magnitude of potential costs to the U.S. economy
                from the risk of petroleum price shocks, and together they suggest that
                this overstatement may be significant. Indeed, the values used to
                support this final rule analysis are sourced from a recent paper by
                Brown.\1896\ Brown updates the underlying parameters used to estimate
                the oil security premium and finds a range of $0.60-$3.45 per barrel of
                imported oil, with a mean of $1.26 per barrel. The study, which was
                cited by SAFE, determines that the U.S. is less much less sensitive to
                oil price shocks than earlier estimates imply.\1897\ The values used in
                today's rule reflect that conclusion.
                ---------------------------------------------------------------------------
                 \1896\ See Brown, Stephen P.A., New estimates of the security
                costs of U.S. oil consumption, Energy Policy, Volume 13, 2018, Pages
                171-192.
                 \1897\ Another report cited by SAFE, Krupnick, et. al, similarly
                conclude that the macroeconomic cost of oil price shocks has
                diminished and that the oil security premium is lower than the
                majority of the existing literature would suggest. See Krupnick,
                Alan, Morgenstern, Richard, Balke, Nathan, Brown, Stephen P.A.,
                Herrera, Ana Maria, and Mohan, Shashank, ``Oil Supply Shocks, US
                Gross Domestic Product, and the Oil Security Premium,'' Resources
                for the Future, November 2017, available at: https://media.rff.org/documents/RFF-Rpt-OilSecurity.pdf (last accessed 01/2020).
                 \1898\ In order to convert per-barrel costs into per-gallon
                costs, we make the common assumption (used throughout the analysis)
                that each barrel of petroleum produces 42 gallons of motor gasoline.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24728]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.408
                [[Page 24729]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.409
                BILLING CODE 4910-59-C
                 Because they are expressed per barrel of petroleum that is imported
                (either in already-refined form as gasoline, or as crude petroleum to
                be refined domestically), applying these estimates requires the
                agencies to project of any changes in U.S. petroleum imports that are
                likely to result from the higher level of fuel consumption anticipated
                to occur as a result of this final rule. As discussed in detail in
                Section VI.D.3.c(b)(i) of this final rule, the agencies have elected to
                retain their previous assumptions that 50 percent of any increase in
                fuel consumption attributable to the rule will be accounted for through
                imports in refined form, and that 90 percent of the remaining increase
                would be refined domestically from imported petroleum. As a
                consequence, the oil security premiums shown in Table VI-200 are
                considered to be an external cost associated with 95 percent of the
                increase in gasoline consumption projected to result from this final
                rule.\1899\
                ---------------------------------------------------------------------------
                 \1899\ The 95 percent figure is calculates at 50 percent plus 90
                percent of the remaining 50 percent, or 50 percent plus 45 percent.
                ---------------------------------------------------------------------------
                (c) Potential Effects of Fuel Consumption and Petroleum Imports on U.S.
                Military Spending
                 A third potential effect of increasing U.S. demand for petroleum is
                an increase in U.S. military spending to secure the supply of oil
                imports from potentially unstable regions of the world and protect
                against their interruption. If an increase in fuel consumption that
                results from reducing CAFE and CO2 standards lead to higher
                military spending to protect oil supplies, this increase in outlays
                would represent an additional external or social cost of the agencies'
                action. Such costs could also include increased costs to maintain the
                U.S. Strategic Petroleum Reserve (SPR), because it is intended to
                cushion the U.S. economy against disruptions in the supply of imported
                oil or sudden increases in the global price of oil.
                 While several commenters argued that current U.S. military
                expenditures are uniquely attributable to securing U.S. supplies of
                petroleum from unstable regions of the globe--the Middle East, in
                particular--should be considered as a cost of this action (CARB, SAFE,
                Zonana), they seemed to confuse those costs with the marginal impact of
                increased oil consumption (relative to the baseline) on U.S. military
                activity and its costs. However, the agencies disagree with commenters
                that incremental changes to domestic consumption of oil for light-duty
                transportation could meaningfully change the scope or scale of the U.S.
                Department of Defense mission in the Persian Gulf region. Instead, they
                side with the Fuel Freedom Foundation, which noted in its comment,
                ``[i]ncrementally decreasing petroleum consumption does not
                significantly
                [[Page 24730]]
                decrease the military spending to protect and ensure its flow around
                the world.'' \1900\
                ---------------------------------------------------------------------------
                 \1900\ NHTSA-2018-0067-12016.
                ---------------------------------------------------------------------------
                 SAFE estimated a per-gallon cost of military externalities
                associated with U.S. dependence on petroleum products, and imported
                petroleum specifically.\1901\ Their low estimate of $0.28/gallon
                assumes $81 billion per year for protection of the global petroleum
                supply and divides those costs by the number of gallons consumed by
                U.S. drivers. In contrast, a similar analysis by Crane et al. stated,
                ``our analysis addresses the incremental cost to the defense budget of
                defending the production and transit of oil. It does not argue that a
                partial reduction of the U.S. dependence on imported oil would yield a
                proportional reduction in U.S. spending that is focused on this
                mission. The effect on military cost from such changes in petroleum use
                would be minimal.'' \1902\ The agencies thus do not believe that any
                incremental petroleum consumption that may result from this final rule
                will influence any fraction of U.S. defense spending that can be
                ascribed to protecting the global oil network.
                ---------------------------------------------------------------------------
                 \1901\ NHTSA-2018-0067-11981.
                 \1902\ Crane, K., A. Goldthau, M. Toman, T. Light, S.E. Johnson,
                A. Nader, A. Rabasa, & H. Dogo, Imported Oil and U.S. National
                Security, Santa Monica, CA, The RAND Corporation (2009) available at
                https://www.rand.org/pubs/monographs/MG838.html.
                ---------------------------------------------------------------------------
                 Eliminating petroleum imports (to both the U.S. and its national
                security allies) entirely might permit the Nation to scale back its
                military presence in oil-supplying regions of the globe to the extent
                that such interventions are driven by narrow concerns for oil
                production rather than other geopolitical considerations, but there is
                little evidence that U.S. military activity and spending in those
                regions have varied over history in response to fluctuations in the
                Nation's oil imports, or are likely to do so over the future period
                spanned by this analysis. Figure VI-80 shows that military spending as
                a share of total U.S. economic activity has gradually declined over the
                past several decades, and that any temporary--although occasionally
                major--reversals of this longer-term decline have been closely
                associated with U.S. foreign policy initiatives or overseas wars.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.410
                 Figure VI-81 superimposes U.S. petroleum consumption and imports on
                the history of military spending shown in the previous figure. Doing so
                shows that variation in U.S military spending throughout this period
                has had little association with the historical pattern of domestic
                petroleum purchases, changes in which instead primarily reflected the
                major increases in global petroleum prices that occurred in 1978-79,
                2008, and 2012-13. More important, Figure VI-81 also shows that U.S.
                military spending varied almost completely independently of the
                nation's imports of petroleum over this period. This history suggests
                that U.S. military activities--even in regions of the world that have
                [[Page 24731]]
                historically represented vital sources of oil imports--serve a far
                broader range of security and foreign policy objectives than simply
                protecting oil supplies. Thus, reducing the nation's consumption or
                imports of petroleum is unlikely by itself to lead to reductions in
                military spending.
                 SAFE further argued in its comments that the America's involvement
                in wars in the Persian Gulf region, starting with the first Gulf War
                and continuing through the Iraq War, has been a direct consequence of
                our dependence upon oil. In particular, they state that ``[w]hile there
                is debate over the precise role of oil in America's wars in the greater
                Middle East, several retired military members of SAFE's ESLC and other
                defense budget experts that were consulted for this report believe the
                connection is clear.'' \1903\ However, neither today's action, nor the
                baseline standards, has the ability to change the historical wealth
                transfer that created powerful nations in the Middle East. Attributing
                the cost of the Iraq War, for example, to oil dependence does not
                directly support an assertion that a marginal reduction in oil
                dependence could have reduced the cost of that conflict.
                ---------------------------------------------------------------------------
                 \1903\ NHTSA-2018-0067-11981.
                 [GRAPHIC] [TIFF OMITTED] TR30AP20.411
                
                 Further, the agencies were unable to find a record of the U.S.
                government attempting to calibrate U.S. military expenditures, force
                levels, or deployments to any measure of the Nation's petroleum use and
                the fraction supplied by imports, or to an assessment of the potential
                economic consequences of hostilities in oil-supplying regions of the
                world that could disrupt the global market.\1904\ Instead, changes in
                U.S. force levels, deployments, and spending in such regions appear to
                have been governed by purposeful foreign policy initiatives,
                [[Page 24732]]
                unforeseen political events, and emerging security threats, rather than
                by shifts in U.S. oil consumption or imports.\1905\
                ---------------------------------------------------------------------------
                 \1904\ Crane et al. (2009) analyzed reductions in U.S. forces
                and associated cost savings that could be achieved if oil security
                were no longer a consideration in military planning, and disagree
                with this assessment. After reviewing recent allocations of budget
                resources, they concluded that ``. . . the United States does
                include the security of oil supplies and global transit of oil as a
                prominent element in its force planning'' at p. 74 (emphasis added).
                Nevertheless, their detailed analysis of individual budget
                categories estimated that even eliminating the protection of foreign
                oil supplies completely as a military mission would reduce the
                current U.S. defense budget by approximately 12-15 percent. See
                Crane, K., A. Goldthau, M. Toman, T. Light, S.E. Johnson, A. Nader,
                A. Rabasa, & H. Dogo, Imported Oil and U.S. National Security.,
                Santa Monica, CA, The RAND Corporation (2009) available at https://www.rand.org/pubs/monographs/MG838.html.
                 \1905\ Crane et al. (2009) also acknowledge the difficulty of
                reliably allocating U.S. military spending by specific mission or
                objective, such as protecting foreign oil supplies. Moore et al.
                (1997) conclude that protecting oil supplies cannot be distinguished
                reliably from other strategic objectives of U.S. military activity,
                so that no clearly separable component of military spending to
                protect oil flows can be identified, and its value is likely to be
                near zero. Similarly, the U.S. Council on Foreign Relations (2015)
                takes the view that significant foreign policy missions will remain
                over the foreseeable future even without any imperative to secure
                petroleum imports. A dissenting view is that of Stern (2010), who
                argues that other policy concerns in the Persian Gulf derive from
                U.S. interests in securing oil supplies, or from other nations'
                reactions to U.S. policies that attempt to protect its oil supplies.
                See Crane, K., A. Goldthau, M. Toman, T. Light, SE Johnson, A.
                Nader, A. Rabasa, and H. Dogo, Imported Oil and U.S. National
                Security., Santa Monica, CA, The RAND Corporation (2009) available
                at https://www.rand.org/pubs/monographs/MG838.html; Moore, John L.,
                E.J. Carl, C. Behrens, and John E. Blodgett, ``Oil Imports--An
                Overview and Update of Economic and Security Effects,''
                Congressional Research Service, Environment and Natural Resources
                Policy Division, Report 98, No. 1 (1997), pp. 1-14; Council on
                Foreign Relations, ``Automobile Fuel Economy Standards in a Lower-
                Oil-Price World,'' November 2015; and Stern, Roger J. ``United
                States cost of military force projection in the Persian Gulf, 1976-
                2007,'' Energy Policy 38, no. 6 (June 2010), pp. 2816-25, https://www.sciencedirect.com/science/article/pii/S0301421510000194?via%3Dihub.
                ---------------------------------------------------------------------------
                 The agencies thus conclude that U.S. military activity and
                expenditures are unlikely to be affected by even relatively large
                changes in consumption of petroleum-derived fuels by light duty
                vehicles. Certainly, the historical record offers no suggestion that
                U.S. military spending is likely to adjust significantly in response to
                the increase in domestic petroleum use that would result from reducing
                CAFE and CO2 standards.
                 Nevertheless, it is possible that more detailed analysis of
                military spending might identify some relationship to historical
                variation in U.S. petroleum consumption or imports. A number of studies
                have attempted to isolate the fraction of total U.S. military spending
                that is attributable to protecting overseas oil supplies.\1906\ These
                efforts have produced varying estimates of how much it might be reduced
                if the U.S. no longer had any strategic interest in protecting global
                oil supplies. However, none has identified an estimate of spending that
                is likely to vary incrementally in response to changes in U.S.
                petroleum consumption or imports.
                ---------------------------------------------------------------------------
                 \1906\ These include Copulos, M R. ``America's Achilles Heel--
                The Hidden Costs of Imported Oil,'' Alexandria VA--The National
                Defense Council Foundation, September 2003-1-153, available at
                http://ndcf.dyndns.org/ndcf/energy/NDCFHiddenCostsofImported_Oil.pdf; Copulos, M R. ``The Hidden Cost
                of Imported Oil--An Update.'' The National Defense Council
                Foundation (2007) available at http://ndcf.dyndns.org/ndcf/energy/NDCF_Hidden_Cost_2006_summary_paper.pdf; Delucchi, Mark A. & James
                J. Murphy. ``US military expenditures to protect the use of Persian
                Gulf oil for motor vehicles,'' Energy Policy 36, no. 6 (June 2008),
                pp. 2253-64; and National Research Council Committee on Transitions
                to Alternative Vehicles and Fuels, Transitions to Alternative
                Vehicles and Fuels (2013).
                ---------------------------------------------------------------------------
                 Nor have any of these studies tracked changes in spending that can
                be attributed to protecting U.S. interests in foreign oil supplies over
                a prolonged period, so they have been unable to examine whether their
                estimates of such spending vary in response to fluctuations in domestic
                petroleum consumption or imports. The agencies conclude from this
                review of research that U.S. military commitments in the Persian Gulf
                and other oil-producing regions of the world contribute to worldwide
                economic and political stability, and insofar as the costs of these
                commitments are attributable to petroleum use, they are attributable to
                oil consumption throughout the world, rather than simply U.S. oil
                consumption or imports.
                 It is thus unlikely that military spending would rise in response
                to any increase in U.S. imports that did result from this final rule.
                As a consequence, the analysis of alternative CAFE and CO2
                emission standards for future model years applies no increase in
                government spending to support U.S. military activities as a potential
                cost of allowing new cars and light trucks to achieve lower fuel
                economy and thus increasing domestic petroleum use.
                 Similarly, while the ideal size of the Strategic Petroleum Reserve
                from the standpoint of its potential stabilizing influence on global
                oil prices may be related to the level of U.S. petroleum consumption or
                imports, its actual size has not appeared to vary in response to either
                of those measures. The budgetary costs for maintaining the SPR are thus
                similar to U.S. military spending in that, while they are not reflected
                in the market price for oil (and thus do not enter consumers' decisions
                about how much to use), they do not appear to have varied in response
                to changes in domestic petroleum consumption or imports.
                 As a consequence, the analysis does not include any potential
                increase in the cost to maintain a larger SPR among the external or
                social costs of the increase in gasoline and petroleum consumption
                likely to result from reducing future CAFE and CO2
                standards. This view aligns with the conclusions of most recent studies
                of military-related costs to protect U.S. oil imports, which generally
                conclude that savings in military spending are unlikely to result from
                incremental reductions in U.S. consumption of petroleum products on the
                scale of those that would resulting from adopting higher CAFE or
                CO2 standards.
                (13) Social Cost of Carbon
                 In the proposal, the agencies projected costs resulting from fuel
                consumption and emissions of CO2 using estimates of
                anticipated climate-related economic damages within U.S. borders per
                ton of CO2 emissions, which the agencies referred to as the
                domestic social cost of carbon (domestic SC-CO2). The
                domestic SC-CO2 estimates, which were originally developed
                by EPA for an earlier regulatory analysis, represent the monetary value
                of damages to the domestic economy likely to be caused by future
                changes in the climate that result from incremental increases in
                CO2 emissions during a given year.\1907\ The agencies did
                not consider climate-related damage costs resulting from emissions of
                other greenhouse gases (GHGs), such as methane or nitrous oxide, in
                their analysis supporting the proposal.
                ---------------------------------------------------------------------------
                 \1907\ For a description of the procedures EPA used to develop
                these values, see U.S. Environmental Protection Agency, Regulatory
                Impact Analysis for the Proposed Emission Guidelines for Greenhouse
                Gas Emissions from Existing Electric Utility Generating Units;
                Revisions to Emission Guideline Implementing Regulations; Revisions
                to New Source Review Program, EPA-452/R-18-006, August 2018 (https://www.epa.gov/sites/production/files/2018-08/documents/utilities_ria_proposed_ace_2018-08.pdf), Section 4.3, at 4-2 to 4-7.
                The sources and potential magnitude of uncertainties surrounding the
                SC-CO2 estimates are described in Chapter 7 of that same
                document, at 7-1 to 7-10.
                ---------------------------------------------------------------------------
                 Climate-related damages caused by emissions of CO2 and
                other GHGs include changes in agricultural productivity, adverse
                effects on human health, property damage from increased flood risk, and
                changes in costs for managing indoor environments in commercial and
                residential buildings (such as costs for heating and air conditioning),
                among other possible damages.
                 The agencies described the SC-CO2 estimates used in the
                NPRM analysis as interim values developed under Executive Order 13783,
                which are to be used in regulatory analyses until revised values that
                incorporate recommendations from NAS can be developed.\1908\ E.O. 13783
                directed
                [[Page 24733]]
                agencies to ensure that estimates of the social cost of greenhouse
                gases used in regulatory analyses are consistent with the guidance
                contained in OMB Circular A-4, ``including with respect to the
                consideration of domestic versus international impacts and the
                consideration of appropriate discount rates.'' \1909\
                ---------------------------------------------------------------------------
                 \1908\ The guidance followed by EPA in developing the SC-
                CO2 values used in the NPRM analysis appears in President
                of the United States, Executive Order 13783, ``Promoting Energy
                Independence and Economic Growth,'' March 28, 2017, Federal
                Register, Vol. 82, No. 61, Friday, March 31, 2017, 16093-97.
                (https://www.govinfo.gov/content/pkg/FR-2017-03-31/pdf/2017-06576.pdf) The recommendations of the National Academies are
                reported in National Academies of Sciences, Engineering, and
                Medicine, Valuing Climate Damages: Updating Estimation of the Social
                Cost of Carbon Dioxide, Washington, DC, January 2017. Revised values
                incorporating this guidance have not yet been developed.
                 https://www.nap.edu/catalog/24651/valuing-climate-damages-updating-estimation-of-the-social-cost-of.
                 \1909\ E.O. 13783, at 16096.
                ---------------------------------------------------------------------------
                 Circular A-4 states that analysis of economically significant
                regulations ``should focus on benefits and costs that accrue to
                citizens and residents of the United States,'' and the agencies
                followed this guidance by using estimates of the SC-CO2 that
                included only domestic economic damages. In response to Circular A-4's
                further guidance that regulatory analyses ``should provide estimates of
                net benefits using [discount rates of] both 3 percent and 7 percent,''
                the agencies presented estimates of the proposed rule's economic
                impacts--including the costs of climate damages likely to result from
                increased CO2 emissions--that incorporated both discount
                rates. The PRIA included a detailed discussion of the analyses used to
                construct estimates of the domestic SC-CO2 using these
                discount rates.\1910\
                ---------------------------------------------------------------------------
                 \1910\ See NHTSA and EPA, PRIA, Chapter 8, Appendix A.
                ---------------------------------------------------------------------------
                 The estimates of the domestic SC-CO2 the agencies used
                in their analysis supporting the proposal increased over future years,
                partly because emissions during future years are anticipated to
                contribute larger incremental costs. Future values of the SC-
                CO2 also increase because U.S. GDP is growing over time, and
                many categories of climate-related damage are estimates as proportions
                of GDP. The agencies' estimates of the domestic SC-CO2 for
                emissions occurring in the year 2020 were $1 and $8 (in 2016$) per
                metric ton of CO2 emissions using 7 and 3 percent discount
                rates, and these values were projected to increase to $2 and $10 (again
                in 2016$) by the year 2050.
                 As the agencies indicated in the NPRM, the SC-CO2
                estimates are subject to several sources of uncertainty. In accordance
                with guidance provided by OMB Circular A-4 for treating uncertainty in
                regulatory analysis, the PRIA included a detailed discussion of how the
                analysis used to develop the interim SC-CO2 estimates
                incorporated sources of uncertainty that could be quantified. It also
                demonstrated how considering the uncertainty introduced by applying
                discount rates over extended time horizons could affect the estimated
                values.\1911\ To reflect this uncertainty, the analysis supporting the
                proposed rule examined the sensitivity of its estimated costs and
                benefits to using higher values for the SC-CO2 ($9-14 per
                metric ton), which were derived using a lower ``intergenerational''
                discount rate of 2.5 percent.\1912\
                ---------------------------------------------------------------------------
                 \1911\ See PRIA, Chapter 8, Appendix A.
                 \1912\ PRIA, Tables 13-8 and 13-9, at 1547-50.
                ---------------------------------------------------------------------------
                (a) Comments on the NPRM Value for the SC-CO2
                 The agencies received extensive comments on the values of the SC-
                CO2 used in the NPRM analysis. Broadly, these comments
                stressed the following concerns:
                 Using a domestic value for SC-CO2 systemically
                underestimates the benefits of adopting stricter standards.
                 The agencies' SC-CO2 omits potential costs due
                to foreign social and political disruptions caused by climate change
                that can affect the U.S.
                 The 7 percent discount rate used in the agencies' main or
                central analysis is inappropriate because it represents an opportunity
                cost of capital rather than a rate of time preference for current
                versus future consumption opportunities, and climate change will affect
                future consumption.
                 (b) Domestic vs. Global Value for SC-CO2
                 Many commenters asserted that it was inappropriate for the agencies
                to use a domestic SC-CO2 value for analyzing benefits or
                costs from changing required levels of fuel economy in the NPRM
                analysis, primarily because doing so could lead regulatory agencies to
                adopt measures that provide inadequate reductions in emissions and
                protection from potential climate change.
                 As noted in the NPRM and above, the SC-CO2 estimates the
                agencies used to estimate climate-related economic costs from adopting
                less demanding fuel economy and CO2 emission were developed
                in response to the issuance of E.O. 13783. The agencies remind
                commenters that E.O. 13783 directed federal agencies to ensure that
                estimates of the social cost of greenhouse gases used in their
                regulatory analyses are consistent with the guidance contained in OMB
                Circular A-4, ``including with respect to the consideration of domestic
                versus international impacts and the consideration of appropriate
                discount rates.'' \1913\ Circular A-4 states that analysis of
                economically significant proposed and final regulations ``should focus
                on benefits and costs that accrue to citizens and residents of the
                United States.'' \1914\ The agencies adhered closely to this guidance
                in evaluating the economic costs and benefits in the proposal and this
                final rule by using the domestic value of the SC-CO2 in our
                central analysis.
                ---------------------------------------------------------------------------
                 \1913\ Executive Order 13,783, at 16096.
                 \1914\ White House Office of Management and Budget, Circular A-
                4: Regulatory Analysis, September 17, 2003, at 15. (https://www.whitehouse.gov/sites/whitehouse.gov/files/omb/circulars/A4/a-4.pdf).
                ---------------------------------------------------------------------------
                 Commenters argued that Circular A-4 allows the agencies to use a
                global SC-CO2 in their central analysis. For example, IPI et
                al. commented that ``Circular A-4's reference to effects `beyond the
                borders' confirms that it is appropriate for agencies to consider the
                global effects of U.S. greenhouse gas emissions.'' \1915\ While the
                agencies agree that Circular A-4 authorizes the agencies to consider
                foreign impacts in certain circumstances, the agencies would also like
                to note that Executive Order 13783 stipulates ``when monetizing the
                value of changes in greenhouse gas emissions resulting from
                regulations, including with respect to the consideration of domestic
                versus international impact [. . .] agencies shall ensure [. . .] any
                such estimates are consistent with the guidance contained in OMB
                Circular A-4.'' \1916\ Using a global SC-CO2 in our central
                analysis would be inconsistent with Circular A-4's directive that any
                non-domestic effects calculated ``should be reported separately.''
                \1917\ As such, if the agencies had used a global SC-CO2,
                this rulemaking would be compelled by Circular A-4 to separate the SC-
                CO2 into domestic and foreign components, and to include
                only the former in our central analysis.
                ---------------------------------------------------------------------------
                 \1915\ IPI et al., DEIS Joint SCC Comments, NHTSA-2017-0069-
                0559, at 20.
                 \1916\ Executive Order 13,783, at 16096.
                 \1917\ Specifically, OMB Circular A-4 directs federal agencies
                as follows: ``Where you choose to evaluate a regulation that is
                likely to have effects beyond the borders of the United States,
                these effects should be reported separately.'' OMB Circular A-4, at
                15.
                ---------------------------------------------------------------------------
                 Furthermore, today's analysis will likely have global impacts
                beyond climate change. For example, freeing manufacturers who compete
                in the U.S. domestic automobile market from burdensome fuel efficiency
                standards may enable them to dedicate time and resources to becoming
                more competitive in global markets, and is thus likely to affect
                product innovation and performance throughout the global auto
                [[Page 24734]]
                market.\1918\ It would be inconsistent to report the global SC-
                CO2 while ignoring other global costs and benefits. The
                agencies do not have a method for analyzing the comprehensive impacts
                of CAFE and CO2 standards--including their many likely
                impacts beyond climate change--on a global scale, and did not receive
                any suggestions about how to conduct such an analysis from commenters.
                Because it would be inconsistent to quantify only climate change and
                none of these other potential global-scale impacts, the agencies have
                decided to focus their attention on domestic impacts, which are more
                readily measurable.
                ---------------------------------------------------------------------------
                 \1918\ Some commenters assert that weakening U.S. fuel economy
                standards could make domestic auto companies less competitive in
                international markets, since several other nations have also adopted
                similar standards. For reasons discussed Section VIII.B.6. of this
                rule, however, the agencies find these comments unpersuasive.
                ---------------------------------------------------------------------------
                 Several commenters argued that the agencies are still obligated to
                report the global impacts of carbon. For example, the North Carolina
                Department of Environmental Quality commented that ``by omitting any
                analysis of the global social cost of carbon, [the agencies] failed to
                adhere to OMB's Circular A-4.'' \1919\ The agencies note Circular A-4
                grants agencies discretion to choose which impacts to report. However,
                to be fully informed of the gamut of potential effects of today's rule,
                the agencies have included two sensitivity cases analyzing the impacts
                of the standards using a global SC-CO2.
                ---------------------------------------------------------------------------
                 \1919\ North Carolina Department of Environmental Quality,
                Comments, NHTSA-2018-0067-12025, at 39.
                ---------------------------------------------------------------------------
                 (c) Scope of Domestic Climate Damages
                 Some commenters asserted that even if the agencies are required to
                use a domestic SC-CO2, the specific value employed by the
                agencies underestimated the domestic impacts of climate change. They
                argued the agencies failed to incorporate economic costs associated
                with social or economic disruptions caused by climate change in regions
                of the world that were more vulnerable to its effects, but that could
                ``spill over'' to impose damages to the U.S. via their effects on
                migration patterns, international trade flows, or other mechanisms that
                connect nations. Other commenters argued that E.O. 13783 does not
                prohibit the agencies from using the estimates or practices developed
                by the IWG to develop new estimates of the SC-CO2, and
                asserted that the IWG's methods and resulting estimates continue to
                represent the best available practices.
                 However, all of the IWG's estimates measure the global SC-
                CO2, and as discussed previously, E.O. 13783, in conjunction
                with Circular A-4, directs the agencies to use a domestic SC-
                CO2 which precludes the use of the IWG estimates. To develop
                interim estimates of the domestic SC-CO2 that were
                consistent with the IWG's procedures, EPA used the same three climate
                economic models the IWG employed previously to calculate the domestic
                SC-CO2. Two of those three models directly estimate the U.S.
                domestic SC-CO2, which represents the economic costs
                resulting from climate change that are likely to be borne within U.S.
                borders.\1920\ The third model the IWG used previously does not
                estimate the domestic SC-CO2 directly, but EPA approximated
                domestic U.S. costs from future climate change as 10 percent of its
                estimate of their global value, based on results from a companion model
                developed by the same author.\1921\ Thus the agencies believed that the
                SC-CO2 values they used in the NPRM analysis represented the
                most reliable estimates of domestic economic costs from future climate
                change that were available for use in evaluating the proposal.
                ---------------------------------------------------------------------------
                 \1920\ The Policy Analysis of the Greenhouse Effect (PAGE) model
                is described in Hope, C., ``The marginal impact of CO2
                from PAGE2002: An integrated assessment model incorporating the
                IPCC's five reasons for concern,'' The Integrated Assessment
                Journal, Vol. 6 No. 1 (2006), at 19-56; and Hope, C., ``Optimal
                carbon emissions and the social cost of carbon under uncertainty,''
                The Integrated Assessment Journal Vol. 8, No. 1 (2008), at 107-22.
                The Climate Framework for Uncertainty, Negotiation, and Distribution
                (FUND) model is documented in Tol, Richard, ``Estimates of the
                damage costs of climate change. Part I: benchmark estimates,'' and
                ``Estimates of the damage costs of climate change. Part II: Dynamic
                estimates.'' Environmental and Resource Economics Vol 21 (2002), at
                47-73 and 135-60.
                 \1921\ The third model is the Dynamic Integrated model of
                Climate and the Economy (DICE), described in Nordhaus, William,
                ``Estimates of the Social Cost of Carbon: Concepts and Results from
                the DICE-2013R Model and Alternative Approaches.'' Journal of the
                Association of Environmental and Resource Economists, Vol. 1, No. 2
                (2014), at 273-312 (https://www.jstor.org/stable/pdf/10.1086/676035.pdf). The 10 percent figure is based on the results from a
                regional version of that model (RICE 2010), as described in
                Nordhaus, William D. 2017, ``Revisiting the social cost of carbon,''
                Proceedings of the National Academy of Sciences of the United
                States, 114 (7), at 1518-23, Table 2. (https://pdfs.semanticscholar.org/f83b/3a7431e0ae2d4e8be3d0ee5f3787a802c34c.pdf?_ga=2.211824467.636056015.1572384992-158339427.1562696454).
                ---------------------------------------------------------------------------
                 The agencies were unable to develop an estimate of the domestic
                value for SC-CO2 that incorporated any of these alleged
                spillover effects, due both to their speculative nature and to the
                absence of credible empirical estimates of their potential magnitude.
                Nor did commenters provide credible explanations for how such
                spillovers might arise, or reliable empirical estimates of their
                potential magnitude.
                (d) Discount Rate Used To Construct the SC-CO2 Value
                 Many commenters also objected to the agencies use of an SC-
                CO2 value that incorporated a 7 percent discount rate in the
                NPRM analysis. Some of these comments reflected a misperception that
                the agencies used such a value in their main or central analysis, when
                in fact it was only used in a sensitivity analysis case as described
                below. Other comments appeared to object to the agencies' use of an SC-
                CO2 value incorporating a 7 percent discount rate even as a
                sensitivity case.
                 E.O. 13783 directed agencies to ensure that any estimates of the
                social cost of CO2 and other greenhouse gases they used for
                purposes of regulatory analyses are consistent with OMB Circular A-4's
                guidance ``with respect to the consideration of. . .appropriate
                discount rates.'' \1922\ In turn, Circular A-4 refers agencies to OMB's
                earlier guidance on discounting contained in its Circular A-94, noting
                that ``[a]s a default position, OMB Circular A-94 states that a real
                discount rate of 7 percent should be used as a base-case for regulatory
                analysis.'' \1923\ OMB continues to use the 7 percent rate to estimate
                the average pre-tax rate of return to private capital investment
                throughout the U.S. economy. Because it is intended to approximate the
                opportunity cost of capital, it is the appropriate discount rate for
                evaluating the economic consequences of regulations that affect
                private-sector capital investments.
                ---------------------------------------------------------------------------
                 \1922\ E.O. 13,783, at 16096.
                 \1923\ OMB Circular A-4, at 33.
                ---------------------------------------------------------------------------
                 At the same time, however, OMB's guidance on discounting also
                recognizes that some federal regulations are more likely to affect
                private consumption decisions made by households and individuals, such
                as when they affect prices or other attributes of consumer goods. In
                these cases, Circular A-4 advises that a lower discount rate is likely
                to be more appropriate, and that a reasonable choice for such a lower
                rate is the real consumer (or social) rate of time preference. This is
                the rate at which individual consumers discount future consumption to
                determine its present value to them.
                 OMB estimated that the rate of consumer time preference has
                averaged 3 percent in real or inflation-adjusted terms over an extended
                period, and continues to use that value. In summary, Circular A-4
                reiterates the guidance provided in OMB's earlier Circular A-
                [[Page 24735]]
                94 that ``[f]or regulatory analysis, you should provide estimates of
                net benefits using both 3 percent and 7 percent.'' \1924\
                ---------------------------------------------------------------------------
                 \1924\ OMB Circular A-4, at 34.
                ---------------------------------------------------------------------------
                 Finally, OMB's guidance on discounting indicates that it may be
                appropriate for government agencies to employ an even lower rate of
                time preference when their regulatory actions entail tradeoffs between
                improving the welfare of current and future generations. Recognizing
                this situation, Circular A-4 advises if the ``rule will have important
                intergenerational benefits or costs [an agency] might consider a
                further sensitivity analysis using a lower but positive discount rate
                in addition to calculating net benefits using discount rates of 3 and 7
                percent.'' \1925\
                ---------------------------------------------------------------------------
                 \1925\ OMB Circular A-4, at 36.
                ---------------------------------------------------------------------------
                 The agencies adhered closely to each of these provisions of OMB's
                guidance on discounting future climate-related economic costs in their
                analysis supporting the NPRM. Specifically, their central analysis
                relied exclusively on a SC-CO2 value that was constructed by
                applying a 3 percent discount rate to future climate-related economic
                damages. This value ranged from $6 per metric ton in 2015 to nearly $11
                per metric ton (both figures in 2016$) by the end of the analysis
                period, the year 2050.
                 Throughout the NPRM central analysis, costs resulting from
                increased emissions of CO2 were also discounted from the
                year when those increases in emissions occurred to the present using a
                3 percent rate, even when all other future costs and benefits were
                discounted at a 7 percent rate. Thus the agencies' central analysis for
                the NPRM did not use SC-CO2 values for future years that
                were constructed by applying a 7 percent rate to discount distant
                future climate-related economic damages, and did not use a 7 percent
                rate to discount costs of increased CO2 from the years when
                they were projected to occur to 2018 (the base year used in the
                analysis).
                 Notwithstanding concerns raised by commenters about including a
                sensitivity analysis that used a higher discount rate, OMB's guidance
                clearly directs the agencies to report estimates of the present value
                of the economic costs resulting from increased CO2 emissions
                that reflect discount rates of both 3 and 7 percent. Thus to supplement
                their central analysis, which as indicated previously employed a 3
                percent discount rate throughout, the agencies also reported an
                estimate of the economic costs of increased CO2 emissions
                based on a value for the SC-CO2 that was constructed using a
                7 percent discount rate as a sensitivity case, which they termed the
                ``Low Social Cost of Carbon'' sensitivity analysis.\1926\ The values
                for the SC-CO2 used in the Low Social Cost of Carbon
                sensitivity analysis varied from $1 per metric ton in 2015 to $3 per
                metric ton (both figures in 2016$) by the end of the analysis period.
                Using these values reduced the loss in total economic benefits
                resulting from the proposed alternative by 1.1 percent, thus increasing
                its net benefits by slightly less than 2 percent.\1927\
                ---------------------------------------------------------------------------
                 \1926\ PRIA, Table 13-1, at 1531-34.
                 \1927\ PRIA, Tables 13-8 and 13-9, at 1547-50. Using a lower
                value for the SC-CO2 had opposite effects on the
                proposal's total and net economic benefits, because its net benefits
                represented the difference between the loss in benefits and the
                savings in costs that would result from adopting the proposed rule,
                compared to the baseline of adopting the Augural standards.
                ---------------------------------------------------------------------------
                 For the proposal, the agencies also included a second sensitivity
                analysis using a value for the SC-CO2 that reflected a lower
                ``intergenerational'' discount rate of 2.5 percent, which is within the
                1 to 3 percent range for discount rates that have previously been
                applied to economic costs and benefits that span multiple generations,
                as reported in OMB guidance.\1928\ Because using a lower discount rate
                results in a higher value for the SC-CO2, this analysis was
                termed the ``High Social Cost of Carbon'' sensitivity case.\1929\ The
                values for the SC-CO2 used in this additional sensitivity
                analysis varied from $8 per metric ton in 2015 to $14 per metric ton
                (both figures in 2016$) in 2050, the last year of the analysis. Using
                these higher values increased the magnitude of the estimated loss in
                economic benefits resulted from adopting the proposed rule (versus
                retaining the Augural standards) by 0.5 percent from that estimated in
                the central analysis, thus reducing its net benefits by 1.0
                percent.\1930\ Thus it appeared that when used to construct alternative
                estimates of the SC-CO2, the range of discount rates
                specified in OMB Circular A-4 had little or no effect on the estimated
                total benefits of the proposed rule, and the sensitivity analyses
                conducted in support of this Final Rule confirm this result.\1931\
                ---------------------------------------------------------------------------
                 \1928\ OMB Circular A-4, at 36.
                 \1929\ PRIA, Table 13-1, at 1531-34.
                 \1930\ PRIA, Tables 13-8 and 13-9, at 1547-50. As in the Low
                Social Cost of Carbon sensitivity case, using a higher value for the
                SC-CO2 had opposite effects on the total and net economic
                benefits, because its net benefits were the difference between the
                sacrifice in benefits and the savings in costs from adopting the
                proposed rule, where both were measured against the baseline of
                adopting the Augural standards.
                 \1931\ See section VII.B. of this Final Rule for results of the
                ``High Social Cost of Carbon'' sensitivity case.
                ---------------------------------------------------------------------------
                (e) SC-CO2 for the Final Rule
                 After carefully considering the concerns raised by commenters, the
                agencies decided to leave the SC-CO2 values unchanged for
                the final rule. This means the SC-CO2 estimate used in this
                analysis is still a domestic value that was constructed using a 3
                percent discount rate, and that costs from increased CO2
                emissions are discounted from the year those emissions occur to the
                present using a 3 percent rate. The agencies have again included ``High
                Social Cost of Carbon'' and ``Low Social Cost of Carbon'' sensitivity
                analyses, which continue to use domestic SC-CO2 values that
                incorporate alternative discount rates of 2.5 percent and 7 percent.
                 The agencies have also added two sensitivity cases using global
                values for the SC-CO2, which reflect discount rates of 3
                percent and 7 percent. Finally, the agencies have also included an
                additional sensitivity case that incorporates estimates of the domestic
                climate damage costs caused by emissions of the GHGs methane
                (CH4) and nitrous oxide (N2O). Like the SC-
                CO2 values used in this analysis, the estimates of the
                domestic values for SC-CH4 and SC-N2O are interim
                estimates developed by EPA for use in regulatory analyses conducted
                under the guidelines specified in E.O. 13783 and OMB Circular A-4, and
                incorporate a 3 percent discount rate.
                (14) External Costs of Congestion and Noise
                (a) Values Used To Analyze the Proposal
                 As explained in the proposal, changes in vehicle use affect the
                levels and economic costs of traffic congestion and highway noise
                associated with motor vehicle use.\1932\ Congestion and noise costs are
                ``external'' to the vehicle owners whose decisions about how much,
                where, and when to drive more--
                [[Page 24736]]
                or less--in response to changes in fuel economy result in these costs.
                Therefore, unlike changes in the costs incurred by drivers for fuel
                consumption or safety risks they willingly assume, changes in
                congestion and noise costs are not offset by corresponding changes in
                the travel benefits drivers experience.\1933\
                ---------------------------------------------------------------------------
                 \1932\ The proposal estimated changes in congestion and noise
                costs associated with the overall change in vehicle use, which
                included changes in the use of new cars and light trucks associated
                with the fuel economy rebound effect as well as with changes in the
                use of older vehicles resulting from the effect of CAFE and
                CO2 standards on turnover in the car and light truck
                fleets. As discussed in more detail elsewhere in this final rule,
                the current analysis assumes that total vehicle use (VMT) differs
                between the baseline and regulatory alternatives only because of
                changes in the use of cars and light trucks produced during the
                model years affected by this rule that occur in response to the fuel
                economy rebound effect.
                 \1933\ The potential contribution of increased vehicle use to
                the costs of injuries and property damage caused by motor vehicle
                crashes may also be partly external to drivers who elect to travel
                more in response to the fuel economy rebound effect. However, these
                costs are dealt with directly and in more detail than the external
                costs of congestion and noise, in section VI.C.2. below.
                ---------------------------------------------------------------------------
                 Congestion costs are limited to road users; however, since road
                users include a significant fraction of the U.S. population, changes in
                congestion costs are treated as part of the rule's economic impact on
                the broader U.S. economy instead of as a cost or benefit to private
                parties. Costs resulting from road and highway noise are even more
                widely dispersed, because they are borne partly by surrounding
                residents, pedestrians, and other non-road users, and for this reason
                are also considered as a cost to the U.S. economy as a whole.
                 To estimate the economic costs associated with changes in
                congestion and noise caused by differences in miles driven, the
                analysis supporting the NPRM used estimates of per-mile congestion and
                noise costs from increased automobile and light truck use that were
                originally developed by FHWA as part of its 1997 Highway Cost
                Allocation Study.\1934\ The agencies previously employed these same
                cost estimates in the 2010, 2011, and 2012 final rules.
                ---------------------------------------------------------------------------
                 \1934\ Federal Highway Administration, 1997 Highway Cost
                Allocation Study, Chapter V, Tables V-22 and V-23, available at
                https://www.fhwa.dot.gov/policy/hcas/final/five.cfm.
                ---------------------------------------------------------------------------
                 The marginal congestion cost estimates reported in the 1997 FHWA
                study were intended to measure the costs of increased congestion
                resulting from incremental growth in travel by different types of
                vehicles (including autos and light trucks), and the delays it causes
                to drivers, passengers, and freight shipments. As explained in the 1997
                FHWA study, the distinction between marginal and average costs is
                extremely important in considering congestion costs on a per-vehicle-
                mile basis. Average congestion costs on a section of highway are
                calculated as the total congestion costs experienced by all vehicles,
                divided by total vehicle miles. In contrast, marginal congestion costs
                are calculated as the increase in congestion costs resulting from an
                incremental increase in vehicle miles.
                 Marginal congestion costs are significantly higher than average
                congestion costs because each additional vehicle that enters a crowded
                roadway slows travel speeds only slightly, thus adding only modestly to
                the average travel time of vehicles already on the road. During
                congested conditions, however, this modest increase is experienced by a
                very large number of vehicles, so the resulting increase in total delay
                experienced by all travelers using the road can be extremely large. As
                a consequence, the increases in total delay and congestion costs
                associated with additional driving are more than proportional to
                changes in VMT that cause them.\1935\
                ---------------------------------------------------------------------------
                 \1935\ Such ``non-linearity'' is a common feature of complex
                systems, such as computing or juggling. Each additional element
                added to a computation, or ball to a cascade, makes performing the
                task more difficult than the last addition.
                ---------------------------------------------------------------------------
                 The FHWA study's estimates of marginal noise costs reflected the
                variation in noise levels resulting from incremental changes in travel
                by autos, light trucks, and other vehicles, and the annoyance and other
                adverse impacts caused by noise. These included adverse impacts on
                pedestrians and residents of the surrounding area, as well as on
                vehicle occupants themselves.
                 To calculate the incremental costs of congestion and noise, the
                agencies multiplied FHWA's ``middle'' estimates of marginal congestion
                and noise costs per mile of auto and light truck travel in urban and
                rural areas by the annual increases in driving attributable to the
                standards to yield increases in total congestion and noise externality
                costs. Because the proposal, and other alternatives that were
                considered, reduced the stringency of CAFE and CO2 standards
                for model years 2021-2026, resulting in lower fuel economy for new cars
                and light trucks produced during those years, the fuel economy rebound
                effect resulted in fewer miles driven relative to the baseline, thus
                generating savings in congestion and noise costs relative to their
                levels under the baseline. Similarly, each of those alternatives also
                reduced the total amount of travel by the used vehicle fleet,
                generating additional savings in these costs.
                (b) Comments on the NPRM Values
                 The agencies received few comments on the estimates of congestion
                and noise costs they used to analyze the economic impacts of the
                proposal. Almost all of these comments focused on the appropriateness
                of the estimated magnitude of the fuel economy rebound effect they used
                to estimate the change in use of new cars and light trucks or the
                plausibility of the reduction in driving by used vehicles, rather than
                to the unit costs estimates themselves. These included comments from
                ICCT and CARB.\1936\
                ---------------------------------------------------------------------------
                 \1936\ ICCT, Comment, NHTSA-2018-0067-11741 at 121; CARB,
                Comment, NHTSA-2018-0067-11873 at 316.
                ---------------------------------------------------------------------------
                 One individual commenter did suggest that recent growth in traffic
                levels, resulting in part from increased use of home delivery services
                for online purchases, has increased congestion and resulting
                delays.\1937\ Although this commenter is correct, traffic growth is not
                strictly a recent phenomenon, and longer-term growth in vehicle use--
                combined with comparatively modest increases in road and highway
                capacity--has contributed to increasing congestion levels. Because
                congestion increases more than proportionately to growing traffic
                volumes, this suggests that FHWA's estimates of congestion costs--now
                more than two decades old--are likely to understate the contribution of
                continuing increases in vehicle use to congestion, resulting delays to
                vehicle occupants and freight shipments, and their associated costs.
                Because noise levels also increase non-linearly with the volume of
                traffic using roads and highways, FHWA's 1997 estimates of marginal
                noise costs may also understate current values.
                ---------------------------------------------------------------------------
                 \1937\ Richard Carriere, NHTSA-2018-0067-12216.
                ---------------------------------------------------------------------------
                (c) Values Used To Analyze the Final Rule
                 The agencies are retaining the same methodology employed in the
                NPRM to estimate congestion and noise costs for the final rule. Like
                other nominal estimates used throughout the analysis, the agencies have
                updated the FHWA estimates to account for current economic and highway
                conditions. The major determinants of marginal congestion costs imposed
                by additional travel include baseline traffic volumes, which determine
                current travel speeds and how they would change in response to further
                increases in travel, together with vehicle occupancy and the value of
                occupants' travel time. These last two factors interact to determine
                the average hourly value of delays to vehicles, which is by far the
                largest component of the total cost of delays that occur under
                congested travel conditions.\1938\ Because travel speeds measure the
                duration of congestion-related delays, while the
                [[Page 24737]]
                value of vehicle occupants' time determines their hourly cost, the
                effects of changes in these variables on overall congestion costs is
                approximately additive, as long as changes in the two are relatively
                modest.
                ---------------------------------------------------------------------------
                 \1938\ Fuel consumption and other operating costs can also
                increase during travel in congested conditions, but their
                relationships to the frequent changes in speed that typically occur
                in congested travel is less well understood, and in any case, they
                vary by far smaller amounts than the value of vehicle occupants'
                travel time.
                ---------------------------------------------------------------------------
                 The agencies approximated the effect of growth in traffic volumes
                on travel speeds and congestion-related delays by increasing congestion
                costs in proportion to the increase in annual vehicle-miles of travel
                per lane-mile on major U.S. highways that occurred between 1997 and
                2017.\1939\ Next, they estimated the increase in the value of travel
                time per vehicle-hour over that same period by combining growth in the
                value of travel time per person-hour--estimated in accordance with DOT
                guidance \1940\--with the increase in average vehicle occupancy by
                persons 16 years of age and older (the same measure of occupancy used
                to estimate the value of refueling time elsewhere in this
                analysis).\1941\ The agencies applied the increases in congestion-
                related delays and the hourly value of travel time to FHWA's 1997
                estimates of marginal congestion costs to update those original values
                to reflect current conditions. The updated values of external
                congestion costs are $0.154 per vehicle-mile of increased travel by
                cars and $0.138 per vehicle-mile for light trucks (expressed in
                constant 2018 dollars), and these values are assumed to remain constant
                throughout the analysis period.
                ---------------------------------------------------------------------------
                 \1939\ Traffic volumes, as measured by the annual number of
                vehicle-miles traveled per lane-mile of roads and highways
                nationwide, rose by 53 percent between 1997 and 2017. Calculated
                from FHWA, Highway Statistics, 1998 and 2018, Tables VM-1 and HM-48,
                available at https://www.fhwa.dot.gov/policyinformation/statistics.cfm.
                 \1940\ See U.S. Department of Transportation, ``Revised
                Departmental Guidance for the Valuation of Travel Time in Economic
                Analysis,'' 2016, at 5-6 and Table 1 at 13.
                 \1941\ The average hourly value of travel time increased by 82
                percent between 1997 and 2017; see U.S. Department of
                Transportation, ``Departmental Guidance for the Valuation of Travel
                Time in Economic Analysis,'' April 9, 1997, Table 4, and U.S.
                Department of Transportation, ``Benefit-Cost Analysis Guidance for
                Discretionary Grant Programs,'' December 2018, Table A-3. From 1995
                to 2017, the average number of light-duty vehicle occupants 16 years
                of age and older increased by 18 percent; values were tabulated from
                FHWA, Nationwide Personal Transportation Survey, 2005 and 2017,
                using online table designer available at https://nhts.ornl.gov/ and
                https://nhts.ornl.gov/index9.shtml.
                ---------------------------------------------------------------------------
                 Similarly, the agencies revised the FHWA estimate of marginal noise
                costs by adjusting for inflation--since the 1994 base year used to
                express values in the FHWA study. Because marginal noise costs are so
                small--less than $0.001 per mile of travel for both cars and light
                trucks--this change did not have a significant impact on the agencies'
                estimates of benefits and costs from the final rule.
                (15) Labor Utilization Assumptions
                 In previous joint CAFE/CO2 rulemakings, the agencies
                considered employment impacts on the automobile manufacturing industry,
                but many of the considerations were qualitative. In the NPRM, the
                agencies presented and took comment on a methodology to quantify
                roughly the direct labor utilization impacts. The agencies recognize
                there is significant uncertainty in any forward-looking
                characterization of labor utilization, including effects resulting from
                CAFE/CO2 rulemakings. Changes to other policies such as
                trade policies and tariff policies are likely substantially to alter
                underlying assumptions presented in the analysis for the rulemaking,
                and these changes could dwarf any differences between policy
                alternatives presented. In this section the agencies discuss the
                assumptions made in the NPRM analysis, summarize comments received on
                that work, and respond to these comments.
                (a) Labor Utilization Baseline (Including Multiplier Effect) and Data
                Description
                 In prior CAFE/CO2 rulemakings, the agencies considered
                an analysis of employment impacts in some form in setting both CAFE and
                tailpipe CO2 emissions standards; NHTSA conducted an
                employment analysis in part to determine whether the standards the
                agency set were economically practicable, that is, whether the
                standards were ``within the financial capability of the industry, but
                not so stringent as to'' lead to ``adverse economic consequences, such
                as a significant loss of jobs or unreasonable elimination of consumer
                choice.'' \1942\ EPA similarly conducted an employment analysis under
                the authority granted to the agency under the Clean Air Act.\1943\ Both
                agencies recognized the uncertainties inherent in estimating employment
                impacts; in fact, both agencies dedicated a substantial amount of
                discussion to uncertainty in employment analyses in the 2012 final rule
                for MYs 2017 and beyond.\1944\ Notwithstanding these uncertainties, by
                imposing costs on new light duty vehicles, CAFE and CO2
                standards can have an impact on the demand for labor. Providing the
                best analysis practicable better informs stakeholders and the public
                about the standards' impact than would omitting any estimates of
                potential labor impacts.
                ---------------------------------------------------------------------------
                 \1942\ 67 FR 77015, 77021 (Dec. 16, 2002).
                 \1943\ See George E. Warren Corp. v. EPA, 159 F.3d 616, 623-24
                (D.C. Cir. 1998) (ordinarily permissible for EPA to consider factors
                not specifically enumerated in the Act).
                 \1944\ See 77 FR 62624, 62952, 63102 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 The NPRM quantified many of the effects that were previously
                qualitatively identified, but not considered. For instance, in the PRIA
                for the 2017-2025 rule EPA identified ``demand effects,'' ``cost
                effects,'' and ``factor shift effects'' as important considerations for
                labor, but the analysis did not attempt to quantify each of these
                effects.\1945\
                ---------------------------------------------------------------------------
                 \1945\ U.S. EPA, ``Regulatory Impact Analysis: Final Rulemaking
                for 2017-2025 Light-Duty Vehicle Greenhouse Gas Emission Standards
                and Corporate Average Fuel Economy Standards,'' at 8-24 to 8-32
                (Aug. 2012).
                ---------------------------------------------------------------------------
                 The NPRM analysis considered direct labor effects on the automotive
                sector. The NPRM evaluated how labor utilization in different facets of
                the automobile manufacturing industry may be affected by the rule,
                including (1) dealership labor related to new light-duty vehicle unit
                sales; (2) assembly labor for vehicles, for engines and for
                transmissions related to new vehicle unit sales; and (3) labor related
                to mandated additional fuel savings technologies, accounting for new
                vehicle unit sales. Importantly, this analysis did not consider whether
                price reductions and regulatory savings associated with different
                standards would, because price reductions would allow consumers to save
                or spend that money on other things of value, increase the consumption
                of other vehicle technologies or, more generally, generate growth in
                other sectors of the overall economy. This means that the analysis is
                inherently and artificially narrow in its focus, and does not represent
                an attempt to quantify the overall labor or economic effects of this
                rulemaking. All labor effects were estimated and reported at a national
                level, in person-years, assuming 2,000 hours of labor per person-
                year.\1946\
                ---------------------------------------------------------------------------
                 \1946\ The agencies recognize a few local production facilities
                may contribute meaningfully to local economies, but the analysis
                reported only on national effects.
                ---------------------------------------------------------------------------
                 The NPRM analysis estimated labor effects from the forecasted CAFE
                model technology costs and from review of automotive labor for the MY
                2016 fleet. For each vehicle in the CAFE model analysis, the locations
                for vehicle assembly, engine assembly, and transmission assembly and
                estimated labor in MY 2016 were recorded. The percent of U.S. content
                for each vehicle was also recorded.\1947\ The analysis also
                [[Page 24738]]
                took into account the portion of parts that are made in the U.S. by
                holding constant the percent of U.S. content for each vehicle as
                manufacturers add fuel-savings technologies. The analysis further
                assumes that the U.S. labor added would be proportional to U.S.
                content, which means that the analysis assumes that U.S. labor inputs
                would remain constant over time, but this does not reflect a prediction
                that U.S. labor inputs actually will remain constant.\1948\ From this
                foundation, the analysis forecasted automotive labor effects as the
                CAFE model added fuel economy technology and adjusted future sales for
                each vehicle.
                ---------------------------------------------------------------------------
                 \1947\ NHTSA provides reports under 49 CFR part 583, ``American
                Automobile Labeling Act Reports'' with information NHTSA received
                from vehicle manufacturers about the U.S./Canadian content (by
                percentage value) of the equipment (parts) used to assemble
                passenger motor vehicles. See https://www.nhtsa.gov/part-583-american-automobile-labeling-act-reports.
                 \1948\ This is a key assumption that should be revisited as
                trade deals and tax or tariff policies materially change.
                ---------------------------------------------------------------------------
                 The NPRM analysis also accounted for sales projections in response
                to the different regulatory alternatives; the labor analysis considers
                changes in new vehicle prices and new vehicle sales (for further
                discussion of the sales model, see Section VI.D.1.b(2)). As vehicle
                prices rise, the analysis expected consumers to purchase fewer vehicles
                than they would have at lower prices.\1949\ As manufacturers sell fewer
                vehicles, the manufacturers may need less labor to produce the vehicles
                and dealers may need less labor to sell the vehicles. However, as
                manufacturers add equipment to each new vehicle, the industry will
                require labor resources to develop, sell, and produce additional fuel-
                saving technologies. The analysis also accounted for the possibility
                that new standards could shift the relative shares of passenger cars
                and light trucks in the overall fleet (see Section VI.D.1.b(2));
                insofar as different vehicles involved different amounts of labor, this
                shifting impacts the quantity of estimated labor. The labor analysis
                took into account the anticipated reduction in vehicle sales, shifts in
                the mix of passenger cars and light trucks, and addition of fuel-
                savings technologies that result from the regulation--and,
                subsequently, the anticipated increase in sales and reduction of fuel-
                savings technologies that are expected to result from a reduction in
                stringency.
                ---------------------------------------------------------------------------
                 \1949\ Many commenters contend that higher prices for more
                efficient goods will have no effect on unit sales and hence
                necessary production resources and employment. The sales aspect of
                labor utilization is addressed in the sales section. NHTSA-2018-
                0067-12000-35, Center for Biological Diversity, et al.
                ---------------------------------------------------------------------------
                 For the NPRM analysis, the agencies assumed that some observations
                about the production of MY 2016 vehicles would carry forward, unchanged
                into the future. For instance, assembly plants would remain the same as
                MY 2016 for all products now, and in the future. The analysis assumed
                the percent of U.S. content would remain constant, even as
                manufacturers updated vehicles and introduced new fuel-saving
                technologies. The analysis further assumed that assembly labor hours
                per unit would remain at estimated MY 2016 levels for vehicles,
                engines, and transmissions, and the factor between direct assembly
                labor and parts production labors would remain the same. When
                considering shifts from one technology to another, the analysis assumed
                revenue per employee at suppliers and original equipment manufacturers
                would remain in line with MY 2016 levels, even as manufacturers added
                fuel-saving technologies and realized cost reductions from learning.
                 The NPRM analysis focused on automotive labor because adjacent
                employment factors and consumer spending factors for other goods and
                services are uncertain and difficult to predict. The analysis did not
                consider how direct labor changes may affect the macro economy and
                possibly change employment in adjacent industries. For instance, the
                analysis did not consider possible labor changes in vehicle maintenance
                and repair, nor did it consider changes in labor at retail gas
                stations. The analysis did not consider possible labor changes due to
                raw material production, such as production of aluminum, steel, copper,
                and lithium, nor did the agencies consider possible labor impacts due
                to changes in production of oil and gas, ethanol, and electricity. The
                analysis did not analyze potential labor effects arising from
                consumption of other products that would not have occurred but for
                improved fuel economy, nor did the analysis assess the effects arising
                from reduced consumption of other products that results from more
                expensive fuel savings technologies at the time of purchase. The
                effects of increased usage of car-sharing, ride-sharing, and automated
                vehicles were not analyzed. The analysis did not estimate how changes
                in labor from any of these industries could affect gross domestic
                product and possibly affect other industries as a result.
                 Many commenters voiced concerns that the NPRM analysis only
                included automotive direct employment, and did not explicitly consider
                other important factors, and that these factors would be better
                addressed with a macroeconomic model. For instance, the International
                Council on Clean Transportation contended that the dollars saved at the
                pump as a result of fuel saving technologies would be spent elsewhere
                in the economy, creating jobs.\1950\ The Association of Global
                Automakers also referenced macroeconomic studies that project long-term
                job gains due to savings at the pump, but also highlight short-term
                setbacks for jobs as money spent to purchase additional fuel saving
                technologies on new vehicles is not spent in other job creating sectors
                of the U.S. economy, which were not considered in an analysis that only
                addresses direct automotive employment.\1951\ The Union of Concerned
                Scientists and Environmental Defense Fund argued that the modeling of
                short-term job losses in the macroeconomic models is incorrect, and
                that purchasing a new vehicle, especially if financed, should increase
                disposable income, because monthly savings at the pump outpace the
                monthly financed cost of the fuel saving equipment, but also that
                consumers will not choose this equipment unless a stringent standard is
                chosen.\1952\ The Institute for Policy Integrity commented that an
                analysis looking only at direct employment is incomplete, and
                encouraged the agencies to include long-term and economy-wide effects
                in scope on employment discussions.\1953\
                ---------------------------------------------------------------------------
                 \1950\ NHTSA-2018-0067-11741-145, ICCT.
                 \1951\ NHTSA-2018-0067-12032-30, Association of Global
                Automakers.
                 \1952\ NHTSA-2018-0067-12039-38, Union of Concerned Scientists;
                NHTSA-2018-0067-12397-4, Environmental Defense Fund, et al.
                 \1953\ NHTSA-2018-0067-12213-66, Institute for Policy Integrity.
                ---------------------------------------------------------------------------
                 The agencies have not quantified employment effects outside of
                automotive sector direct employment for this final rule. The agencies
                agree with commenters that the reductions in production costs of new
                vehicles will free up resources for other productive pursuits. Some
                producers may shift resources away from the development and production
                of fuel saving technologies and into the development and production of
                other vehicle attributes. In this case, there would be a transfer of
                labor resources within a firm. Other producers may instead pass along
                the reduction in production costs to consumers in the form of price
                reductions or avoided price increases, allowing those consumers to
                allocate those new funds between expenditure in other consumption
                categories or savings. The increased expenditure in other consumption
                categories would more efficiently create new employment in sectors
                expanding to cover new market-based (as opposed to regulatory-
                [[Page 24739]]
                based) demand. Increased savings also creates additional investment in
                new productive capital, which will generate employment opportunities in
                the future. However, the extent and nature of these effects are all
                highly uncertain, and the agencies have therefore not quantified the
                effect of the rule on economy-wide employment in the final rule
                analysis.
                 Many commenters expressed concern that America would cede
                leadership in development and production of fuel saving technologies,
                and fuel-saving technology investment would be gutted if augural
                standards were not kept in place. For instance, the Mayor of the City
                of Chillicothe, and Mayors of other Ohio cities, pointed out that many
                light duty vehicles are built in Ohio and neighboring geographies, and
                that workers designing and producing fuel economy equipment make an
                average annual salary of $61,500, expressing concern that if standards
                are lowered, some of these jobs may no longer be necessary.\1954\ The
                BlueGreen Alliance pointed out that over the last twenty years,
                manufacturers have invested billions of dollars into fuel saving
                technologies, and that multinational companies may shift jobs to other
                countries if the standards do not require continued, strong, additional
                investment in even more fuel saving technologies.\1955\
                ---------------------------------------------------------------------------
                 \1954\ NHTSA-2018-0067-12318-2, Mayors of the City of
                Chillicothe and other Ohio cities.
                 \1955\ NHTSA-2018-0067-12009-6, BlueGreen Alliance.
                ---------------------------------------------------------------------------
                 The agencies recognize that development of fuel saving technologies
                can be capital intensive. However, high fuel economy standards do not,
                per se, guarantee multinational companies will invest in American
                research and development or production. For example, the larger percent
                U.S. content in the MY 2017 light truck vs. the MY 2017 passenger car
                new vehicle fleet may be tied to the so-called ``Chicken Tax,'' a long-
                established tariff on the import of light duty trucks.\1956\ On
                average, a light truck in the MY 2017 fleet contained 47.8 percent U.S.
                content, while a passenger car contained 36.0 percent U.S. content. To
                the extent that other policies encourage multi-national corporations to
                build and invest in U.S. production facilities, these organizations
                will need access to capital to do so. Notably, as part of the sales
                module, as fuel economy of the fleet improves, the agencies assume
                customers increasingly choose light trucks, meaning that a shift
                towards light-trucks is already considered in the CAFE model under the
                augural standards.
                ---------------------------------------------------------------------------
                 \1956\ On average, a light truck in the MY 2017 fleet contained
                47.8 percent U.S. content, while a passenger car contained 36.0
                percent U.S. content.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.412
                 Finally, no assumptions were made about part-time-level of
                employment in the broader economy and the availability of human
                resources to fill positions. When the economy is at full employment, a
                fuel economy regulation is unlikely to have much impact on net overall
                U.S. employment; instead, labor would primarily be shifted from one
                sector to another. These shifts in employment impose an opportunity
                cost on society, as regulation diverts workers from other market-based
                activities in the economy. In this situation, any effects on net
                employment are likely to be transitory as workers change jobs (e.g.,
                some workers may need to be retrained or require time to search for new
                jobs, while short-term labor shortages in some sectors or regions could
                result in firms bidding up wages to attract workers). On the other
                hand, if a regulation comes into effect during a period of less-than-
                full employment, a change in labor demand due to regulation would
                affect net overall U.S. employment because the labor market is not in
                equilibrium. Schmalensee and Stavins point out that net positive
                employment effects are possible in the near term when the economy is at
                less than full employment due to the potential hiring of idle labor
                resources by the regulated sector to meet new requirements (e.g., to
                install new equipment) and new economic activity
                [[Page 24740]]
                in sectors related to the regulated sector longer run.\1957\ However,
                the net effect on employment in the long run is more difficult to
                predict and will depend on the way in which the related industries
                respond to regulatory requirements. For that reason, this analysis does
                not include multiplier effects but instead focuses on labor impacts in
                the most directly affected industries, which would face the most
                concentrated labor impacts.
                ---------------------------------------------------------------------------
                 \1957\ Schmalensee, Richard, and Robert N. Stavins. ``A Guide to
                Economic and Policy Analysis of EPA's Transport Rule.'' White paper
                commissioned by Excelon Corporation, March 2011 (Docket EPA-HQ-OAR-
                2010-0799-0676).
                ---------------------------------------------------------------------------
                (b) Estimating Labor for Fuel Economy Technologies, Vehicle Components,
                Final Assembly, and Retailers
                 The following sections discuss the approaches to estimating factors
                related to dealership labor, final assembly labor and parts production,
                and fuel economy technology labor.
                (i) Dealership Labor
                 The NPRM analysis evaluated dealership labor related to new light-
                duty vehicle sales, and estimated the labor hours per new vehicle sold
                at dealerships, including labor from sales, finance, insurance, and
                management. The effect of new car sales on the maintenance, repair, and
                parts department labor is expected to be limited, as this need is based
                on the vehicle miles traveled of the total fleet. To estimate the labor
                hours at dealerships per new vehicle sold, the agencies referenced the
                National Automobile Dealers Association 2016 Annual Report, which
                provides franchise dealer employment by department and function.\1958\
                The analysis estimated that slightly less than 20 percent of dealership
                employees' work relates to new car sales (versus approximately 80
                percent in service, parts, and used car sales), and that on average
                dealership employees working on new vehicle sales labor for 27.8 hours
                per new vehicle sold. The analysis presented today retains assumptions
                about dealership labor hours per vehicle sold.
                ---------------------------------------------------------------------------
                 \1958\ NADA Data 2016: Annual Financial Profile of America's
                Franchised New-Car Dealerships, National Automobile Dealers
                Association, https://www.nada.org/2016NADAdata/ (last visited
                December 20, 2019).
                ---------------------------------------------------------------------------
                (ii) Final Assembly Labor and Parts Production
                 As new vehicle sales increase or decrease, the amount of labor
                required to assemble parts and vehicles changes accordingly. The NPRM
                evaluated how the quantity of assembly labor and parts production labor
                for MY 2016 vehicles would increase or decrease in the future as new
                vehicle unit sales increased or decreased. Specific assembly locations
                for final vehicle assembly, engine assembly, and transmission assembly
                for each MY 2016 vehicle were identified. In some cases, manufacturers
                assembled products in more than one location, and the analysis
                identified such products and considered parallel production in the
                labor analysis.
                 The analysis estimated average direct assembly labor per vehicle
                (30 hours), per engine (four hours), and per transmission (five hours)
                based on a sample of U.S. assembly plant employment and production
                statistics and other publicly available information. The analysis used
                the assembly locations and averages for labor per unit to estimate U.S.
                assembly labor hours for each vehicle. U.S. assembly labor hours per
                vehicle ranged from as high as 39 hours if the manufacturer assembled
                the vehicle, engine, and transmission at U.S. plants, to as low as zero
                hours if the manufacturer imported the vehicle, engine, and
                transmission.
                 The analysis also considered labor for parts production. The
                agencies surveyed motor vehicle and equipment manufacturing labor
                statistics from the U.S. Census Bureau, the Bureau of Labor Statistics,
                and other publicly available sources. The agencies found that the
                historical average ratio of vehicle assembly manufacturing employment
                to employment for total motor vehicle and equipment manufacturing for
                new vehicles was roughly constant over the period from 2001 through
                2013, at a ratio of 5.26.\1959\ Observations from 2001-2013 included
                many combinations of technologies and technology trends, and many
                economic conditions, yet the ratio remained about the same over time.
                Accordingly, the analysis scaled up estimated U.S. assembly labor hours
                by a factor of 5.26 to consider U.S. parts production labor in addition
                to assembly labor for each vehicle. The estimates for vehicle assembly
                labor and parts production labor for each vehicle scaled up or down as
                unit sales scaled up or down over time in the CAFE model.
                ---------------------------------------------------------------------------
                 \1959\ NAICS Code 3361, 3363.
                ---------------------------------------------------------------------------
                 The analysis presented today retains assumptions about coefficients
                for final assembly labor and parts production, and updates production
                and final assembly locations for the MY 2017 fleet. As discussed in
                Section VI.D.1.b(2), today's analysis also applies updated methods for
                estimating the extent to which changes in CAFE and CO2
                standards might lead to changes in quantities of new vehicles sold each
                year. These estimated changes in sales lead to changes in estimated
                changes in domestic employment.
                (iii) Fuel Economy Technology Labor
                 As manufacturers spend additional dollars on fuel-saving
                technologies, parts suppliers and manufacturers require labor to bring
                those technologies to market. Manufacturers may add, shift, or replace
                employees in ways that are difficult for the agencies to predict;
                however, it is expected that the revenue per labor hour at original
                equipment manufacturers (OEMs) and suppliers will remain about the same
                as in MY 2016 even as manufacturers include additional fuel-saving
                technology. To estimate the average revenue per labor hour at OEMs and
                suppliers, the analysis looked at financial reports from publicly
                traded automotive businesses.\1960\ Based on recent figures, it was
                estimated that OEMs would add one labor year per each $633,066
                increment in revenue and that suppliers would add one labor year per
                $247,648 in revenue.\1961\ These global estimates are applied to all
                revenues, and U.S. content is applied as a later adjustment. In today's
                analysis, the agencies assume these ratios would remain constant for
                all technologies rather than that the increased labor costs would be
                shifted toward foreign countries. There are some reasons to believe
                that this may be a conservative assumption. For instance, domestic
                manufacturers may react to increased labor costs by searching for
                lower-cost labor in other countries.
                ---------------------------------------------------------------------------
                 \1960\ The analysis considered suppliers that won the Automotive
                News ``PACE Award'' from 2013-2017, covering more than 40 suppliers,
                more than 30 of which are publicly traded companies. Automotive News
                gives ``PACE Awards'' to innovative manufacturers, with most recent
                winners earning awards for new fuel-savings technologies.
                 \1961\ The analysis assumed incremental OEM revenue as the
                retail price equivalent for technologies, adjusting for changes in
                sales volume. The analysis assumed incremental supplier revenue as
                the technology cost for technologies before retail price equivalent
                mark-up, adjusting for changes in sales volume.
                ---------------------------------------------------------------------------
                 The analysis presented today retains assumptions about coefficients
                for fuel economy technology labor, and updates the percent of U.S.
                content for the MY 2017 fleet.
                (iv) Labor Calculations
                 The agencies estimated the total labor effect as the sum of three
                components: changes to dealership hours, final assembly and parts
                production, and labor for fuel-economy technologies (at OEMs and
                suppliers) that are due to the final rule. The CAFE model calculated
                [[Page 24741]]
                additional labor hours for each vehicle, based on current vehicle
                manufacturing locations and simulation outputs for additional
                technologies, and sales changes. The analysis applied some constants to
                all vehicles.\1962\ Other constants were vehicle specific, for all
                years considered in the analysis.\1963\ Still, other constants were
                year-specific for a vehicle.\1964\ While a multiplier effect of all
                U.S. automotive related labor on non-auto related U.S. jobs was not
                considered for the final rule's analysis, the analysis did incorporate
                a ``global multiplier'' that can be used to scale up or scale down the
                total labor hours. This parameter exists in the parameters file, and
                for the final rule's analysis the analysis set the value at 1.00. The
                results of this analysis can be found in Table VI-201 below.
                ---------------------------------------------------------------------------
                 \1962\ The analysis applied the same assumptions to all
                manufacturers for annual labor hours per employee, dealership hours
                per unit sold, OEM revenue per employee, supplier revenue per
                employee, and factor for the jobs multiplier.
                 \1963\ The analysis made vehicle-specific assumptions about
                percent of U.S. content and U.S. assembly employment hours.
                 \1964\ The analysis estimated technology cost for each vehicle,
                for each year based on the technology content applied in the CAFE
                model, year-by-year.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.413
                 Results of this analysis can be found in Section VII. Considering
                that, all else equal, increases in new vehicle sales lead to increases
                in domestic employment while decreases in technology outlays lead to
                decreases in domestic employment, the agencies estimate that less
                stringent standards could slightly reduce domestic employment. It is
                important to note, however, that the reduction in person-years
                described in this table merely reflects the fact that, when compared to
                the standards set in 2012, fewer jobs will be specifically created to
                meet regulatory requirements that, for other reasons, are not
                economically practicable. It is also important to note that avoided
                outlays for technology can be invested by manufacturers into other
                areas, or passed on to consumers. Moreover, consumers can either take
                those cost savings in the form of a reduced vehicle price, or used
                toward the purchase of specific automotive features that they desire
                (potentially including a more-efficient vehicle), which would increase
                employment among suppliers and manufacturers.
                2. Simulating Safety Impacts of Regulatory Alternatives
                 The primary objectives of CAFE and CO2 standards are to
                achieve maximum feasible fuel economy and reduce CO2
                emissions, respectively, from the light-duty vehicle fleet. In setting
                standards to achieve these intended effects, the potential of the
                standards to affect vehicle safety is also considered. As a safety
                agency, NHTSA has long considered the potential for adverse safety
                consequences when establishing CAFE standards, and under the CAA, EPA
                considers factors related to public health and human welfare, including
                safety, in regulating emissions of air pollutants from mobile sources.
                 Safety trade-offs associated with increases in fuel economy
                standards have occurred in the past--particularly before CAFE standards
                became attribute-based--because manufacturers chose to comply with
                stricter standards by building smaller and lighter vehicles. In cases
                where fuel economy improvements were achieved through reductions in
                vehicle size and mass, the smaller, lighter vehicles did not protect
                their occupants as effectively in crashes as larger, heavier vehicles,
                on average. Although the agencies now use attribute-based standards, in
                part to reduce the incentive to downsize vehicles to comply with CAFE
                and CO2 standards, the agencies must continue to be mindful
                of the possibility of safety-related trade-offs.
                 Although prior analyses acknowledged that CAFE and CO2
                standards could influence factors that affect safety other than vehicle
                mass, those impacts were not estimated quantitatively.\1965\ Instead,
                the agencies focused exclusively on the safety impacts of changes in
                vehicle mass. In the proposal, the safety analysis was expanded to
                include a broader and more comprehensive measure of safety impacts. The
                final rule retains this comprehensive approach and analyzes the safety
                impact of three factors:
                ---------------------------------------------------------------------------
                 \1965\ The agencies included a quantification of rebound-
                associated safety impacts in its Draft TAR analysis, but because the
                scrappage model is new for this rulemaking, did not include safety
                impacts associated with the effect of standards on new vehicle
                prices and thus on fleet turnover. The fact that the scrappage model
                did not exist prior to this rulemaking does not mean that the
                effects that it aims to show were not important considerations,
                simply that the agencies were unable to account for them
                quantitatively prior to the current rulemaking.
                ---------------------------------------------------------------------------
                 (1)Changes in Vehicle Mass. Similar to previous analyses, the
                agencies calculate the safety impact of changes in vehicle mass made to
                reduce fuel consumption and comply with the standards. The agencies'
                statistical analysis of historical crash data indicates reducing mass
                in heavier vehicles generally improves safety, while reducing mass in
                lighter vehicles generally reduces safety. NHTSA's crash simulation
                modeling of vehicle design
                [[Page 24742]]
                concepts for reducing mass revealed similar effects.
                 (2)Impacts of Vehicle Prices. Vehicles have become safer over time
                through a combination of new safety regulations and voluntary safety
                improvements. The agencies expect this trend to continue as emerging
                technologies, such as advanced driver assistance systems, are
                incorporated into new vehicles. Safety improvements will likely
                continue regardless of changes to CAFE standards. However, the pace of
                such improvements may be modified if manufacturers choose to delay or
                forgo investments in safety technology because of the demands that
                complying with stricter CAFE and CO2 standards impose on
                scarce research, development, and manufacturing resources.
                 As discussed in Section VI.D.1.b), technologies added to comply
                with fuel economy standards have an impact on vehicle prices, and, by
                extension, on the affordability of newer, safer vehicles, and therefore
                on the rates at which newer vehicles are acquired and older, less safe
                vehicles are retired from use. The delays in fleet turnover caused by
                the effect of new vehicle prices on sales and scrappage rates affect
                safety, by slowing the penetration of new safety technologies into the
                fleet.
                 The standards also influence the composition of the light-duty
                fleet. As the safety provided by light trucks, SUVs and passenger cars
                responds differently to technology that manufacturers employ to meet
                the standards--particularly mass reduction--fleets with different
                compositions of body styles will have varying numbers of fatalities, so
                changing the share of each type of light-duty vehicle in the projected
                future fleet impacts safety outcomes.
                 (3)Increased driving because of better fuel economy.The ``rebound
                effect'' predicts consumers will drive more when the cost of driving
                declines. More stringent standards reduce vehicle operating costs, and
                in response, some consumers may choose to drive more. Additional
                driving increases exposure to risks associated with motor vehicle
                travel, and this added exposure translates into higher fatalities and
                injuries.
                 We measure the impact of these factors as differences in fatalities
                across the alternatives. Fatalities are calculated by deriving a fleet-
                wide fatality rate (fatalities per vehicle mile of travel)
                incorporating the different factors and multiplying it by the
                alternative's expected VMT. Fatalities are converted into a societal
                cost by multiplying fatalities with the DOT-recommended value of a
                statistical life (VSL). As with the NPRM, traffic injuries and property
                damage are not modeled directly; \1966\ rather, traffic injuries and
                property damage continue to be estimated using adjustment factors that
                reflect the observed relationship between societal costs of fatalities
                and costs of injuries and property damage.
                ---------------------------------------------------------------------------
                 \1966\ The agencies noted in the NPRM that traffic injuries and
                property damage are not directly modeled because of insufficient
                data. See PRIA at 43108.
                ---------------------------------------------------------------------------
                 All three factors influence predicted fatalities, but only two of
                them--changes in vehicle mass and in the composition of the light-duty
                fleet in response to changes in vehicle prices--impose increased risks
                on drivers and passengers that are not compensated for by accompanying
                benefits. In contrast, increased driving associated with the rebound
                effect is a consumer choice that reveals the benefit of additional
                travel. Consumers who choose to drive more have apparently concluded
                that the utility of additional driving exceeds the additional costs for
                doing so--including the crash risk that they perceive additional
                driving involves. As discussed in Section VI.D.2.c), the agencies
                account for the benefits of rebound driving by offsetting a portion of
                the added safety costs.
                 Some commenters argued that the agencies should be measuring the
                change in the fatality rate rather than the change in the number of
                fatalities. For example, EDF argued that changes in fatalities was a
                measurement of VMT and number of passengers rather than safety, and
                that ``NHTSA's job is to decrease the fatality rate per mile, not to
                decrease the number of miles people drive.'' \1967\ EDF also commented
                that the agencies were required to report the ``fatality rate data for
                the overall safety impacts.'' The agencies disagree with EDF. The
                agencies are responsible for measuring the impacts of fuel economy and
                CO2 standards, including changes to VMT. While other NHTSA
                safety rules have minimal impacts upon aggregate VMT, CAFE standards
                have a large impact on VMT and VMT-related costs, including fatalities.
                ---------------------------------------------------------------------------
                 \1967\ EDF, Appendix A, NHTSA-2018-0067-12108, at 7-9.
                ---------------------------------------------------------------------------
                 Although NHTSA often uses changes in fatality rates as a metric to
                evaluate the impact of regulations on safety, these rates are just a
                tool utilized to derive the relevant safety impact--namely the
                estimated change in fatalities. Furthermore, as part of the cost-
                benefit analysis required by Executive Order 12866 and specified in OMB
                Circular A-4, the agencies must quantify and value safety impacts to
                compare them to the costs of the regulation. The fundamental metric for
                valuing loss of life is the VSL. To apply this metric, the agencies
                must first produce estimates of any change in the number of fatalities
                that results from the regulatory action. Fatalities prevented, as well
                as other safety impacts such as non-fatal injuries prevented and
                property damage crashes avoided, are appropriate measures of rules that
                affect motor vehicle safety.
                (a) Impact of Weight Reduction on Safety
                 Vehicle mass reduction can be one of the more cost-effective means
                of increasing fuel economy and reducing CO2 emissions to
                meet standards--particularly for makes and models not already built
                with much high strength steel or aluminum closures or low mass
                components. Manufacturers have stated that they will continue to reduce
                vehicle mass to meet more stringent standards, and therefore, this
                expectation is incorporated into the modeling analysis supporting the
                standards. Safety trade-offs associated with mass-reduction have
                occurred in the past, particularly before CAFE standards were
                attribute-based; past safety trade-offs may have occurred because
                manufacturers chose at the time, in response to CAFE standards, to
                build smaller and lighter vehicles. In cases where fuel economy
                improvements were achieved through reductions in vehicle size and mass,
                the smaller, lighter vehicles did not fare as well in crashes as
                larger, heavier vehicles, on average. Although the agencies now use
                attribute-based standards, in part to reduce or eliminate the incentive
                to downsize vehicles to comply with CAFE and CO2
                standards,\1968\ the agencies must be mindful of the possibility of
                related safety trade-offs.
                ---------------------------------------------------------------------------
                 \1968\ CAFE and CO2 standards are ``footprint-
                based,'' with footprint being defined as a measure of a vehicle's
                size, roughly equal to the wheelbase times the average of the front
                and rear track widths. Footprint-based standards create a
                disincentive for manufacturers to produce smaller-footprint
                vehicles. This is because, as footprint decreases, the corresponding
                fuel economy/CO2 emission target becomes more stringent.
                We also believe that the shape of the footprint curves themselves is
                such that the curves should neither encourage manufacturers to
                increase nor decrease the footprint of their fleets.
                ---------------------------------------------------------------------------
                 Historically, as shown in FARS data analyzed by the agencies, mass
                reduction concentrated among the heaviest vehicles (chiefly, the
                largest LTVs, CUVs and minivans) is estimated to reduce overall
                fatalities, while mass reduction concentrated among the lightest
                vehicles (chiefly, smaller passenger cars) is estimated to increase
                [[Page 24743]]
                overall fatalities. Mass reduction in heavier vehicles is more
                beneficial to the occupants of lighter vehicles than it is harmful to
                the occupants of the heavier vehicles. Mass reduction in lighter
                vehicles is more harmful to the occupants of lighter vehicles than it
                is beneficial to the occupants of the heavier vehicles. In response to
                questions of whether designs and materials of more recent model year
                vehicles may have weakened the historical statistical relationships
                between mass, size, and safety, the agencies updated our public
                database for statistical analysis consisting of crash data. The
                analysis considered the full range of real-world crash types.
                 The methodology used for the statistical analysis of historical
                crash data has evolved over many years. The methodology used for the
                NPRM and unchanged for the final rule reflects learnings and
                refinements from: NHTSA studies in 2003, 2010, 2011, 2012, and 2016;
                independent peer review of 23 studies by the University of Michigan
                Transportation Research Institute;\1969\ two public workshops hosted by
                NHTSA;\1970\ interagency collaboration among NHTSA, DOE and EPA; and
                comments to CAFE and CO2 rulemakings in 2010, 2012, the 2016
                Draft TAR, and the 2018 NPRM. As explained in greater detail below, the
                methodology used for the statistical analysis of historical crash data
                for the NPRM and final rule is the best and most up to date available.
                ---------------------------------------------------------------------------
                 \1969\ Green, Paul E., Kostyniuk, Lidia P., Gordon, Timothy J.,
                and Reed, Matthew P., Independent Review of Statistical Analyses of
                Relationship between Vehicle Curb Weight, Track Width, Wheelbase and
                Fatality Rates, UMTRI-2011-12, University of Michigan of
                Transportation Research Institute (2011). Available at http://www.umtri.umich.edu/our-results/publications/independent-review-statistical-analyses-relationship-between-vehicle-curb.
                 \1970\ The workshops were held on February 25, 2011 and May 13-
                14, 2013. Video, transcripts, and presentations are available on the
                NHTSA website (recommended search terms include ``workshop,''
                ``mass,'' ``safety,'' and the dates of the workshops).
                ---------------------------------------------------------------------------
                 Additionally, to assess whether future vehicle designs may impact
                the relationship of vehicle mass reduction on safety, NHTSA sponsored a
                fleet crash simulation study using future mass reduction vehicle design
                concepts (see Fleet Simulation Study below). The results of the
                simulation research showed that future mass reduction techniques
                continue to exhibit impacts on safety and were consistent with the
                statistical analysis of FARS crash data. The agencies considered the
                findings of the study and concluded it was reasonable and appropriate
                to continue to consider the impact of mass reduction on safety for
                future vehicles because the data indicate the relationship between mass
                and safety will continue in the future.
                 For the rulemaking analysis, the CAFE model tracks the amount of
                mass reduction applied to each vehicle model, and then applies
                estimated changes in societal fatality risk per 100 pounds of mass
                reduction determined through the statistical analysis of FARS crash
                data. This process allows the CAFE model to tally changes in fatalities
                attributed to mass reduction across all of the analyzed future model
                years. In turn, the CAFE model is able to provide an overall impact of
                the final standards and alternatives on fatalities attributed to mass
                reduction.
                 A number of comments were received on technical aspects of the
                mass-safety analysis in the NPRM. The agencies carefully considered all
                comments. Where warranted, the agencies conducted additional analyses
                to determine whether commenters' suggestions would improve the
                analysis. The agencies found that the methodology employed by the
                proposal, which was developed over many years, subject to extensive
                review and feedback, remains the most rigorous methodology. The
                agencies found the alternative approaches raised in comments would
                provide less likely estimates, were statistically problematic, or, in
                some cases, advocated discarding or ignoring the most likely estimates
                altogether. The agencies' assessments of comments are discussed in
                detail in the subsections below.
                 Overall, consistent with prior analyses, the data show that mass
                reduction concentrated in heavier vehicles is generally beneficial to
                overall safety, and mass reduction concentrated in lighter vehicles is
                harmful.
                (1) Crash Data
                 The agencies use real-world crash data as the basis for projecting
                the future safety implications for regulatory changes. To support the
                2012 rulemaking, NHTSA created a common, updated database for
                statistical analysis consisting of crash data. The initial iteration
                contained crash data for model years 2000-2007 vehicles in calendar
                years 2002-2008. NHTSA made the preliminary version of the new
                database, which was the basis for NHTSA's 2011 preliminary report
                (hereinafter 2011 Kahane report),\1971\ available to the public in May
                2011, and an updated version in April 2012 (used in NHTSA's 2012 final
                report, hereinafter 2012 Kahane report), \1972\ enabling other
                researchers to analyze the same data and, hopefully, minimize
                discrepancies in results caused by reporting inconsistencies across
                databases.\1973\ NHTSA updated the crash and exposure databases for the
                2016 Draft TAR analysis.
                ---------------------------------------------------------------------------
                 \1971\ Kahane, C, J. Relationships Between Fatality Risk, Mass,
                and Footprint in Model Year 2000-2007 Passenger Cars and LTVs--Final
                Report, National Highway Traffic Safety Administration (Aug. 2012).
                Available at https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/811665.
                 \1972\ Kahane, C, J. Relationships Between Fatality Risk, Mass,
                and Footprint in Model Year 2000-2007 Passenger Cars and LTVs--
                Preliminary Report. Docket No. NHTSA-2010-0152-0023. Washington, DC:
                National Highway Traffic Safety Administration.
                 \1973\ See 75 FR 25324, 25395-96 (May 7, 2010).
                ---------------------------------------------------------------------------
                 For the proposed rule and unchanged for today's final rule, the
                crash and exposure databases were updated again. The databases are the
                most up-to-date possible (MY 2004-2011 vehicles in CY 2006-2012), given
                the processing time for crash data and the need for enough crash cases
                to permit statistically meaningful analyses. As in previous analyses,
                NHTSA has made the new databases available to the public on its
                website.\1974\
                ---------------------------------------------------------------------------
                 \1974\ ftp://ftp.nhtsa.dot.gov/CAFE/2018_mass_size_safety/.
                ---------------------------------------------------------------------------
                (2) Methodology
                 The relationship between a vehicle's mass, size, and fatality risk
                is complex, and it varies in different types of crashes. The agencies
                have been examining this relationship for more than two decades. The
                basic analytical method used to analyze the impacts of weight reduction
                on safety for the proposal, and unchanged for this final rulemaking, is
                the same as in 2016 Puckett and Kindelberger report.\1975\ NHTSA
                released the 2016 Puckett and Kindelberger report as a preliminary
                report on the relationship between fatality risk, mass, and footprint
                in June 2016 in advance of the Draft TAR. The 2016 Puckett and
                Kindelberger report covered the same scope as previous NHTSA
                reports,\1976\ offering a detailed description of the crash and
                exposure databases, modeling approach, and analytical results on
                relationships among vehicle size, mass, and fatalities that informed
                the Draft TAR. The
                [[Page 24744]]
                modeling approach described in the 2016 Puckett and Kindelberger report
                was developed with the collaborative input of NHTSA, EPA and DOE, and
                subject to extensive public review, scrutiny in two NHTSA-sponsored
                workshops, and a thorough peer review that compared it with the
                methodologies used in other studies.\1977\
                ---------------------------------------------------------------------------
                 \1975\ Puckett, S.M. and Kindelberger, J.C. (2016, June).
                Relationships between Fatality Risk, Mass, and Footprint in Model
                Year 2003-2010 Passenger Cars and LTVs--Preliminary Report. (Docket
                No. NHTSA-2016-0068). Washington, DC: National Highway Traffic
                Safety Administration, available at https://www.nhtsa.gov/sites/nhtsa.dot.gov/files/2016-prelim-relationship-fatalityrisk-mass-footprint-2003-10.pdf.
                 \1976\ The 2016 Puckett and Kindelberger report is an extension
                of 2011 Kahane report and 2012 Kahane report.
                 \1977\ Previous reports from which the 2016 Puckett and
                Kindelberger report was derived from, were also subject to extensive
                peer reviews. Farmer, Green, and Lie, who reviewed the 2010 Kahane
                report, also peer-reviewed the 2011 Kahane report. In preparing his
                2012 report (along with the 2016 Puckett and Kindelberger report and
                Draft TAR), Kahane also took into account Wenzel's assessment of the
                preliminary report and its peer reviews, DRI's analyses published
                early in 2012, and public comments such as the International Council
                on Clean Transportation's comments submitted on NHTSA and EPA's 2010
                notice of joint rulemaking. These comments prompted supplementary
                analyses, especially sensitivity tests, discussed at the end of this
                section.
                ---------------------------------------------------------------------------
                 In computing the impact of changes in mass on safety, the agencies
                are faced with competing challenges. Research has consistently shown
                that mass reduction affects ``lighter'' and ``heavier'' vehicles
                differently across crash types. The 2016 Puckett and Kindelberger
                report found mass reduction concentrated amongst the heaviest vehicles
                is likely to have a beneficial effect on overall societal fatalities,
                while mass reduction concentrated among the lightest vehicles is likely
                to have a detrimental effect on fatalities.\1978\ To accurately capture
                the differing effect on lighter and heavier vehicles, the agencies must
                split vehicles into lighter and heavier vehicle classifications in the
                analysis.\1979\ However, this poses a challenge of creating
                statistically-meaningful results. There is limited relevant crash data
                to use for the analysis. Each partition of the data reduces the number
                of observations per vehicle classification and crash type, and thus
                reduces the statistical robustness of the results. The methodology
                employed by the agencies was designed to balance these competing forces
                as an optimal trade-off to accurately capture the impact of mass-
                reduction across vehicle curb weights and crash types while preserving
                the potential to identify robust estimates.
                ---------------------------------------------------------------------------
                 \1978\ The findings of the 2016 Puckett and Kindelberger report
                are consistent with the results of the 2012 Kahane report and Draft
                TAR.
                 \1979\ If lighter and heavier vehicles are left undistinguished,
                the agencies analysis would be restricted to identifying a single
                effect of mass reduction for passenger cars and a single effect of
                mass reduction for truck-based LTVs. As discussed below, distinct
                effects have been estimated historically for lighter versus heavier
                vehicles for cars and LTVs, confirming the validity of
                distinguishing by curb weight where feasible.
                ---------------------------------------------------------------------------
                 For the proposal and the final rule, the agencies employed the
                modeling technique developed in the 2016 Puckett and Kindelberger
                report to analyze the updated crash and exposure data by examining the
                cross sections of the societal fatality rate per billion vehicle miles
                of travel (VMT) by mass and footprint, while controlling for driver
                age, gender, and other factors, in separate logistic regressions for
                five vehicle groups and nine crash types. ``Societal'' fatality rates
                include fatalities to occupants of all the vehicles involved in the
                collisions, plus any pedestrians, cyclists, or occupants of other
                conveyances (e.g., motorcyclists). The agencies utilize the
                relationships between weight and safety from this analysis, expressed
                as percentage increases in fatalities per 100-pound weight reduction,
                to examine the weight impacts applied in this CAFE analysis. The
                effects of mass reduction on safety were estimated relative to
                (incremental to) the regulatory baseline (augural standards) in the
                CAFE analysis, across all vehicles for MYs 2018 and beyond.
                 As in the 2012 Kahane report, 2016 Puckett and Kindelberger report,
                and the Draft TAR, the vehicles are grouped into three classes:
                Passenger cars (including both two-door and four-door cars); CUVs and
                minivans; and truck-based LTVs. The curb weight of passenger cars is
                formulated, as in the 2012 Kahane report, 2016 Puckett and Kindelberger
                report, and Draft TAR, as a two-piece linear variable to estimate one
                effect of mass reduction in the lighter cars and another effect in the
                heavier cars. The boundary between ``lighter'' and ``heavier'' cars is
                3,201 pounds (which is the median mass of MY 2004-2011 cars in fatal
                crashes in CY 2006-2012, up from 3,106 pounds for MY 2000-2007 cars in
                CY 2002-2008 in the 2012 NHTSA safety database, and up from 3,197
                pounds for MY 2003-2010 cars in CY 2005-2011 in the 2016 NHTSA safety
                database). Likewise, for truck-based LTVs, curb weight is a two-piece
                linear variable with the boundary at 5,014 pounds (again, the MY 2004-
                2011 median, higher than the median of 4,594 pounds for MY 2000-2007
                LTVs in CY 2002-2008 and the median of 4,947 pounds for MY 2003-2010
                LTVs in CY 2005-2011). CUVs and minivans are grouped together in a
                single group covering all curb weights of those vehicles; as a result,
                curb weight is formulated as a simple linear variable for CUVs and
                minivans. Historically, CUVs and minivans have accounted for a
                relatively small share of new-vehicle sales over the range of the data,
                resulting in less crash data available than for cars or truck-based
                LTVs. In sum, vehicles are distributed into five groups by class and
                curb weights: Passenger cars < 3,201 pounds; passenger cars 3,201
                pounds or greater; truck-based LTVs < 5,014 pounds; truck-based LTVs
                5,014 pounds or greater; and all CUVs and minivans.
                 There are nine types of crashes specified in the analysis for each
                vehicle group: three types of single-vehicle crashes, five types of
                two-vehicle crashes; and one classification of all other crashes.
                Single-vehicle crashes include first-event rollovers, collisions with
                fixed objects, and collisions with pedestrians, bicycles and
                motorcycles. Two-vehicle crashes include collisions with: heavy-duty
                vehicles; cars, CUVs, or minivans < 3,187 pounds (the median curb
                weight of other, non-case, cars, CUVs and minivans in fatal crashes in
                the database); cars, CUVs, or minivans >= 3,187 pounds; truck-based
                LTVs < 4,360 pounds (the median curb weight of other truck-based LTVs
                in fatal crashes in the database); and truck-based LTVs >= 4,360
                pounds. Grouping partner-vehicle CUVs and minivans with cars rather
                than LTVs is more appropriate because their front-end profile and
                rigidity more closely resemble a car than a typical truck-based LTV. An
                additional crash type includes all other fatal crash types (e.g.,
                collisions involving more than two vehicles, animals, or trains).
                Splitting the vehicles from this crash type involved in crashes
                involving two light-duty vehicles into a lighter and a heavier group
                permits more accurate analyses of the mass effect in collisions of two
                vehicles.
                 For a given vehicle class and weight range (if applicable),
                regression coefficients for mass (while holding footprint constant) in
                the nine types of crashes are averaged, weighted by the number of
                baseline fatalities that would have occurred for the subgroup MY 2008-
                2011 vehicles in CY 2008-2012 if these vehicles had all been equipped
                with electronic stability control (ESC). The adjustment for ESC, a
                feature of the analysis added in 2012, takes into account results will
                be used to analyze effects of mass reduction in future vehicles, which
                will all be ESC-equipped, as required by NHTSA's safety regulations.
                 The agencies received multiple comments on how they distribute
                vehicles into classifications. IPI, quoting a study by Tom Wenzel,
                commented that sorting vehicles into footprint deciles shows positive
                impacts from mass reduction for the majority of the
                [[Page 24745]]
                footprint deciles.\1980\ CARB commented that the agencies should have
                used the curb weight of all vehicles to calculate the thresholds for
                ``lighter'' and ``heavier'' vehicle types rather than just the curb
                weights of vehicles involved in fatal crashes.\1981\ CARB also
                commented that pickup trucks and SUVs that are not subject to CAFE
                regulation (i.e., Class 2b and Class 3 vehicles, such as \3/4\-ton and
                one-ton pick-up trucks, vans and related SUVs) should not be included
                in the assessment of the impact of mass on safety and doing so raises
                the median weight of trucks.\1982\ CARB also commented that the median
                weights are static values representing the historical fleet, but the
                median weights and proportions of crash types involving given vehicle
                weight categories should change with median weight of the fleet modeled
                by the CAFE model.\1983\ Commenters generally believed that the
                agencies' approach ``results in inappropriate apportioning of cars and
                trucks into the corresponding lighter or heavier bins,'' which in turn
                causes the agencies to overestimate the fatalities associated with mass
                reduction.\1984\
                ---------------------------------------------------------------------------
                 \1980\ IPI, Detailed Comments, Docket No. NHTSA-2018-0067-12213,
                at 127 (quoting Tom Wenzel, Assessment of NHTSA's Report
                ``Relationships Between Fatality Risk, Mass, and Footprint in Model
                Year 2004-2011 Passenger Cars and LTVs,'' (LBNL Phase 1, 2018).
                Available at https://escholarship.org/uc/item/4726g6jq.
                 \1981\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
                11873, at 276.
                 \1982\ Tom Wenzel of Lawrence Berkeley National Laboratories,
                Comment, EPA-HQ-OAR-2018-0283-4118, at 1; see also CARB, Detailed
                Comments, Docket No. NHTSA-2018-0067-11873, at 259.
                 \1983\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
                11873, at 260.
                 \1984\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
                11873, at 276.
                ---------------------------------------------------------------------------
                 Dividing vehicles into footprint deciles and excluding Class 2b and
                3 vehicles pose sample size and data coverage issues. If vehicles were
                grouped into footprint deciles, the sample sizes in each decile would
                be approximately one-fifth as large as the corresponding sample sizes
                in each of the agencies' four passenger car and LTV vehicle classes
                (and one-tenth as large as the sample size for CUVs and minivans).
                Smaller parameter estimates require correspondingly smaller standard
                errors (i.e., relatively precise estimates) to achieve statistical
                significance, but splitting the limited data into deciles yields larger
                standard errors, restricting the ability to identify statistically-
                significant estimates. Likewise, by extending the footprint-curb
                weight-fatality data to include Class 2b and 3 trucks that are
                functionally and structurally similar to corresponding \1/2\-ton models
                that are subject to CAFE regulation,\1985\ the sample size and ranges
                of curb weights and footprint are improved. Sample size is a challenge
                for estimating relationships between curb weight and fatality risk for
                individual crash types in the main analysis; dividing the sample
                further or removing observations makes it exceedingly difficulty to
                identify meaningful estimates and the relationships that are present in
                the data.
                ---------------------------------------------------------------------------
                 \1985\ Class 2b and 3 pickup trucks, vans and SUVs have physical
                characteristics and usage profiles that are substantially similar to
                their Class 2a counterparts. For example, the Class 2a version of
                the Ford F-150 has similar physical characteristics to and has a
                similar usage profile to the Class 2b Ford F150. Same for the Class
                2a Ford F150 relative to the Class 2b and 3 Ford F250, and for the
                GMC Yukon relative to the Yukon XL. The Class 2b and 3 pickup trucks
                in the sample generally have gross vehicle weight ratings of 10,000
                pounds or less, and thus are subject to the same Federal motor
                vehicle safety standards as their light-duty counterparts. Likewise,
                these vehicles generally have similar physical dimensions (e.g.,
                ground clearance, width) as related light-duty vehicles. Key
                differentiating factors among these vehicles are height, payload,
                and towing capacity. There are likely to be unobserved differences
                in how these vehicles are driven relative to light-duty
                alternatives; however, the crash data include a census of fatal
                crashes involving case vehicles and the Class 2b and 3 vehicles
                included in the analysis, in turn representing the relative risk of
                differences in curb weight in crashes involving Class 2b and 3
                vehicles. Despite being regulated by different fuel economy and
                emissions regulations as they become heavier (i.e., once a given
                model crosses a mass threshold changes classes), the vehicles may
                continue to be used in similar ways over time; in turn, the safety
                implications of the presence of these vehicles may continue to be
                similar. In contrast, other types of heavy-duty vehicles, such as
                box trucks, buses, refuse trucks, fire trucks, and other heavy-duty
                commercial vehicles are substantially different from light duty
                vehicles in their physical characteristics and usage profiles, and
                it would not be appropriate to include them in the statistical
                analysis to determine the impact of mass on crash fatalities.
                ---------------------------------------------------------------------------
                 Compounding the issue is the fact the analysis focuses on societal
                fatality risk (i.e., all fatalities, including crash partners and
                people outside of vehicles, such as pedestrians, cyclists, and
                motorcyclists) rather than merely in-vehicle fatality risk, which
                yields estimates that are smaller in magnitude (and thus more difficult
                to identify meaningful differences from zero) than estimates
                representing changes in in-vehicle fatality risk. That is, compared to
                an analysis of in-vehicle fatality risk (which would tend to yield
                relatively large estimated effects of mass reduction), the focus on
                societal fatalities tends to yield relatively small (net) effects of
                mass reduction on fatality risk.
                 Including Class 2b and 3 vehicles in the analysis to determine the
                relationship of vehicle mass on safety has the added benefit of
                improving correlation constraints. Notably, curb weight increases
                faster than footprint for large light trucks and Class 2b and 3 pickup
                trucks and SUVs, in part because the widths of vehicles are constrained
                more tightly (i.e., due to lane widths) than their curb weights.
                Including data from Class 2b and 3 pick-up truck and SUV fatal crashes
                provides data over a wider range of vehicle weights, which improves the
                ability to estimate the mass-crash fatality relationship. The agencies
                believe the decision of whether to include Class 2b and 3 vehicles in
                the analysis should be made based on whether the additional data
                improves the estimate of the safety impact of mass reduction in light
                trucks, and that the fatality data should not be simplistically
                excluded because the vehicles are not regulated under the CAFE and
                CO2 emissions programs. Ultimately, the agencies find that:
                (1) The fundamental objective is to capture the strongest, meaningful
                signal regarding societal fatality risk as a function of the mass of
                light trucks; (2) that incorporating information on fatal incidents
                involving Class 2b and 3 trucks improves the quality of the signal the
                agencies can capture, and (3) including the vehicles provides the best
                estimate of the impacts of mass on societal fatalities.
                 In assessing whether to calculate the median curb weight threshold
                from all vehicles involved in accidents or on the road, the agencies
                weighed changing the process used to establish the thresholds and the
                potential impact on the robustness of the statistical analysis. From a
                statistical perspective, using thresholds that allocate a similar
                number of fatal crash cases to both the lower vehicle weight group and
                the higher vehicle weight group for a given vehicle type will minimize
                the average standard errors of estimates for both groups, which
                provides the best estimates for each group. Because reducing average
                standard errors strengthens the statistical analysis, the agencies
                conclude using only the curb weight of vehicles involved in fatal
                crashes to calculate the median curb weight threshold produces the best
                estimate. This conclusion is the same that was reached previously when
                considering the same issue for the 2011 Kahane, 2012 Kahane, and 2016
                Puckett and Kindelberger analyses.
                 On a related note, the regression models are estimated based on
                with respect to the total number of fatalities associated within each
                vehicle weight group classification (referred to as vehicle group
                below, for brevity). Shifting the threshold would change the estimated
                incremental impact of changes in curb weight in each vehicle
                [[Page 24746]]
                group, but the net effects would offset each other across vehicle
                groups, resulting in the same overall estimated effect of changes in
                vehicle mass on societal fatality risk. For example, if one restricted
                the ``lightest'' group for a vehicle type to include only the bottom
                ten percentiles of vehicle weight, one would expect to identify a very
                strong detrimental effect (or weakest beneficial effect) of mass
                reduction for that group. However, the estimated effect of mass
                reduction in that group has minimal implications for the fleet (i.e.,
                because there are fewer vehicles in the group), and the corresponding
                estimated effect of mass reduction for other groups would also mute the
                impact (i.e., because there are many vehicles in the group that vary in
                mass to a much larger degree than in the ``lighter'' group).
                Ultimately, the mean effect of mass reduction across the lighter and
                heavier groups would be the same as when using the median as the
                threshold (or at least, similar, subject to limitations in statistical
                optimization), but with a different point of reference when comparing
                the groups. Thus, the agencies believe the selection of curb weight
                threshold has a minimal impact on the estimated effects of mass
                reduction across all vehicle types.
                 Full consideration of CARB's comment on mass thresholds, and
                whether they should change as the median weight of the fleet modeled by
                the CAFE model changes, requires a deeper look at each of the crash
                types considered in the analysis. That is, the point estimates
                presented in Table VI-202 represent weighted averages across nine
                separate, mutually-exclusive and exhaustive crash models (analyzed
                separately for cars, LTVs, and CUVs and minivans). For example, an
                individual model for first-event rollovers yields estimates of the
                percentage change in societal fatality risk per 100-pound mass
                reduction for lighter and heavier (or, in the case of CUVs and
                minivans, all) vehicles in the target vehicle class. The final, overall
                point estimate for a given vehicle type is found by: (1) Multiplying
                the estimate associated with an individual crash type by the estimated
                share of societal fatalities involving the vehicle class (adjusting for
                two-vehicle collisions that span vehicle classes to avoid double-
                counting); and (2) summing the values estimated in (1) across all crash
                types. In its comments, CARB noted that if the distribution of vehicles
                in terms of curb weight changes through lightweighting, the shares of
                (fatal) two-vehicle crashes involving a given pair of vehicles as
                defined by weight class (e.g., car below a given threshold colliding
                with a LTV above a given threshold) would change. In turn, the
                appropriate weighting across the crash types modeled in the analysis
                would likewise be different (involving an increasing share of vehicles
                below a given curb weight threshold). Due to these potential
                limitations, CARB questioned the stability of the summary point
                estimates relative to changes in the shares of fatalities within each
                crash type in the analysis.\1986\
                ---------------------------------------------------------------------------
                 \1986\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
                11873, at 278-79.
                ---------------------------------------------------------------------------
                 To evaluate CARB's concerns regarding future crash mixes and
                definitions of vehicle weight classes, the agencies performed an
                exploratory analysis examining the scope and impacts of potential model
                changes. In doing so, the agencies examined the degree of change in the
                median vehicle fleet weight in the NPRM analysis relative to the fixed
                mass threshold values, and also how sensitive the curb weight safety
                point estimates are to assumptions about the distribution of curb
                weights in future vehicle fleets. The agencies also considered the
                feasibility of changing the shares of fatalities by crash type as a
                function of forthcoming or developing vehicle safety technologies. This
                information would help inform adjustments to fatality rate impacts for
                each vehicle type, because the likelihood of observing individual fatal
                crash types could change in different ways across vehicle types in the
                analysis as the vehicle mix changes. However, the agencies identified
                no studies on the effectiveness of forthcoming or developing vehicle
                safety technologies that could inform projections of shares of
                fatalities across crash types, nor did the commenters reference any
                such studies. Likewise, commenters provided no data that would enable
                projections of these factors. Thus, for a given vehicle mix, the
                agencies have no information available to justify changing the shares
                of fatalities across crash types over time. Therefore, the agencies
                decided to keep the distribution of fatality shares constant for:
                First-event rollovers; fixed-object collisions; collisions with
                pedestrians, bicyclists, and motorcycles; collisions with heavy
                vehicles; collisions with one other light-duty vehicle (i.e., a
                constant share across the sum of these crashes, but not constant for
                any given type of crash partner); and all other crashes.
                 The agencies had sufficient information to evaluate the effects of
                changes in the fatal crash mix for cases involving two light-duty
                vehicles. The agencies agreed that it was internally consistent to
                adjust fatality shares by crash type proportionally to the distribution
                of vehicle types and curb weight classes for a given focal MY. An
                important technical question associated with this approach is the level
                of disaggregation. The agencies considered an alternative in which the
                agencies would estimate and apply unique curb weight point estimates
                for each calendar year in the analysis for each regulatory alternative.
                This alternative would account for changes in the distribution of crash
                types associated with changes in both vehicle type shares (i.e., shifts
                from passenger cars to CUVs and LTVs) and vehicle mass shares (i.e.,
                shifts from vehicles above the curb weight thresholds to vehicles below
                the thresholds). As in the status quo analysis of curb weight and
                fatality risk, the resulting point estimates would be weighted averages
                across the individual crash type models as presented in the NPRM, but
                re-weighted to reflect projected changes to the fleet.
                 The agencies investigated this alternative and identified several
                concerns. A key functional constraint is that the curb weight safety
                point estimates are applied in the CAFE model as a lump-sum, lifetime
                effect to a given vehicle. This characteristic of the model limits the
                ability to apply calendar-year-specific effects of changes in curb
                weight and vehicle type distributions when evaluating safety impacts of
                changes in curb weights. The safety point estimates also represent net
                effects of changes in curb weights over the lifetime of a given vehicle
                in the CAFE model; any changes in the calculation of safety point
                estimates would need to preserve this characteristic. More broadly, the
                vehicle fleet is not static over a vehicle's lifetime (i.e., the
                distributions of curb weight and vehicle type change each year), so the
                effective probabilities of each crash type over a given vehicle's
                lifetime are a function of many calendar-year-level curb weight and
                vehicle type distributions. To capture any effects of changes in
                vehicle mass distributions over time within the current CAFE model
                structure, the agencies would need to enact a method that: (1)
                Identifies defensible changes in fatality risk associated with vehicle
                mass as the distribution of vehicle mass changes (e.g., accounting for
                changes in the likelihood of observing particular fatal crash types
                that reflect projected changes in the distribution of vehicle types and
                curb weights across vehicles); and (2) allocates calendar-year-specific
                impacts of curb weight on fatality risk to each vehicle in the fleet
                across the
                [[Page 24747]]
                analysis horizon. Identifying how best to achieve this would be
                complex, and would require the development of an alternative analytical
                approach that would be outside the scope of this rulemaking.
                 With these concerns in mind, the agencies explored an alternative
                approach to test the sensitivity of the safety point estimates to
                distributions of vehicles by curb weight and vehicle type. The starting
                point for the alternative approach is maintaining the understanding
                that the nine crash type models that are present in the curb weight
                safety analysis represent the best statistical alternatives for
                evaluating the crash data in the database (i.e., optimal statistical
                precision conditional on the coverage of the data). Furthermore, the
                nine crash type models are defined in terms of physical relationships
                (i.e., crashes involving vehicles of particular curb weight ranges and
                vehicle types) that are invariant to changes in the distributions of
                vehicles for those same characteristics. That is, the estimated changes
                in societal fatality risk as curb weights change for a focal vehicle
                (i.e., of a particular type and weight range) that is involved in a
                particular type of crash apply equally to any scenario involving such
                vehicle, regardless of changes in the probability of observing such a
                scenario. For example, the agencies would expect the societal fatality
                risk for a crash involving a passenger car lighter than 3,201 pounds
                colliding with a LTV heavier than 4,360 pounds to be the same
                regardless of how many such collisions take place. Thus, the agencies
                would expect the net effect of a given change in curb weight for a
                given vehicle type in a given crash type to scale proportionally with
                the probability of such crashes occurring. Put simply, if there are
                half as many potential crash partners of a given type in a future year
                compared to a base year, the agencies would expect a given curb weight
                reduction to have half as large of a net effect on fatalities in the
                future year relative to the base year. In the extreme, curb weight
                changes would have no net effect on fatalities at all for a given crash
                type if such crashes had a zero percent probability of occurring (i.e.,
                if there are no potential crash partner vehicles).
                 Based on this maintained hypothesis, the agencies examined test
                curb weight safety point estimates under alternative scenarios, in
                which fatality shares by crash type were proportional to the
                distribution of vehicle types and curb weight classes across a range of
                outcomes reflecting different model years and policy alternatives
                represented in the NPRM. The sensitivities of the safety point
                estimates to changes in the distributions of vehicle curb weights and
                vehicle types were tested by adjusting fatality shares across the
                relevant crash types in the analysis (i.e., involving two light-duty
                vehicles) in a manner consistent with potential changes in the vehicle
                fleet, while holding the outputs of the individual crash type models
                the same as in the NPRM.
                 For example, compare the safety point estimate for LTVs lighter
                than 5,014 pounds in the NPRM with an alternative point estimate for an
                extreme hypothetical future year where 80 percent of the LTV fleet is
                lighter than the median curb weight for crash partners (4,360 pounds):
                [GRAPHIC] [TIFF OMITTED] TR30AP20.414
                [[Page 24748]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.415
                 The estimated net societal effect of a 100-pound mass reduction is
                equal to: (1) The sum of the estimated net effects across all crash
                types, divided by (2) the baseline estimate of annual fatalities
                involving the vehicle class (adjusted to avoid double-counting) for the
                most recent four MYs in the database (MYs 2008-2011), or 1,782
                fatalities per year. In the NPRM, the estimated net societal effect of
                a 100-pound mass reduction for lighter LTVs was a 5.5 fatality
                increase, or a 0.31 percent increase relative to a baseline of 1,782
                fatalities. Changing the share of crash fatalities involving heavier
                LTVs to be consistent with a fleet with only 20 percent of LTVs above
                the curb weight threshold yields: (1) An increase in incremental
                fatalities in crashes involving lighter LTVs (from 0.5 fatality to 0.7
                fatality); and (2) a decrease in incremental fatalities in crashes
                involving heavier LTVs (from 1.5 fatalities to 0.7 fatality); for a
                total net increment of 4.9 fatalities compared to the NPRM's estimate
                of 5.5 fatalities. Thus, the agencies estimate that, in a future year
                where the fleet differs from the baseline by having an extreme case of
                80 percent of LTVs below the crash-partner curb weight threshold, the
                net societal effect of a 100-pound mass reduction in LTVs lighter than
                5,014 pounds would be 4.9 divided by 1,782, or 0.28 percent, versus
                0.31 percent in the baseline.
                 This simple example confirms that the estimates do indeed change as
                the distribution of curb weights changes. In this case, the change is
                intuitive: As the LTV fleet becomes lighter, mass reduction among LTVs
                below 5,014 pounds becomes less detrimental to society. However, the
                incremental effect is estimated to be quite small: Shifting from an
                even mix of LTVs above and below the threshold to an extreme 20 percent
                to 80 percent split only changes the estimated net societal effect by
                0.03 percent in absolute terms. Thus, the model results for lighter
                LTVs appear relatively insensitive to the LTV curb weight distribution.
                Indeed, in the limit, where all LTVs are below the crash-partner curb
                weight threshold (and thus there are no fatality impacts for crashes
                involving heavier LTVs), the estimated net societal effect of a 100-
                pound mass reduction for LTVs below 5,014 pounds (i.e., all LTVs in
                this case) is 0.25 percent, a difference of 0.06 percent in absolute
                terms compared to the baseline. This result is driven by the dominating
                effects of crash types involving either: (1) No crash partner (e.g.,
                first-event rollovers); (2) one crash partner from a group not
                associated with a given change in a curb weight distribution (e.g.,
                heavy vehicles, bicyclists, passenger cars); or (3) multiple crash
                partners (an element of ``all other crashes''). That is, even extreme
                changes in the distribution of curb weights for a given vehicle type
                will not change the role that vehicle mass plays in crashes for a focal
                vehicle when that vehicle does not collide with another vehicle from
                the distribution in question. In the above example involving lighter
                LTVs, 90 percent of fatalities involve incidents that do not include a
                single LTV crash partner, and 66 percent of fatalities involve
                incidents that do not include a single light-duty crash partner.
                 Continuing with this example scenario, the point estimate for LTVs
                heavier than 5,014 pounds becomes larger in magnitude (i.e., more
                societally beneficial mass reduction) to a similar degree as the
                reduction in magnitude for lighter LTVs when moving to an extreme 20
                percent to 80 percent split of crash partner LTVs above (versus below
                in the case above) the curb weight threshold:
                [[Page 24749]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.416
                 In the NPRM and this analysis, the estimated net societal effect of
                a 100-pound mass reduction for lighter LTVs was a 20.0 fatality
                decrease, or a 0.61 percent decrease relative to a baseline of 3,304
                fatalities. Changing the share of crash fatalities involving heavier
                LTVs to be consistent with a fleet with only 20 percent of LTVs above
                the curb weight threshold yields: (1) A larger reduction in fatalities
                in crashes involving lighter LTVs per 100-pound mass reduction (from
                4.0 fatalities to 6.1 fatalities); and (2) a decrease in incremental
                fatalities in crashes involving heavier LTVs (from 1.6 fatalities to
                0.7 fatality); for a total net change of -22.9 fatalities compared to a
                baseline of -20.0 fatalities. Thus, the agencies estimate that, in a
                future year where the fleet differs from the baseline by having 80
                percent of LTVs below the crash-partner curb weight threshold, the net
                societal effect of a 100-pound mass reduction in LTVs 5,014 pounds or
                heavier would be -22.9 divided by 3,304, or -0.69 percent, versus -0.61
                percent in the baseline. Consistent with the test results for lighter
                LTVs, the model results for heavier LTVs appear relatively insensitive
                to the LTV curb weight distribution. In the limit, where all LTVs
                (except for one remaining heavier LTV in consideration) are below the
                crash-partner curb weight threshold (and thus there are no effective
                fatality impacts for crashes involving heavier LTVs), the estimated net
                societal effect of a 100-pound mass reduction for the remaining LTV
                above 5,014 pounds is -0.76 percent, a difference of 0.15 percent in
                absolute terms compared to the baseline.
                 Expanding the analysis to account for changes in the relative sales
                shares of each vehicle type dampens the net effects further. As the
                fleet share of passenger cars decreases, the net effects of mass
                reduction among LTVs become less societally beneficial. That is, as
                there are fewer relatively vulnerable passenger cars in the fleet,
                there become fewer opportunities to reduce fatalities in collisions
                between LTVs and passenger cars through mass reduction. In some
                scenarios considered in the exploratory analysis, the effects of sales
                shifts from passenger cars to LTVs at least fully offset the estimated
                improvements in net fatalities associated with mass reduction
                summarized above as the LTV fleet becomes lighter.
                 Ultimately, the exploratory analysis using extreme example cases
                confirmed that the baseline safety point estimates are very reasonable
                for the feasible ranges of mixes of vehicle types and curb weights
                across the model years in the CAFE model analysis. The sensitivities of
                the point estimates are relatively low across relative shares of
                lighter versus heavier LTVs (especially relative to the uncertainty in
                the baseline estimates), and similarly low and offsetting across
                decreasing fleet shares for passenger cars. Because shifts in mass in
                the rulemaking analysis would have insignificant impacts on the safety
                estimated values and therefore rulemaking decision making, the agencies
                conclude no changes are warranted for this final rule analysis.
                Mass Safety Results
                 Table VI-204 presents the estimated percent increase in U.S.
                societal fatality risk per 10 billion VMT for each 100-pound reduction
                in vehicle mass, while holding footprint constant, for each of the five
                vehicle classes:
                [[Page 24750]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.417
                 Techniques developed in the 2011 (preliminary) and 2012 (final)
                Kahane reports have been retained to test statistical significance and
                to estimate 95-percent confidence bounds (sampling error) for mass
                effects and to estimate the combined annual effect of removing 100
                pounds of mass from every vehicle (or of removing different amounts of
                mass from the various classes of vehicles), while holding footprint
                constant.
                 None of the estimated effects have 95-percent confidence bounds
                that exclude zero, and thus are not statistically significant at the
                95-percent confidence level. The NPRM reported that two estimated
                effects are statistically significant at the 85-percent level. Societal
                fatality risk is estimated to: (1) Increase by 1.2 percent if mass is
                reduced by 100 pounds in the lighter cars; and (2) decrease by 0.61
                percent if mass is reduced by 100 pounds in the heavier truck-based
                LTVs. The estimated increases in societal fatality risk for mass
                reduction in the heavier cars and the lighter truck-based LTVs, and the
                estimated decrease in societal fatality risk for mass reduction in CUVs
                and minivans are not significant, even at the 85-percent confidence
                level. Although 85-percent statistical significance is not a
                traditional metric of meaningful differences to zero, this result
                confirms that the estimated effects for vehicles with curb weights most
                dissimilar to the median vehicle are the most likely to be
                significantly different to zero.
                 The agencies judge the central value estimates are the best and
                most up-to-date estimates available; the estimates offer a stronger
                statistical representation of relationships among vehicle curb weight,
                footprint and fatality risk than an assumption of no correlation
                whatsoever. The agencies appropriately present the statistical
                uncertainty. For example, the central values for the highest vehicle
                weight group (LTVs 5,014 pounds or heavier) and the lowest vehicle
                weight group (passenger cars lighter than 3,201 pounds) (which, based
                on fundamental physics, are expected to have the greatest impact of
                mass reduction on safety) are economically significant,\1987\ and are
                in line with the prior analyses used in past NHTSA CAFE and EPA
                CO2 rulemakings. As shown in Table VI-205, the estimated
                coefficients have trended to lower numerical values in successive
                studies, but remain positive for lighter cars and negative for heavier
                LTVs. The 85-percent confidence level was reported only to show the
                scope of uncertainty at the first rounded (to five percent) threshold
                where the coefficient estimates were significantly different to zero
                for the two vehicle groups at the extremes of the curb weight
                distribution. No preference was suggested for an 85-percent confidence
                bound. Rather, the agencies found value in reporting confidence
                intervals for all five coefficients at the threshold where the
                estimates for the two extremes of the curb weight distribution were
                significantly different to zero. The agencies determined it was better
                to include the estimates, despite the slightly lower confidence level,
                than knowingly omitting economically significant results.
                ---------------------------------------------------------------------------
                 \1987\ The agencies use ``economically significant results'' to
                mean values that have an important, practical implication, but may
                be derived from estimates that do not meet traditional levels of
                statistical significance. For example, if the projected economic
                benefit of a project equaled $100 billion, the agencies would
                consider the impact economically significant, even if the estimates
                used to derive the impact were not statistically significant at the
                95-percent confidence level. Conversely, if the projected economic
                benefit of a project equaled $1, the agencies would not consider the
                impact economically significant, even if the estimates used to
                derive the impact were statistically significant at the 99.99-
                percent confidence level. In the case above, we considered the
                results associated with the lightest and heaviest vehicle types to
                be economically significant because the associated safety costs were
                large and the estimates had magnitudes meaningfully different from
                zero and were statistical significant at the 85-percent confidence
                level.
                ---------------------------------------------------------------------------
                 The regression results are constructed to project the effect of
                changes in mass, independent of all other factors, including footprint.
                With each additional change from the current environment (e.g., the
                scale of mass change, presence and prevalence of safety features,
                demographic characteristics), the results may become less
                representative. That is, although safety features and demographic
                factors are accounted for separately, the estimated effects of mass are
                identified under the specific mix of vehicles and drivers in the data.
                The agencies note that the analysis accounts for safety features that
                are optional but available across all MYs in the sample (most notably
                electronic stability control, which was not yet mandatory for all model
                years in the sample), and calibrates historical safety data to account
                for future fleets with full ESC penetration to reflect the mandate.
                 The agencies considered the near multicollinearity of mass and
                footprint to be a major issue in the 2010 Kahane report and voiced
                concern about inaccurately estimated regression coefficients. High
                correlations between mass and footprint and variance inflation factors
                (VIF) have persisted from MY 1991-1999 to MY 2004-2011; large footprint
                vehicles continued to be, on the average, heavier than small footprint
                vehicles to the same extent as in the previous decade.
                 Nevertheless, multicollinearity appears to have become less of a
                problem in the 2012 Kahane, 2016 Puckett and Kindelberger/Draft TAR,
                and current analyses. Ultimately, only three of the 27 core models of
                fatality risk by vehicle type in the current analysis indicate the
                potential presence of effects of multicollinearity, with estimated
                effects of mass and footprint
                [[Page 24751]]
                reduction greater than two percent per 100-pound mass reduction and
                one-square-foot footprint reduction, respectively; these three models
                include passenger cars and CUVs in first-event rollovers, and CUVs in
                collisions with LTVs greater than 4,360 pounds. This result is
                consistent with the 2016 Puckett and Kindelberger report, which also
                found only three cases out of 27 models with estimated effects of mass
                and footprint reduction greater than two percent per 100-pound mass
                reduction and one-square-foot footprint reduction.
                 For comparison, Table VI-205 shows the fatality coefficients from
                the 2012 Kahane report (MY 2000-2007 vehicles in CY 2002-2008) and the
                2016 Puckett and Kindelberger report and Draft TAR (MY 2003-2010
                vehicles in CY 2005-2011).
                [GRAPHIC] [TIFF OMITTED] TR30AP20.418
                 The new results are directionally the same as in 2012; in the 2016
                analysis, the estimate for lighter LTVs was of opposite sign (but small
                magnitude). Consistent with the 2012 Kahane and 2016 Puckett and
                Kindelberger reports, mass reductions in lighter cars are estimated to
                lead to increases in fatalities, and mass reductions in heavier LTVs
                are estimated to lead to decreases in fatalities.
                ---------------------------------------------------------------------------
                 \1988\ Median curb weights in the 2012 Kahane report: 3,106
                pounds for cars, 4,594 pounds for truck-based LTVs. Median curb
                weights in the 2016 Puckett and Kindelberger report: 3,197 pounds
                for cars, 4,947 pounds for truck-based LTVs.
                ---------------------------------------------------------------------------
                 The estimated mass effect for heavier truck-based LTVs is stronger
                in this analysis and in the 2016 Puckett and Kindelberger report than
                in the 2012 Kahane report; both estimates are statistically significant
                at the 85-percent confidence level, unlike the corresponding estimate
                in the 2012 Kahane report. The estimated mass effect for lighter truck-
                based LTVs is insignificant and positive in this analysis and the 2012
                Kahane report, while the corresponding estimate in the 2016 Puckett and
                Kindelberger report was insignificant and negative.
                 Multiple commenters, including the South Coast Air Quality
                Management District and States and Cities, challenged the practical
                value of using estimates with statistical significance at the 85-
                percent level, arguing that below 95 (or 90) percent are insufficiently
                reliable.\1989\ For example, CARB stated, ``[d]ue to the lack of
                statistical significance, NHTSA should not be attributing any increase
                in fatalities due to mass reduction'' and argues that the ``effect of
                mass reduction on fatality risk should be set to zero since the
                estimates are not statistically different to zero.'' \1990\
                ---------------------------------------------------------------------------
                 \1989\ See South Coast Air Quality Management District, Detailed
                Comments, Docket No. NHTSA-2018-0067-11813, at 6 (internal citation
                omitted); States and Cities, Detailed Comments, Docket No. NHTSA-
                2018-0067-11735, at 95.
                 \1990\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
                11873, at 269.
                ---------------------------------------------------------------------------
                 The agencies believe the updated analysis that was presented in the
                NPRM represents the most up to date and best estimate of the impacts of
                mass reduction on crash fatalities; and, that it is appropriate for the
                analysis to use the best and most likely estimates for safety, even if
                the estimates are not statistically significant at the 95-percent
                confidence level. Significance at the 85-percent confidence level is
                important evidence that the relevant point estimates are meaningfully
                different to zero (e.g., approximately five to six times more likely to
                be non-zero than zero). The agencies believe it would be misleading to
                ignore these data or to use values of zero for the rulemaking analysis,
                as doing so would not properly inform decision makers on the safety
                impacts of the regulatory alternatives and final standards. Similar to
                past analyses, the NPRM and this final rule analysis use the best
                available estimates. The agencies feel it is inappropriate to ignore
                likely impacts of the standards simply because the best available
                estimates have confidence levels below 95 percent; uniform estimates of
                zero are statistically weaker than the estimates identified in the
                analysis, and thus are not the best available. Because the point
                estimates are derived from the best-fitting estimates for each crash
                type (all of which are non-zero), the confidence bounds around an
                overall estimate of zero would necessarily be larger than the
                corresponding confidence bounds around the point estimates presented
                here.
                 The sensitivity analysis in Section VII.C Sensitivity Analysis
                provides an evaluation of extreme cases in which all of the estimated
                net fatality rate impacts of mass reduction are either at their fifth-
                or 95th-percentile values. The range of net impacts in the sensitivity
                analysis not only covers the relatively more likely case that
                uncertain, yet
                [[Page 24752]]
                generally offsetting, effects are distinct from the central estimates
                considered here (e.g., in a plausible case where mass reduction in the
                heaviest LTVs is less beneficial than indicated by the central
                estimates, it would also be relatively likely that mass reduction in
                the lightest passenger cars would be less harmful, yielding a similar
                net impact), but also covers the relatively unlikely case that all of
                the estimates are uncertain in the same direction.
                 At a broader level, multiple commenters asserted that the role of
                safety-related estimates should be restricted because of what they
                claim is a weak historical relationship between fuel economy and
                vehicle safety. For example, the Green Energy Institute at Lewis &
                Clark Law School commented, ``[o]ver the past 40 years, per-capita
                vehicle fatalities decreased by 50%, while average fuel economy
                doubled.'' \1991\ However, this statistic is misleading because it does
                not account for vehicle safety factors and changes in driving behavior
                external to fuel economy (e.g., FMVSS and other safe design advances,
                reductions in drunk driving, increases in seat belt use). That is,
                fatality rates have decreased due to a range of factors that are
                unrelated to fuel economy efforts. The methodology in the 2012 Kahane
                report, the 2016 Puckett and Kindelberger, the Draft TAR, the 2018 NPRM
                analysis and today's final rule analysis addresses these other changes
                in order to isolate the impacts of mass reduction alone. The role of
                the safety analysis outlined in this document is to isolate incremental
                effects on safety outcomes that are related to changes in fuel economy.
                ---------------------------------------------------------------------------
                 \1991\ Green Energy Institute at Lewis & Clark Law School,
                Docket No. NHTSA-2018-0067-12213, at 3.
                ---------------------------------------------------------------------------
                 Multiple commenters disagreed with the results in Table VI-204,
                maintaining that mass reduction need not reduce societal safety. EDF
                cited a Michigan Manufacturing Technology Center (MMTC) review as
                supporting that widespread lightweighting would decrease crash severity
                through reduced kinetic energy in multiple-vehicle crashes. Similarly,
                the Aluminum Association commented, ``[v]ehicle size, not weight, has
                been shown to be the leading safety determinant.'' \1992\ Other
                commenters cited Anderson and Auffhammer (2014), which finds that the
                safety effects of mass reduction in one vehicle are offset by the
                safety effects in the crash partner vehicle.\1993\ The South Coast Air
                Quality Management District asserted that NHTSA and EPA appear to argue
                ``that fuel-efficient vehicles are lighter than other vehicles, and
                therefore, less safe.'' The North Carolina Department of Environmental
                Quality asserted that a takeaway from the preferred alternative is that
                larger vehicles are safer than smaller vehicles. The agencies'
                conclusion is that, at the societal level, it is the distribution of
                changes in vehicle mass that matter (i.e., mitigating mass reduction in
                the lightest vehicles is societally beneficial, while mitigating mass
                reduction in the heaviest vehicles is societally harmful).
                ---------------------------------------------------------------------------
                 \1992\ The Aluminum Association, Detailed Comments, Docket No.
                NHTSA-2018-0067-12213, at 3.
                 \1993\ Anderson, M.L. and M. Auffhammer (2014). ``Pounds that
                Kill,'' Review of Economic Studies, Vol. 81, No. 2, pp. 535-71.
                ---------------------------------------------------------------------------
                 The 2012 Kahane report, the 2016 Puckett and Kindelberger, the
                Draft TAR, the 2018 NPRM analysis and today's final rule analysis all
                have shown that both mass and vehicle size impact societal safety.
                Across recent rulemakings, the analyses have confirmed a protective
                effect of vehicle size (i.e., societal fatality risk decreases as
                footprint increases). As mentioned previously, the agencies believe
                vehicle footprint-based standards help to discourage vehicle
                manufacturers from downsizing their vehicles, and therefore assume
                changes in CAFE and CO2 standards will not impact vehicle
                size and size-related safety impacts. On the other hand, mass reduction
                is a cost-effective technology for increasing fuel economy and reducing
                CO2 emissions. Therefore, the agencies do include the
                assessment of safety impacts related to mass reduction. As discussed
                throughout this mass-safety subsection, the agencies present
                comprehensive consideration of the various studies and workshops on the
                impact of vehicle mass on safety, and conclude there is in fact a
                relationship. The fleet simulation study, discussed in the next
                subsection, further supports the existence of this relationship and
                that this relationship will continue to exist in future vehicle
                designs.
                 The principal difference between heavier vehicles, especially
                truck-based LTVs, and lighter vehicles, especially passenger cars, is
                that mass reduction has a different effect in collisions with another
                car, LTV, or other object such as a lamp post. When two vehicles of
                unequal mass collide, the change in velocity (delta-V) is greater in
                the lighter vehicle. Through conservation of momentum, the degree to
                which the delta-V in the lighter vehicle is greater than in the heavier
                vehicle is proportional to the ratio of mass in the heavier vehicle to
                mass in the lighter vehicle:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.419
                [GRAPHIC] [TIFF OMITTED] TR30AP20.420
                 Because fatality risk is a positive function of delta-V, the
                fatality risk in the lighter vehicle in two-vehicle collisions is also
                higher. Vehicle design can reduce the magnitude of delta-V to some
                degree (e.g., changing the stiffness
                [[Page 24753]]
                of a vehicle's structure could dampen delta-V for both crash partners).
                These considerations drive the overall result: mass reduction is
                associated with an increase in fatality risk in lighter cars, a
                decrease in fatality risk in heavier LTVs, CUVs, and minivans, and has
                smaller effects in the intermediate groups. Mass reduction may also be
                harmful in a crash with a movable object such as a small tree, which
                may break if hit by a high mass vehicle resulting in a lower delta-V
                than may occur if hit by a lower mass vehicle which does not break the
                tree and therefore has a higher delta-V. However, in some types of
                crashes not involving collisions between cars and LTVs, especially
                first-event rollovers and impacts with fixed objects, mass reduction
                may not be harmful and may even be beneficial.
                 Ultimately, delta-V is a direct function of relative vehicle mass
                for given vehicle structures. Removing some mass from the heavier
                vehicle involved in an accident with a lighter vehicle reduces the
                delta-V in the lighter vehicle, where fatality risk is higher,
                resulting in a large benefit to the passengers of the lighter vehicle.
                This is partially offset by a small increase in the delta-V in the
                heavy vehicle; however, the fatality risk is lower in the heavier
                vehicle and remains relatively low despite the increase in delta-V. In
                sum, the change in mass and delta-V from mass reduction in heavier
                vehicles results in a net societal benefit.
                 Multiple commenters claimed that the agencies' analysis does not
                allow for the likely outcome that mass reduction would be concentrated
                among relatively heavy vehicles.\1994\ For example, Global Automakers
                commented that the agencies should not include weight reduction in
                their safety analysis because ``very few vehicles [have] implemented
                lightweight material substitution strategies.'' \1995\
                ---------------------------------------------------------------------------
                 \1994\ See also, e.g., South Coast Air Quality Management
                District, Detailed Comments, Docket No. NHTSA-2018-0067-11813, at 6.
                 \1995\ Association of Global Automakers, Attachment A, Docket
                No. NHTSA-2018-0067-12032, at A-32.
                ---------------------------------------------------------------------------
                 Neither CAFE standards nor this analysis mandate mass reduction, or
                mandate mass reduction occur in any specific manner. However, mass
                reduction is a highly cost effective technology for improving fuel
                economy and CO2 emissions. The steel, aluminum, plastics,
                composite, and other material industries are developing new materials
                and manufacturing equipment and facilities to produce those materials.
                In addition, suppliers and manufacturers are optimizing designs to
                maintain or improve functional performance with lower mass.
                Manufacturers have stated that they will continue to reduce vehicle
                mass to meet more stringent standards, and therefore, this expectation
                is incorporated into the modeling analysis supporting the standards to:
                (1) Determine capabilities of manufacturers; and (2) to predict costs
                and fuel consumption effects of CAFE standards. The CAFE and
                CO2 rulemakings in 2012, the Draft TAR and EPA Preliminary
                Determination, imposed an artificial constraint on vehicle mass
                reduction to achieve a desired safety-neutral outcome. For the current
                rulemaking, this artificial constraint is eliminated so the analysis
                reflects manufacturers applying the most cost effective technologies to
                achieve compliance with the regulatory alternatives and the final
                standards; this approach allows mass reduction to be applied across the
                fleet. This is consistent with industry trends.\1996\ To the extent
                that mass reduction is only cost-effective for the heaviest vehicles,
                the CAFE model would create the outcome predicted by commenters. In
                reality, however, mass reduction is a cost-effective means of improving
                fuel economy and does take place across vehicles of all sizes and
                weights. Accordingly, the model reflects that manufacturers may reduce
                vehicle mass--regardless of vehicle class--when doing so is cost
                effective.
                ---------------------------------------------------------------------------
                 \1996\ The baseline MY 2016 (for the NPRM) and MY 2017 (for this
                final rule analysis) vehicle fleet data show manufacturers have in
                fact implemented mass reduction technology across vehicle types and
                sizes- including smaller and lighter vehicles.
                ---------------------------------------------------------------------------
                 The National Tribal Air Association claimed the 2015 NAS study
                found ``evidence suggest[ing] that the [2012] standards will lead the
                nation's light-duty vehicle fleet to become lighter but not less
                safe.'' \1997\ The agencies note the NAS quote is one phrase from the
                press release that accompanied the NHTSA sponsored 2015 NAS
                study,\1998\ and the agencies do not believe the phrase in isolation
                reflects the findings of the NAS Committee, which are discussed in over
                3 pages of the report.\1999\ The 2015 NAS report supported the
                analytical methodology used for the 2012 NHTSA CAFE and EPA
                CO2 rulemaking and found it reasonable. As discussed in the
                subsections further above, a nearly identical methodology was used for
                the NPRM analysis and for this final rule.
                ---------------------------------------------------------------------------
                 \1997\ National Tribal Air Association, Detailed Comments,
                Docket No. NHTSA-2018-0067-11948, at 2.
                 \1998\ NAS (2015). Press Release. ``Analysis Used by Federal
                Agencies to Set Fuel Economy and Greenhouse Gas Standards for U.S.
                Cars Was Generally of High Quality; Some Technologies and Issues
                Should Be Re-examined.'' June 18, 2015. Available at http://www8.nationalacademies.org/onpinews/newsitem.aspx?RecordID=21744.
                 \1999\ Key excerpts from the report include: ``[o]ccupants of
                smaller vehicles are at a greater risk of fatality in crashes,
                particularly in a crash with a vehicle of greater mass;'' and
                ``[t]he 2012 studies (by NHTSA, Lawrence Berkeley National
                Laboratories, and Dynamic Research, Inc.) indicate that mass
                reduction while holding footprint constant is associated with a
                small increase in risk for lighter-than-average cars only; the
                estimated effect on other vehicle types is not statistically
                significant.'' National Research Council (2015). Cost,
                Effectiveness, and Deployment of Fuel Economy Technologies for
                Light-Duty Vehicles, available at https://doi.org/10.17226/21744.
                pp. 224-28.
                ---------------------------------------------------------------------------
                 The agencies received several comments about the relationship
                between mass and crash avoidance. The NRDC commented that the analysis
                should account for the expected result that mass reduction makes it
                easier to avoid crashes.\2000\ Conversely, IPI quoted a finding by LNL
                that ``found that mass reductions may increase the number of accidents
                but that each crash results in fewer fatalities.'' \2001\
                ---------------------------------------------------------------------------
                 \2000\ NRDC, Detailed Comments, Docket No. NHTSA-2018-0067-
                11973.
                 \2001\ IPI, Detailed Comments, Docket No. NHTSA-2018-0067-12213,
                at 129.
                ---------------------------------------------------------------------------
                 The phenomenon touched upon by IPI and NRDC has been identified in
                past rulemakings as well, and highlights that the relationship between
                mass reduction and societal fatality risk include two partially-
                offsetting components (i.e., increased exposure to crashes is offset
                partially by decreased risk in some vehicles conditional on a crash
                occurring). The agencies note that this relationship, while not
                reported separately, is in fact embedded within the analysis detailed
                in this document, as the extent to which some vehicles are more
                maneuverable and faster-braking, the crash data reflect those
                characteristics through lower observed fatality rates. However, when
                considering the purposes of estimating effects of mass reduction on
                fatalities, it is immaterial what share of the effect is comprised of
                crash avoidance factors and crashworthiness factors, the ultimate
                effect is present within the data evaluated in the analysis. The mass-
                safety impacts estimated by the statistical analysis of crash data are
                based on the safety technologies and mass levels present among the
                vehicle fleets for the calendar and model years in the data. As
                discussed below in this section, the analysis separately accounts for
                the effects of future safety technologies.
                (4) Sensitivity Analysis
                 Table VI-206 shows the principal findings and includes sampling-
                error
                [[Page 24754]]
                confidence bounds for the five parameters used in the CAFE model. The
                confidence bounds represent the statistical uncertainty that is a
                consequence of having less than a census of data. NHTSA's 2011, 2012,
                and 2016 reports acknowledged another source of uncertainty: The
                central (baseline) statistical model can be varied by choosing
                different control variables or redefining the vehicle classes or crash
                types, which for example, could produce different point estimates.
                 Beginning with the 2012 Kahane report, NHTSA has provided results
                of 11 plausible alternative models that serve as sensitivity tests of
                the baseline model. Each alternative model was tested or proposed by:
                Farmer (IIHS) or Green (UMTRI) in their peer reviews; Van Auken (DRI)
                in his public comments; or Wenzel in his parallel research for DOE. The
                2012 Kahane and 2016 Puckett and Kindelberger reports provide further
                discussion of the models and the rationales behind them.
                 Alternative models use NHTSA's databases and regression-analysis
                approach but differ from the central model in one or more explanatory
                variables, assumptions, or data restrictions. The agencies applied the
                11 techniques to the latest databases to generate alternative CAFE
                model coefficients. The range of estimates produced by the sensitivity
                tests offers insight to the uncertainty inherent in the formulation of
                the models, subject to the caveat that these 11 tests are, of course,
                not an exhaustive list of conceivable alternatives.
                 The central and alternative results follow, ordered from the lowest
                to the highest estimated increase in societal risk per 100-pound
                reduction for cars weighing less than 3,201 pounds:
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.421
                [[Page 24755]]
                BILLING CODE 4910-59-C
                 The sensitivity tests illustrate both the fragility and the
                robustness of central estimates. On the one hand, the variation among
                the coefficients is quite large relative to the central estimate: In
                the preceding example of cars < 3,201 pounds, the estimated
                coefficients range from almost zero to almost double the central
                estimate. This result underscores the key relationship that the
                societal effect of mass reduction is small. In other words, varying how
                to model some of these other vehicle, driver, and crash factors, which
                is exactly what sensitivity tests do, can appreciably change the
                estimate of the societal effect of mass reduction.
                 On the other hand, variations are not particularly large in
                absolute terms. The ranges of alternative estimates are generally in
                line with the sampling-error confidence bounds for the central
                estimates. Generally, in alternative models as in the central model,
                mass reduction tends to be relatively more harmful in the lighter
                vehicles and more beneficial in the heavier vehicles, just as they are
                in the central analysis. In all models, the point estimate of the
                coefficient is positive for the lightest vehicle class, cars < 3,201
                pounds. In 10 out of 11 models, the point estimate is negative for CUVs
                and minivans, and in nine out of 11 models the point estimate is
                negative for LTVs >= 5,014 pounds. The agencies believe the central
                case uses the most rigorous methodology, as discussed further above,
                and provides the best estimates of the impacts of mass reduction on
                safety.
                 Tom Wenzel commented confirming a preference for the alternative
                model with footprint separated into track width and wheelbase, and with
                the induced exposure data limited to stopped vehicle cases.\2002\
                Wenzel asserts that splitting footprint into its components reduces
                multicollinearity with curb weight, and that limiting induced exposure
                cases to stopped vehicles mitigates bias against driver-vehicle pairs
                that are less likely to be involved in crashes. Based on this feedback
                and the intuitiveness of the approach, the agencies further considered
                the alternative model with footprint split into track width and
                wheelbase. Consistent with previous analyses and assessments, there are
                problems with splitting footprint into its components within the mass-
                size-safety models because of strong correlations among curb weight,
                track width and wheelbase. For all vehicle classes in the analysis,
                curb weight is correlated either nearly as high or higher with track
                width as with footprint. Track width and wheelbase are also highly
                correlated with one another (ranging from around 0.64 to 0.80, with the
                exceptions of smaller correlations for large pickups and minivans).
                Viewed from another angle, wheelbase is almost perfectly correlated
                with footprint (with correlations ranging from around 0.95 to 0.97).
                ---------------------------------------------------------------------------
                 \2002\ Wenzel, T., Lawrence Berkeley National Laboratories,
                Docket No. EPA-HQ-OAR-2018-0283-4118.
                ---------------------------------------------------------------------------
                 Considered in concert, the track width and wheelbase model not only
                essentially incorporates the full correlation issues from the baseline
                model (curb weight highly correlated with another independent
                variable), but also adds a further correlation issue (the variable that
                is highly correlated with curb weight is also highly correlated with a
                separate independent variable). The agencies examined supplementary
                means of confirming the relative methodological merit of the footprint-
                based model and the track-width-wheelbase-based alternative. The
                supplementary analysis centered on the condition index, which
                quantifies the invertibility of the matrix of independent variables in
                a given model through its measure, the condition number.\2003\ A model
                with a low condition number has relatively low correlations among its
                independent variables, and thus its invertibility and the corresponding
                model outputs are robust to variations in model input values. A model
                with a high condition number has relatively high correlations among its
                independent variables, and thus its invertibility and model outputs are
                not robust to variations in model input values. That is, a model with a
                high condition number is likely to be subject to the problems
                associated with multicollinearity. Although there is no strict
                threshold condition number value to indicate multicollinearity, higher
                values indicate greater likelihood that the independent variables are
                correlated to a problematic degree.
                ---------------------------------------------------------------------------
                 \2003\ See Belsley, D.A., Kuh, E., and Welsch, R.E. (1980).
                ``The Condition Number.'' Regression Diagnostics: Identifying
                Influential Data and Sources of Collinearity. New York: John Wiley &
                Sons; Freund, R.J. and Littell, R.C. (2000). SAS System for
                Regression, Third Edition. Cary, NC: SAS Institute, Inc.; and
                Hallahan, C. (1995). ``Understanding the Multicollinearity
                Diagnostics in SAS/Insight and Proc Reg.'' SAS Conference
                Proceedings, Washington, DC, October 8-10, 1995.
                ---------------------------------------------------------------------------
                 The condition index offers an alternative means of capturing the
                same forces as the variance inflation factor (VIF), which the agencies
                have used historically (including in this rulemaking) as a diagnostic
                of multicollinearity. However, the condition index offers some
                advantages relative to the VIF. Notably, the condition index applies
                regardless of the econometric form of the model (i.e., the
                decomposition of the independent variables is the same regardless of
                how the variables are applied in the model). This is distinct from the
                VIF, which is limited to a linear diagnostic of the data that may not
                map well to non-linear econometric models, including the logistic
                regression models that form the core of the curb weight-fatality risk
                analysis. The condition index estimates the incremental effects of
                individual variables, which is helpful in an analysis of which
                independent variables are the most problematic. Conversely, the
                diagnostic values from the VIF are not necessarily sensitive to
                incremental correlated variables, as the VIF value (1/(1-R\2\) does not
                necessarily change much once correlations are relatively high (i.e.,
                when R\2\ is already high, the inclusion of one or more highly
                correlated variables may not change R\2\, and in turn, the VIF, by
                much.
                 An incremental comparison of VIF estimates for the data confirmed
                the potential weakness of the VIF in this case. For the CUV-minivan
                model data, the VIF decreases from 9.4 to 6.7 when: (1) Substituting
                either track width or footprint for footprint that has an identical
                correlation with curb weight as footprint; and (2) adding the other
                component of footprint. This result is counterintuitive (i.e., the
                simpler model should necessarily have fewer issues of
                multicollinearity), and may be an artifact of differences in model fit
                (e.g., a higher R\2\ in the simpler model could indicate better model
                fit rather than anything problematic in terms of correlation
                structure). This result led the agencies to question how well the VIF
                identifies relative impacts of multicollinearity across related models,
                especially in non-linear applications.
                 The calculated condition numbers for the curb weight-footprint
                models and their corresponding curb weight-wheelbase-track width
                alternatives were consistent with expectations regarding
                multicollinearity, however. The condition numbers for the curb weight-
                wheelbase-track width models are approximately two to three times
                higher than the condition numbers for the curb weight-footprint models.
                This indicates that the level of imprecision in model estimates using
                track width and wheelbase would be expected to be between approximately
                two to three times higher than in the baseline models using footprint.
                Unlike the VIF, the condition index supports a hypothesis that
                multicollinearity would not be mitigated in an alternative with
                disaggregated variables that are highly
                [[Page 24756]]
                correlated with both the variable of interest and the variable they are
                replacing. Considering these results, the agencies that using footprint
                to represent vehicle size in the safety models provides a more reliable
                estimate of safety impacts than splitting footprint into track width
                and wheelbase.
                 The agencies also considered the use of stopped-vehicle data as an
                alternative. The primary problem with this approach is that the
                agencies do not observe as large of a share of cases on roads with
                higher travel speeds (e.g., interstate highways) when including only
                stopped vehicles; this relationship influences the extent to which the
                induced exposure data reflect the distributions of driver attributes
                and contextual effects across national VMT. Based on this assessment,
                the agencies believe the methodology used for the analysis in the
                proposal provides a more reliable and representative estimate of safety
                impacts, and thus is not changing the methodology for today's final
                rule.
                 In a related comment, Wenzel proposes that future analyses should
                directly account for differences in curb weight between vehicles in
                two-vehicle crashes. The agencies believe that would require the
                development of a model that directly accounts for the relative weights
                of vehicles in two-vehicle crashes, and that such a model would require
                peer review. Key alternatives to test would vary in terms of the
                functional form of the mass disparity between two crash partners (e.g.,
                a relative mass ratio consistent with the delta-V calculation presented
                above, linear mass difference, non-linear mass difference). The
                agencies will consider initiating work to explore such a model in the
                future.
                 DRI requested the agencies clarify whether the analysis accounts
                for all road users (i.e., including pedestrians, bicyclists,
                motorcyclists, and other crash partners), while the Pennsylvania
                Department of Environmental Protection commented, ``[i]t is inadequate
                for the agencies' analysis for this Proposed Rule to only focus on
                frontal crashes while omitting near-frontal collisions, side-impact
                collisions, rear-end collisions, rollover accidents, impacts with
                stationary objects and accidents involving pedestrians.'' \2004\ The
                agencies confirm that the analysis presented in this section continues
                to apply the methodology developed by Kahane, which incorporates all
                road users, without double-counting, to identify societal fatality rate
                impacts. Because every fatal crash (across crash types) is included in
                the analysis, not just frontal crashes, the agencies find this comment
                lacks a basis. The agencies believe the commenter's confusion may stem
                from the use of front-to-back crashes to generate estimates of the
                proportions of all driving for each vehicle model associated with
                particular characteristics of drivers (e.g., age, gender) and crashes
                (e.g., urban/rural, day/night). These crashes represent the best
                available trade-off among sample size, representativeness of overall
                vehicle and driver exposure, and mitigating bias in a sample that is
                intended to be effectively random (i.e., the probability of being
                struck from behind by an at-fault driver is assumed to be a function of
                characteristics of other drivers and travel demand, but not of the
                struck driver or the struck vehicle).
                ---------------------------------------------------------------------------
                 \2004\ Pennsylvania Department of Environmental Protection,
                Detailed Comments, Docket No. NHTSA-2018-0067-11956, at 9.
                ---------------------------------------------------------------------------
                (5) Fleet Simulation Study
                 Commenters to recent CAFE rulemakings, including some vehicle
                manufacturers, have suggested designs and materials of more recent
                model year vehicles may have weakened the historical statistical
                relationships between mass, size, and safety. NHTSA and EPA agreed that
                the statistical analysis would be improved by using an updated crash
                and exposure database reflecting more recent safety technologies,
                vehicle designs and materials, and reflecting changes in the vehicle
                fleet. As mentioned above, a new crash and exposure database was
                created with the intention of capturing modern vehicle engineering and
                has been employed for assessing safety effects for CAFE rules since
                2012.
                 The agencies have traditionally relied solely on real-world crash
                data as the basis for projecting the future safety implications for
                regulatory changes. The agencies are required to consider relevant data
                in setting standards.\2005\ Every fleet regulated by the agencies'
                standards differs from the fleet used to establish said standard, and
                as such, the light-duty vehicle fleet in the MY 2021-2026 timeframe
                will be different from the MY 2004-2011 fleet analyzed above. This is
                not a new or unique phenomenon, but instead is an inherent challenge in
                regulating an industry reliant on continual innovation. This is the
                agencies' sixth evaluation of effects of mass reduction and/or
                downsizing,\2006\ comprising databases ranging from MYs 1985 to 2011.
                Despite continual claims that modern lightweight engineering will
                render current data obsolete, results of the six studies, while not
                identical, have been generally consistent in showing a small, negative
                impact related to mass reduction. The agencies strongly believe that
                real-world crash data remains the best, relevant data to measure the
                effect of mass reduction on safety.
                ---------------------------------------------------------------------------
                 \2005\ See Center for Biological Diversity v. NHTSA, 538 F.3d
                1172, 1203 (9th Cir. 2008).
                 \2006\ As outlined throughout this section, NHTSA's six related
                studies include the new analysis supporting this rulemaking, and:
                Kahane, C.J. Vehicle Weight, Fatality Risk and Crash Compatibility
                of Model Year 1991-99 Passenger Cars and Light Trucks, National
                Highway Traffic Safety Administration (Oct. 2003), available at
                https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/809662;
                Kahane, C.J. Relationships Between Fatality Risk, Mass, and
                Footprint in Model Year 1991-1999 and Other Passenger Cars and LTVs
                (Mar. 24, 2010), in Final Regulatory Impact Analysis: Corporate
                Average Fuel Economy for MY 2012-MY 2016 Passenger Cars and Light
                Trucks, National Highway Traffic Safety Administration (Mar. 2010)
                at 464-542; Kahane, C.J. Relationships Between Fatality Risk, Mass,
                and Footprint in Model Year 2000-2007 Passenger Cars and LTVs--
                Preliminary Report, National Highway Traffic Safety Administration
                (Nov. 2011), available at Docket ID NHTSA-2010-0152-0023); Kahane,
                C.J. Relationships Between Fatality Risk, Mass, and Footprint in
                Model Year 2000-2007 Passenger Cars and LTVs: Final Report, NHTSA
                Technical Report. Washington, DC: NHTSA, Report No. DOT-HS-811-665;
                and Puckett, S.M., & Kindelberger, J.C. Relationships between
                Fatality Risk, Mass, and Footprint in Model Year 2003-2010 Passenger
                Cars and LTVs--Preliminary Report, National Highway Traffic Safety
                Administration (June 2016), available at https://www.nhtsa.gov/sites/nhtsa.dot.gov/files/2016-prelim-relationship-fatalityrisk-mass-footprint-2003-10.pdf.
                ---------------------------------------------------------------------------
                 However, because lightweight vehicle designs introduce fundamental
                changes to the structure of the vehicle, there remains a persistent
                question of whether historical safety trends will apply. To address
                this concern and to verify that real-world crash data remain an
                appropriate source of data for projecting mass-safety relationships in
                the future fleet, in 2014, NHTSA sponsored research to develop an
                approach to utilize experimental lightweight vehicle designs to
                evaluate safety in a broader range of real-world representative
                crashes.\2007\ NHTSA contracted with George Washington University to
                perform a fleet simulation model to study the impact and relationship
                of light-weighted vehicle design with injuries and fatalities.\2008\
                The study involved simulating crashes on eight test vehicles, five of
                which were equipped with lightweight materials
                [[Page 24757]]
                and advanced designs not yet incorporated into the U.S. fleet. The
                study assessed a range of frontal crashes, including crashes with fixed
                objects and other vehicles, across wide range of vehicle speeds, and
                with mid-size male and mid-size female dummies.\2009\ In all, more than
                440 vehicle crashes with 1,520 dummy passengers were simulated for a
                range of crash speeds and crash configurations. Results from the fleet
                simulation study showed the trend of increased societal injury risk for
                light-weighted vehicle designs occurs for both single vehicle and two-
                vehicle crashes. Results are listed in Table VI-207.\2010\
                ---------------------------------------------------------------------------
                 \2007\ See also 83 FR at 43133 (Aug 24, 2018).
                 \2008\ Samaha, R.R., Prasad, P., Marzougui, D., Cui, C., Digges,
                K., Summers, S., Patel S., Zhao, L., & Barsan-Anelli, A. (2014,
                August). Methodology for evaluating fleet protection of new vehicle
                designs--Application to lightweight vehicle designs. Report No. DOT
                HS 812 051A, Washington, DC--National Highway Traffic Safety
                Administration.
                 \2009\ Regulatory and consumer information crash safety tests
                are performed at high speeds, and the dummy occupant is generally a
                mid-size male. In the real world, crashes occur at various impact
                velocities and configurations; with various impact partners (e.g.,
                rigid obstacles, lighter or heavier vehicles); and involve occupants
                of various sizes and ages.
                 \2010\ This fleet simulation study does not provide information
                that can be used to modify coefficients derived for the NPRM
                regression analysis because of the restricted types of crashes and
                vehicle designs. Additionally, the fleet simulation study assumed
                restraint equipment to be as in the baseline model, in which
                restraints/airbags are not redesigned to be optimal with light-
                weighting.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.422
                [GRAPHIC] [TIFF OMITTED] TR30AP20.423
                 The change in the safety risk from the fleet simulation study was
                directionally consistent with results for passenger cars from the 2012
                Kahane report,\2011\ the 2016 Puckett and Kindelberger report, and the
                analysis used for the proposal and today's final rule. As noted, fleet
                simulations were performed in frontal crash mode and did not consider
                other crash modes such as rollover crashes.\2012\ The fleet simulation
                analysis confirmed that real-world crash data were still a reliable
                source for analyzing mass safety impacts.
                ---------------------------------------------------------------------------
                 \2011\ The 2012 Kahane study considered only fatalities,
                whereas, the fleet simulation study considered severe (AIS 3+)
                injuries and fatalities (DOT HS 811 665).
                 \2012\ The risk assessment for CUV in the regression model
                combined CUVs and minivans in all crash modes and included belted
                and unbelted occupants.
                ---------------------------------------------------------------------------
                 Despite the results of the fleet simulation analysis, which was
                republished in the proposal, the agencies received additional comments
                questioning the assumption that relationships among vehicle mass, size,
                and fatality risk will continue in the future. For example, the
                Alliance for Vehicle Efficiency asserted that using lighter frame
                materials has no impact on safety, noting that any mass reduction
                strategies are applied to components that are unrelated to crash safety
                and crash ratings have not declined for vehicles over the past five
                years.\2013\ CARB commented that the agencies did not account for new
                vehicle improvements and claimed the data used for the analysis was
                ``not a good indicator of the safety performance of future purpose-
                designed lightweighted vehicles.'' \2014\ Consumers Union offered a
                similar appraisal, indicating that the MYs in the sample are ``unlikely
                to capture the current and future mass/fatality relationship of modern
                vehicles.'' \2015\ While the Aluminum Association commented vehicle
                size, not mass, is the only physical feature that impacts safety.\2016\
                The American Chemistry Council, Hyundai, and Tesla commented that it is
                feasible to utilize
                [[Page 24758]]
                design improvements and technologies to offset the incremental risk for
                vehicle occupants associated with mass reduction.\2017\ EDF said the
                mass-safety analysis did not agree with conclusions from a study by the
                Michigan Manufacturing Technology Center.\2018\ Comments from States
                and Cities, American Honda, ICCT, and NRDC shared these
                sentiments.\2019\
                ---------------------------------------------------------------------------
                 \2013\ Alliance for Vehicle Efficiency, Detailed Comments,
                Docket No. NHTSA-2018-0067-11696, at 11.
                 \2014\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
                11873, at 270.
                 \2015\ Consumers Union, Detailed Comments, Docket No. NHTSA-
                2018-0067-12068, at 18.
                 \2016\ Aluminum Association, Detailed Comments, Docket No.
                NHTSA-2018-0067-11952, at 3.
                 \2017\ American Chemistry Council, Detailed Comments, Docket No.
                EPA-HQ-OAR-2018-0283-1415, at 2-8; Hyundai-Kia America Technical
                Center, Detailed Comments, Docket No. EPA-HQ-OAR-2018-0283-4411, at
                13; Tesla, Detailed Comments, Docket No. EPA-HQ-OAR-2018-0283-4186,
                at 21-23.
                 \2018\ Michigan Manufacturing Technology Center study ``Vehicle
                Lightweighting: A Review of the Safety of Reduced Weight Passenger
                Cars and Light Duty Trucks,'' October 2018, available at https://advocacy.consumerreports.org/wp-content/uploads/2018/10/CU-MMTC-Safety-Study-10-24-2018.pdf.
                 \2019\ States and Cities, Detailed Comments, Docket No. NHTSA-
                2018-0067-11735 at 81 and 95; American Honda, Detailed Comments,
                Docket No. NHTSA-2018-0067-11818, at 15; ICCT, Detailed Comments,
                Docket No. NHTSA-2018-0067-11741, at II-10-11. National Resources
                Defense Council, Detailed Comments, Docket No. EPA-HQ-OAR-2018-0283-
                4410, at 11-14.
                ---------------------------------------------------------------------------
                 These comments and the MMTC study ignored the results of the fleet
                simulation study and seem premised on the notion that a vehicles'
                performance on NHTSA FMVSS, NHTSA voluntary NCAP, and IIHS voluntary
                safety tests is the only measure for assessing societal safety impacts
                for mass reduction. The regulatory and consumer information tests are
                representative of real-world, single-vehicle crash configurations.
                However, the tests are performed at constant speeds, and the dummy
                occupant is generally a mid-size male. In the real world, crashes occur
                at various impact velocities and configurations; with various impact
                partners (e.g., rigid obstacles, lighter or heavier vehicles); and
                involve occupants of various sizes and ages. The fleet simulation
                study, summarized above, assessed additional types of frontal crashes,
                including crashes with fixed objects and other vehicles at a wide range
                of vehicle speeds, and with mid-size male and mid-size female dummies.
                The fleet simulation study was more comprehensive and focused on the
                need to assess overall societal safety impacts. The fleet simulation
                study found that vehicle mass does impact safety with future
                lightweight vehicle designs that perform well on regulatory and
                consumer information tests.
                 The agencies received one comment regarding the fleet simulation
                analysis. CARB commented that the analysis tested too few vehicles and
                crash types, should have optimized restraints in the lightweighted
                models to simulate future safety improvements instead of using modern
                restraints, and lacked credibility because the results of the fleet
                simulation analysis did not reproduce the same results of other
                studies.\2020\ CARB's comments demonstrate a general misunderstanding
                of the fleet simulation analysis; the analysis was not intended to
                serve as a prediction of how the future vehicle fleet will perform, but
                rather was an exploration of whether expected lightweighting techniques
                would alter the dynamic between mass reduction and safety. The analysis
                was not an attempt to model every potential vehicle construction or
                crash scenario. Attempting to simulate every future crash would be
                impractical and ineffective. The combination of vehicles and crash
                simulations were purposely selected to provide the strongest insight
                into the effective of lightweighting techniques. For passenger cars and
                light trucks, frontal crashes account for 58 percent of fatal crashes;
                \2021\ it is appropriate to focus research on understanding the effects
                of mass reduction where the largest issue exists. For the study, the
                use of generic restraint systems as the foundations for the models was
                intentional so that the models would be more representative of a
                vehicle class rather than a specific vehicle. The models of the
                restraint systems represented designs currently in production at time
                of the study in terms of pretensioners, load limiters and air bag
                inflators. It is worth noting that in general, driver air bags are
                similar in most vehicles. And finally, the analysis was not an attempt
                to reproduce the 2012 Kahane report or any other study. The fact that
                the fleet simulation analysis showed mass-reduction to be detrimental
                in more types of vehicles than in the FARS data only further highlights
                the need to consider how today's standards may impact mass-safety.
                While in the future there may be resources and opportunity to expand
                the fleet simulation approach to other crash scenarios and, if they
                become available, to include additional vehicle mass reduction
                concepts, the lack of potential future data does not justify ignoring
                the data that currently exist.
                ---------------------------------------------------------------------------
                 \2020\ CARB, Detailed Comments, Docket No. NHTSA-2018-0067-
                11873, at 272-73.
                 \2021\ Samaha, R.R., Prasad, P., Marzougui, D., Cui, C., Digges,
                K., Summers, S., Patel S., Zhao, L., & Barsan-Anelli, A. (2014,
                August). Methodology for evaluating fleet protection of new vehicle
                designs--Application to lightweight vehicle designs. Report No. DOT
                HS 812 051A, Washington, DC--National Highway Traffic Safety
                Administration.
                ---------------------------------------------------------------------------
                 From a higher perspective, the comments, and in particular CARB's
                comment, identify the problem with abandoning real-world crash data:
                There is no alternate methodology or data that can account for the full
                diversity of crash scenarios that occur in the real world. Real-world
                crash data is the only data type that can achieve that. Therefore, the
                agencies have determined that, while simulations can prove helpful to
                understanding potential effects of key crash scenarios and as a check
                on the agencies' preferred analysis, real-world data still is still the
                best, most relevant data available for assessing safety.
                (6) Summary of Mass Safety Impacts
                 Table VI-208 through Table VI-213 show results of NHTSA's vehicle
                mass-size-safety analysis over the cumulative lifetime of MY 1977-2029
                vehicles, for both the CAFE and CO2 programs, based on the
                MY 2017 baseline fleet, accounting for the projected safety baselines.
                Results are driven extensively by the degree to which mass is reduced
                in relatively light passenger cars and in relatively heavy vehicles
                because their coefficients in the logistic regression analysis have the
                most significant values. The agencies assume any impact on fatalities
                will occur over the lifetime of the vehicle, and the chance of a
                fatality occurring in any particular year is directly related to the
                weighted vehicle miles traveled in that year.
                BILLING CODE 4910-59-P
                [[Page 24759]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.424
                [[Page 24760]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.425
                [[Page 24761]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.426
                [[Page 24762]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.427
                [[Page 24763]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.428
                [[Page 24764]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.429
                 As shown in the tables above, all of the alternatives are estimated
                to lead to a decrease in the number of mass-related fatalities over the
                cumulative lifetime of MY 1977-2029 vehicles. The effects of mass
                changes on fatalities
                [[Page 24765]]
                range from a combined decrease (relative to the augural standards, the
                baseline) of 143 fatalities for Alternative #7 to a combined decrease
                of 288 fatalities for Alternatives #1 and #2. The difference in results
                by alternative depends upon how much weight reduction is used in that
                alternative and the types and sizes of vehicles to which the weight
                reduction applies. The decreases in fatalities are driven by impacts
                within passenger cars (decreases of between 167 and 380 fatalities) and
                are offset by impacts within light trucks (increases of between 9 and
                92 fatalities).
                 Changes in vehicle mass are estimated to decrease social safety
                costs over the lifetime of the nine model years by between $2.5 billion
                (for Alternative #7) and $5.1 billion (for Alternatives #1 and #2)
                relative to the augural standards at a three-percent discount rate and
                by between $1.5 billion and $3.1 billion at a seven-percent discount
                rate. The estimated decreases in social safety costs are driven by
                estimated decreases in costs associated with passenger cars, ranging
                from $3.0 billion (for Alternative #7) to $6.7 billion (for
                Alternatives #1 and #2) relative to the augural standards at a three-
                percent discount rate and by between $1.8 billion and $4.0 billion at a
                seven-percent discount rate. The estimated decreases in costs
                associated with passenger cars are offset partially by estimated
                increases in costs associated with light trucks, ranging from $0.1
                billion (for Alternative #5) to $1.6 billion (for Alternatives #1 and
                #2) relative to the Augural standards at a three-percent discount rate
                and by between $0.1 billion and $0.9 billion at a seven-percent
                discount rate.
                 In this analysis, the profile of mass reduction across vehicle
                models leads to a small, but beneficial effect on fatalities as fuel
                economy standards are tightened. Table VI-212 through Table VI-219
                present average annual estimated safety effects of vehicle mass
                changes, for CYs 2036-2045. The CY-level values offer a complementary
                view of the impacts of fuel economy standards on mass-related
                fatalities relative to model-year-level results. Effects by CY over the
                interval selected (2036-2045) enable a summary view of (a flow of)
                annual fatality impacts during a period where vehicles subjected to the
                standards have not only fully entered the fleet, but also interact with
                both older and newer vehicles. Conversely, the MY-level values offer a
                summary view of (a stock of) the impacts of fuel economy standards for
                the lifetime of a given MY:
                [[Page 24766]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.430
                [[Page 24767]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.431
                [[Page 24768]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.432
                [[Page 24769]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.433
                [[Page 24770]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.434
                [[Page 24771]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.435
                 For all light-duty vehicles, mass changes are estimated to lead to
                an average annual decrease in fatalities in all alternatives evaluated
                for CYs 2035-2045. The effects of mass changes on fatalities range from
                a combined
                [[Page 24772]]
                decrease (relative to the augural standards) of 20 fatality per year
                for Alternative #7 to a combined decrease of 37 fatalities per year for
                Alternative #4. The difference in the results by alternative depends
                upon how much weight reduction is used in that alternative and the
                types and sizes of vehicles to which the weight reduction applies. The
                decreases in fatalities are generally driven by impacts within
                passenger cars (decreases of between 22 and 50 fatalities per year
                relative to the augural standards) and are offset by impacts within
                light trucks (increases of between 2 and 12 fatalities per year).
                 Changes in vehicle mass are estimated to decrease average annual
                social safety costs in CY 2035-2045 by between $0.3 billion (for
                Alternative #7) and $0.6 billion (for Alternative #4) at a three-
                percent discount rate relative to the augural standards (decrease of
                between $0.1 and $0.2 billion at a seven-percent discount rate).
                Average annual social safety costs associated with passenger cars in CY
                2035-2045 are estimated to decrease by between $0.3 billion and $0.7
                billion at a three-percent discount rate (decrease of between $0.1
                billion and $0.3 billion at a seven-percent discount rate), but this
                effect is partially offset by a corresponding increase in costs
                associated with light trucks (increase of $0.2 billion or less across
                alternatives at three-percent and seven-percent discount rates).
                 To help illuminate effects at the model year level, Table VI-220
                presents the lifetime fatality impacts associated with vehicle mass
                changes for passenger cars, light trucks, and all light-duty vehicles
                by model year under the preferred alternative, relative to the augural
                standards for the CAFE Program. Table VI-221 presents an analogous
                table for the CO2 Program.
                [[Page 24773]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.436
                 Under the preferred alternative, passenger car fatalities
                associated with mass changes are estimated to decrease relative to the
                augural standards steadily from MYs 2018-19 (decrease of 5 fatalities)
                through MY 2028 (decrease of
                [[Page 24774]]
                53 fatalities). Conversely, light truck fatalities associated with mass
                changes under the preferred alternative are estimated to increase
                relative to the augural standards from MY 2019 (increase of 2
                fatalities) through MY 2029 (increase of 9 fatalities).
                 Table VI-222 and Table VI-223 present estimates of monetized
                lifetime social safety costs associated with mass changes by model year
                at three-percent and seven-percent discount rates, respectively for the
                CAFE Program. Table VI-224 and Table VI-225 show comparable tables from
                the perspective of the CO2 Program.
                [[Page 24775]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.437
                [[Page 24776]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.438
                 Lifetime social safety costs associated with mass change in
                passenger cars are estimated to decrease by between $0.1 billion (for
                MYs 2020-22) and $0.3 billion (for MYs 2026-29) at a three-percent
                discount rate. At a seven-
                [[Page 24777]]
                percent discount rate, lifetime social safety costs associated with
                mass change in passenger cars are estimated to decrease by between $0.1
                billion and $0.2 billion from MY 2021 through MY 2029. Lifetime social
                safety costs associated with mass change in light trucks are estimated
                to increase by $0.1 billion or less for all MYs at three-percent and
                seven-percent discount rates.
                BILLING CODE 4910-59-P
                [[Page 24778]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.439
                [[Page 24779]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.440
                [[Page 24780]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.441
                [[Page 24781]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.442
                [[Page 24782]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.443
                [[Page 24783]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.444
                 As shown in the tables above, all of the alternatives are estimated
                to lead to a decrease in the number of mass-related fatalities over the
                cumulative lifetime of MY 1977-2029 vehicles. The effects of mass
                changes on fatalities
                [[Page 24784]]
                range from a combined decrease (relative to the augural standards, the
                baseline) of 126 fatalities for Alternative #7 to a combined decrease
                of 253 fatalities for Alternatives #1 and #2. The difference in results
                by alternative depends upon how much weight reduction is used in that
                alternative and the types and sizes of vehicles to which the weight
                reduction applies. The decreases in fatalities are driven by impacts
                within passenger cars (decreases of between 146 and 33 fatalities) and
                are offset by impacts within light trucks (increases of between 8 and
                81 fatalities).
                 Changes in vehicle mass are estimated to decrease social safety
                costs over the lifetime of the nine model years by between $2.2 billion
                (for Alternative #7) and $4.5 billion (for Alternatives #1 and #2)
                relative to the augural standards at a three-percent discount rate and
                by between $1.3 billion and $2.7 billion at a seven-percent discount
                rate. The estimated decreases in social safety costs are driven by
                estimated decreases in costs associated with passenger cars, ranging
                from $2.6 billion (for Alternative #7) to $5.9 billion (for
                Alternatives #1 and #2) relative to the Augural standards at a three-
                percent discount rate and by between $1.6 billion and $3.5 billion at a
                seven-percent discount rate. The estimated decreases in costs
                associated with passenger cars are offset partially by estimated
                increases in costs associated with light trucks, ranging from $0.1
                billion (for Alternative #5) to $1.4 billion (for Alternatives #1 and
                #2) relative to the Augural standards at a three-percent discount rate
                and by between $0.1 billion and $0.8 billion at a seven-percent
                discount rate.
                 In this analysis, the profile of mass reduction across vehicle
                models leads to a small, but beneficial effect on fatalities as fuel
                economy standards are tightened. Table VI-232 through Table VI-237
                present average annual estimated safety effects of vehicle mass
                changes, for CYs 2035-2045:
                BILLING CODE 4910-59-P
                [[Page 24785]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.445
                [[Page 24786]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.446
                [[Page 24787]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.447
                [[Page 24788]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.448
                [[Page 24789]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.449
                [[Page 24790]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.450
                BILLING CODE 4910-59-C
                 For all light-duty vehicles, mass changes are estimated to lead to
                an average annual decrease in fatalities in all alternatives evaluated
                for CYs 2035-
                [[Page 24791]]
                2045. The effects of mass changes on fatalities range from a combined
                decrease (relative to the augural standards) of 17 fatality per year
                for Alternative #7 to a combined decrease of 34 fatalities per year for
                Alternative #4. The difference in the results by alternative depends
                upon how much weight reduction is used in that alternative and the
                types and sizes of vehicles to which the weight reduction applies. The
                decreases in fatalities are generally driven by impacts within
                passenger cars (decreases of between 19 and 44 fatalities per year
                relative to the augural standards) and are offset by impacts within
                light trucks (increases of between 2 and 11 fatalities per year).
                 Changes in vehicle mass are estimated to decrease average annual
                social safety costs in CY 2035-2045 by between $0.2 billion (for
                Alternative #7) and $0.5 billion (for Alternative #4) at a three-
                percent discount rate relative to the augural standards (decrease of
                between $0.1 and $0.2 billion at a seven-percent discount rate).
                Average annual social safety costs associated with passenger cars in CY
                2035-2045 are estimated to decrease by between $0.3 billion and $0.6
                billion at a three-percent discount rate (decrease of between $0.1
                billion and $0.3 billion at a seven-percent discount rate), but this
                effect is partially offset by a corresponding increase in costs
                associated with light trucks (increase of $0.1 billion or less across
                alternatives at three-percent and seven-percent discount rates).
                 To help illuminate effects at the model year level, Table VI-238
                presents the lifetime fatality impacts associated with vehicle mass
                changes for passenger cars, light trucks, and all light-duty vehicles
                by model year under the preferred alternative, relative to the Augural
                standards for the CAFE Program. Table VI-239 presents an analogous
                table for the CO2 Program.
                [[Page 24792]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.451
                 Under the preferred alternative, passenger car fatalities
                associated with mass changes are estimated to decrease relative to the
                augural standards steadily from MYs 2018-19 (decrease of 4 fatalities)
                through MYs 2028-29
                [[Page 24793]]
                (decrease of 46 fatalities). Conversely, light truck fatalities
                associated with mass changes under the preferred alternative are
                estimated to increase relative to the augural standards from MY 2019
                (increase of 1 fatality) through MY 2029 (increase of 8 fatalities).
                 Table VI-240 and Table VI-241 present estimates of monetized
                lifetime social safety costs associated with mass changes by model year
                at three-percent and seven-percent discount rates, respectively for the
                CAFE Program. Table VI-242 and Table VI-243 show comparable tables from
                the perspective of the CO2 Program.
                [[Page 24794]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.452
                [[Page 24795]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.453
                BILLING CODE 4910-59-C
                 Lifetime social safety costs associated with mass change in
                passenger cars are estimated to decrease by between $0.1 billion (for
                MYs 2020-23) and $0.3
                [[Page 24796]]
                billion (for MYs 2026-29) at a three-percent discount rate. At a seven-
                percent discount rate, lifetime social safety costs associated with
                mass change in passenger cars are estimated to decrease by between $0.1
                billion and $0.2 billion from MY 2022 through MY 2029. Lifetime social
                safety costs associated with mass change in light trucks are estimated
                to increase by less than $0.1 billion for all MYs at three-percent and
                seven-percent discount rates.
                b) Impact of Vehicle Prices on Fatalities
                 The sales and scrappage responses discussed above have important
                safety consequences and influence safety outcomes through the same
                basic mechanism, fleet turnover. In the case of the scrappage response,
                delaying fleet turnover keeps drivers in older vehicles which are less
                safe than newer vehicles.\2022\ Similarly, the sales response slows the
                rate at which newer vehicles, and their associated safety improvements,
                enter the on-road population. The sales response also influences the
                mix of vehicles on the road--with more stringent CAFE standards leading
                to a higher share of light trucks sold in the new vehicle market,
                assuming all else is equal. Light trucks have higher rates of fatal
                crashes when interacting with passenger cars and, as earlier sections
                discussed, different directional responses to mass reduction technology
                based on the existing mass and body style of the vehicle.\2023\
                ---------------------------------------------------------------------------
                 \2022\ See Passenger Vehicle Occupant Injury Severity by Vehicle
                Age and Model Year in Fatal Crashes, Traffic Safety Facts Research
                Note, DOT-HS-812-528, National Highway Traffic Safety
                Administration, April, 2018, and The Relationship Between Passenger
                Vehicle Occupant Injury Outcomes and Vehicle Age or Model Year in
                Police-Reported Crashes, Traffic Safety Facts Research Note, DOT-HS-
                (812-937), National Highway Traffic Safety Administration, March,
                2020.
                 \2023\ See Section 6. Analytical Approach as Applied to
                Regulatory Alternatives] for a full explanation of the sales and
                scrappage effects and how they are modeled.
                ---------------------------------------------------------------------------
                 With an integrated fleet model now part of the analytical framework
                for CAFE analysis, any effects on fleet turnover (either from delayed
                vehicle retirement or deferred sales of new vehicles) will affect the
                distribution of both ages and model years present in the on-road fleet.
                Because each of these vintages carries with it inherent rates of fatal
                crashes, and newer vintages are generally safer than older ones,
                changing that distribution will change the total number of on-road
                fatalities under each regulatory alternative. Similarly, the dynamic
                fleet share model captures the changes in the fleet's composition of
                cars and trucks. As cars and trucks have different fatality rates,
                differences in fleet composition across the alternatives will affect
                fatalities.
                 At the highest level, the agencies calculate the impact of the
                sales and scrappage effects by multiplying the VMT of a vehicle by the
                fatality risk of that vehicle. For this analysis, calculating VMT is
                rather simple: the agencies use the distribution of miles calculated in
                Section VI.D.1.b)(5)(b). The trickier aspect of the analysis is
                creating fatality rate coefficients. The fatality risk measures the
                likelihood that a vehicle will be involved in fatal accident per mile
                driven. As explained below, the agencies' methodology changed from the
                proposal to this final rule in response to comments, but the basic
                analytical framework remains the same. The agencies calculate the
                fatality risk of a vehicle based on the vehicle's model year, age, and
                style, while controlling for factors which are independent of the
                intrinsic nature of the vehicle, such as behavioral characteristics.
                (1) How the Agencies Modeled Impacts of Vehicle Scrappage and Sales on
                Fatalities in the NPRM
                 In the proposal, the sales-scrappage safety model comprised two
                components.\2024\ First, the agencies estimated an empirical
                relationship among vehicle age, model year or vintage, and fatalities
                using the FARS database of fatal crashes, vehicle registration data
                from Polk to represent the on-road vehicle population, and the mileage
                accumulation schedules discussed in Section VI.D.1.b)(5) Vehicles Miles
                Traveled to estimate total vehicle use.\2025\ These data were used to
                construct per-mile fatality rates that varied by vehicle vintage, and
                also accounted for the influence of vehicle age. To accomplish this,
                the agencies used FARS data at a lower level of resolution; rather than
                looking at each crash and the specific factors that contributed to its
                occurrence, the agencies looked at the total number of fatal crashes
                involving light-duty vehicles over time with a focus on the influence
                of vehicle age and vehicle vintage. The model used in the proposal
                incorporated a weighted quartic polynomial regression (with each
                observation weighted by the number of registered vehicles it
                represented) on vehicle age, and included fixed effects for each model
                year present in the dataset. The model reproduced the observed
                fatalities of a given model year, at each age, reasonably well with
                more recent model years estimated with smaller errors. These estimates
                were used to account for the inherent safety risks of the legacy fleet
                and the influence of age on a vehicle's fatality rate.
                ---------------------------------------------------------------------------
                 \2024\ The derivation of the NPRM analysis is discussed in
                detail in Section 7 of the FRIA.
                 \2025\ The analysis supporting the CAFE rule for MYs 2017 and
                beyond did not account for differences in exposure or inherent
                safety risk as vehicles aged throughout their useful lives. However,
                the relationship between vehicle age and fatality risk is an
                important one. In a 2013 Research Note, NHTSA's National Center for
                Statistics and Analysis (NCSA) concluded a driver of a vehicle that
                is 4-7 years old is 10% more likely to be killed in a crash than the
                driver of a vehicle 0-3 years old, accounting for the other factors
                related to the crash. This trend continued for older vehicles more
                generally, with a driver of a vehicle 18 years or older being 71%
                more likely to be killed in a crash than a driver in a new vehicle.
                ``How Vehicle Age and Model Year Relate to Driver Injury Severity in
                Fatal Crashes,'' DOT HS 811 825, NHTSA NCSA, August 2013. While
                there are more registered vehicles that are 0-3 years old than there
                are 20 years or older (nearly three times as many) because most of
                the vehicles in earlier vintages are retired sooner, the average age
                of vehicles in the United States is 11.6 years old and has risen
                significantly in the past decade.
                ---------------------------------------------------------------------------
                 In the proposal, the agencies noted that factors other than the
                advent of new safety technologies have affected the historical trend in
                fatality and injury rates and are likely to continue to do so in the
                future. These include changes in driver behavior, including seat belt
                use, driving under the influence of alcohol or drugs, and driver
                distraction, particularly from the use of hand-held electronic devices
                such as smartphones, all of which affect either the frequency with
                which drivers are involved in crashes or the severity of accidents.
                They also include changes in the demographic composition of driving,
                since drivers of different ages, gender, income levels, and educational
                attainment have differing accident-involvement rates, as well as in the
                geographic distribution of motor vehicle travel, since road and driving
                conditions (visibility, etc.) tend to be poorer in rural areas than in
                urban locations, thus leading to more frequent and more severe crashes.
                Other factors affecting safety trends include infrastructure
                investments and road maintenance practices that improve road design and
                travel conditions, thus reducing the frequency and severity of crashes,
                improvements in accident response and emergency medical care, and
                cyclical variation in economic activity, which affects the demographic
                composition of drivers on the road.
                 Seat belts have historically been the single most effective safety
                technology, preventing roughly half of all fatalities in the event of a
                potentially fatal crash, and accounting for over half the lives
                cumulatively saved by all FMVSS-related safety technologies since
                1960.\2026\ While belts have been in passenger vehicles since the
                1960s, few
                [[Page 24797]]
                drivers or passengers initially used them. Over the past 3 decades,
                seat belt usage rates have steadily climbed from under 60 percent in
                the early 1990s to roughly 90 percent in 2018 and has been the single
                most significant factor in reducing fatality rates over time.
                Additional changes in seat belt use are possible but challenging to
                achieve, since the last drivers to buckle up are typically the most
                likely to be risk takers and are often the most resistant to changing
                their habits. Moreover, with usage rates already at 90 percent, there
                is less potential for continued improvement.
                ---------------------------------------------------------------------------
                 \2026\ Kahane, C.J., Lives Saved by Vehicle Safety Technologies
                and Associated Federal Motor Vehicle Safety Standards, 1960 to
                2012--Passenger Cars and LTVs, National Highway Traffic Safety
                Administration, Paper Number 15-0291. https://www-esv.nhtsa.dot.gov/Proceedings/24/files/24ESV-000291.PDF.
                ---------------------------------------------------------------------------
                 Overall, the agencies believe improvement in seat belt use is
                unlikely to have the impact going forward that it has in the past.
                Technological fixes are possible for seat belt use and impaired
                driving, but would likely require the promulgation of new regulation,
                and therefore cannot be assumed. Similarly, individual States could
                take steps to address impaired driving, speeding, driver distraction,
                seat belt use and roadway infrastructure improvements, but the pace and
                impact of such improvements is speculative. The agencies also note that
                improvements in roadway infrastructure and human factors such as belt
                and alcohol use potentially affect both old and new vehicles alike. If
                improvements in these non-vehicle factors are equally spread across
                vehicles of all MY age groups, the differences in their fatality rates
                would not change. In other words, these types of improvements might
                shift the entire MY fatality rate curve down rather than change its
                slope.
                 Nonetheless, the agencies stated that it was reasonable to expect
                some continuation in the generalized trend from non-vehicle technology
                factors such as these. In the analysis supporting the NPRM, our
                statistical model controlled for non-vehicle safety factors by
                accounting for the well-documented fact that older vehicles tend to be
                owned and driven by drivers whose demographic characteristics,
                behavior, and geographic location tends are associated with more
                frequent or severe crashes.
                 Second, the agencies created estimates of future fatality rates.
                The agencies noted that predicting future safety trends has an inherent
                degree of uncertainty, which was amplified due to the dearth of
                academic and empirical research available at the time of the proposal.
                Although the agencies expected further safety improvements because of
                advanced driver assistance systems, such as automatic braking and
                eventually fully automated vehicles, the pace of development and extent
                of consumer acceptance of these improvements was uncertain. Thus,
                instead of attempting to model the impact of future safety features
                directly, the agencies relied on two different trend models to predict
                future safety trends. The first model relied on the results from a
                previous NCSA study that measured the effect of known safety
                regulations on fatality rates by performing statistical evaluations of
                the effectiveness of motor vehicle safety technologies based on real
                world performance in the on-road vehicle fleet to determine the
                effectiveness of each safety technology.\2027\ The agencies used this
                information to forecast future fatality rates. The second model
                employed was simpler. The agencies used actual, aggregate fatality
                rates measured from 2000 through 2016 and modeled the fatality rate
                trend based on these historical data.
                ---------------------------------------------------------------------------
                 \2027\ Blincoe, L. and Shankar, U., ``The Impact of Safety
                Standards and Behavioral Trends on Motor Vehicle Fatality Rates,''
                National Highway Traffic Safety Administration, DOT HS 810 777,
                Washington, DC, January, 2007.
                ---------------------------------------------------------------------------
                 The agencies noted that both models had significant limitations and
                predicted significantly different safety trends. The NCSA study focused
                on projections to reflect known technology adaptation requirements, but
                it was conducted prior to the 2008 recession, which disrupted the
                economy and changed travel patterns throughout the country, and
                predated the emergence of newer technologies in the 2010s. The NCSA
                anticipated continued improvement well beyond 2020. By contrast, the
                historical fatality rate model reflected shifts in safety not captured
                by the NCSA model, but gave arguably implausible results after 2020
                because of an observed upward shift in fatalities between 2014 and
                2015. It essentially represented a scenario in which economic, market,
                or behavioral factors minimize or offset much of the potential impact
                of future safety technology. To reconcile the two projections of safety
                improvements beyond 2015, the agencies averaged the NCSA and historical
                fatality rate models, accepting each as an illustration of different
                and conflicting possible future scenarios.
                 The agencies received a number of comments on the provisional model
                used in the NPRM, which focused mainly on its omission of variables
                that change over time and can affect the safety of all vehicles in use,
                regardless of their original model year or current age. As indicated
                previously, these include changes in seat belt use, driving under the
                influence of alcohol or drugs, use of hand-held electronic devices,
                driver demographics, the geographic distribution of vehicle use, road
                design and maintenance, emergency response and medical care, and
                overall economic activity.
                 For example, CARB asserted that the NPRM modeling overestimated
                fatality rates for older vehicles because it did not ``control for
                factors that can have a significant influence on fatality risk, such as
                crash circumstances and driver characteristics.'' Elsewhere, CARB
                highlighted the omission of calendar year effects from the NPRM
                analysis, adding ``the agencies only model fatality rate as a function
                of model year, but fatality rate should be a function of both model
                year and calendar year [. . .] [which] would account for systematic
                safety improvements to the entire on-road fleet.'' \2028\ CARB also
                argued that analysis should account for safety differences between body
                styles, noting that passenger cars and other LTVs ``have historically
                had different safety regulations.'' \2029\ Passenger cars and LTVs are
                not always regulated at exactly the same pace and in some
                circumstances, LTV regulations have differed from passenger car
                regulations. However, with a few exceptions, both types of passenger
                vehicles are equipped with safety technologies that address the same
                basic safety hazards. Historically, these involve regulations that
                preserve passenger compartment integrity and protect passengers in the
                event of a crash. These include technologies such as air bags, seat
                belts, stronger roof structures, side door beams, and fuel tank
                integrity. Further, going forward, the agencies expect that both
                vehicle types will eventually all be equipped with the same advanced
                crash avoidance safety technologies that are currently being developed.
                Whatever differences there are have influenced the fatality rates and
                since this rulemaking uses combined average fatality rates (for PCs and
                LTVs) for the model, the results should closely mirror the results from
                an analysis that calculates the two vehicle types separately and then
                adds them together.
                ---------------------------------------------------------------------------
                 \2028\ CARB, Detailed Comments, NHTSA-2018-0067-11873 at 263.
                 \2029\ CARB, Auken Fatality Report, NHTSA-2018-0067-11881, at
                25.
                ---------------------------------------------------------------------------
                 Similarly, States and Cities noted the potential importance of
                factors that can affect trends in vehicle safety over time, pointing
                out that ``increased seat belt use over time, improvements in roadway
                design and life-saving emergency response and treatment, and crash
                compatibility with other vehicles improve the overall safety of
                vehicles currently on the road'' and therefore
                [[Page 24798]]
                concluded that ``the CAFE model's assumption that the fatality rate of
                a 1985 model year vehicle is 23.8 per billion vehicle miles traveled
                for any calendar year is incorrect. That error increases the risk of
                fatalities determined by the NPRM for scrappage by around 25 percent.''
                \2030\ Consumers Union echoed this argument and suggested driver
                characteristics and behavior may ``more strongly influence fatality
                risk than a vehicle's model year.'' \2031\
                ---------------------------------------------------------------------------
                 \2030\ States and Cities, Detailed Comments, NHTSA-2018-0067-
                11735, at 101 (internal citation omitted).
                 \2031\ Consumers Union, et al., NHTSA-2018-0067-11731,
                Attachment 11, at 14.
                ---------------------------------------------------------------------------
                 IPI speculated that omitting the effect of variables that change
                over time in ways that could affect fleet-wide safety may have caused
                the agencies' analysis to over-emphasize the role of safety
                improvements to new vehicles. Specifically, IPI observed that ``the
                agencies could not adequately control for driver behavior trends. And a
                decrease in fatalities could look like it was caused by vehicle
                improvements over time rather than societal changes.'' \2032\
                ---------------------------------------------------------------------------
                 \2032\ IPI, NHTSA-2018-0067-12213, at 71.
                ---------------------------------------------------------------------------
                 The agencies also received a few comments on their modeling
                choices. For example, CARB commented that the agencies equation for the
                legacy fleet was ``either incorrect or [had] limited domain-of-validity
                because it can potentially predict negative fatality rates'' and
                because it was missing an intercept term.\2033\ CARB suggested a
                logarithmic function would fix the problem. The agencies note that the
                polynomial specification of the safety model the agencies developed for
                the legacy fleet was extremely unlikely to predict negative fatality
                rates in light of the estimated values of its coefficients, and that
                its fixed-effects specification in effect included separate intercept
                terms for each model year, with that for the earliest model year
                serving as the ``reference case'' and thus performing the normal role
                of the constant term.
                ---------------------------------------------------------------------------
                 \2033\ CARB, Auken Fatality Report, NHTSA-2018-0067-11881, at
                25.
                ---------------------------------------------------------------------------
                 In electing to offset rebound-related safety consequences for the
                NPRM, the agencies distinguished the rebound effect from mass and fleet
                turnover impacts by describing the former as a voluntary consumer
                choice and the latter as imposed by the standards on consumers.\2034\
                The agencies acknowledged in the NPRM that a reasonable argument might
                be made that consumers' decisions to purchase newer and safer cars or
                light trucks and to keep older models in service are also voluntary
                consumer choices, in which case changes in their decisions in response
                to newly-adopted CAFE and CO2 standards might be accompanied
                by offsetting gains or losses in benefits. The agencies dismissed this
                argument in the NPRM by noting that new vehicle prices act as a barrier
                to entry for some consumers, hence--at least ``marginal'' shoppers--
                purchasing a more expensive vehicle is not a choice; and, without the
                ability to determine how many potential purchasers are `priced out' of
                the new vehicle market, it would be inappropriate to offset sales and
                scrappage safety impacts.\2035\ The agencies sought comment on this
                assumption.
                ---------------------------------------------------------------------------
                 \2034\ See 83 FR at 43107.
                 \2035\ The agencies further augmented the discussion by
                explaining that less stringent standards encouraged new vehicle
                purchases through lower vehicle prices while simultaneously
                discouraging additional driving due to higher operating costs. See
                id.
                ---------------------------------------------------------------------------
                 The agencies did not receive any suggestions for distinguishing
                between consumers who voluntarily delayed purchases and those who were
                forced to delay a purchase due to high vehicle prices. Thus, the
                problem of deciphering the motives behind delayed purchases still
                lingers. However, the agencies did receive several comments advocating
                that the agencies offset fatalities attributable to sales and scrappage
                as they do for the rebound effect. For example, NCAT commented that
                ``consumer purchases are voluntary and this effect should not be
                attributed to the standards.'' \2036\ The environmental group coalition
                commented that miles driven in older vehicles are ``a consumer choice,
                not something the standards compel.'' \2037\ In comparing the decision
                to retain and drive older vehicles to the decision to drive new
                vehicles more, i.e. the rebound effect, EDF concluded, ``to treat these
                identical choices in 180 degree different manners is of course
                manifestly arbitrary.'' \2038\
                ---------------------------------------------------------------------------
                 \2036\ NCAT, Comments, NHTSA-2018-0067-11969, at 32-33.
                 \2037\ Environmental Group Coalition, Appendix A, NHTSA-2018-
                0067-12000, at 40-41.
                 \2038\ EDF, Appendix B, NHTSA-2018-0067-12108, at 58.
                ---------------------------------------------------------------------------
                 On a rudimentary level, the agencies agree with commenters that
                purchasing decisions are a consumer choice. While reducing the
                stringency of the standards should make new vehicles more affordable,
                nothing in today's rule requires consumers to purchase a new vehicle;
                likewise, the analysis does not assume every older vehicle will be
                replaced immediately. There is no strict requirement that the agencies
                must offset consumer choices. In fact, such a viewpoint would be
                untenable. Nothing in today's rule compels private parties to do
                anything. If the agencies assumed all freely chosen or voluntary
                actions, such as driving or manufacturing automobiles, were not
                attributable to the rule, then each regulatory scenario would have the
                same net benefit--zero. As such, the agencies explanation in the
                proposal of freely chosen and voluntary was likely imprecise and led
                commenters to an overly broad conclusion. Deciding which behavioral
                responses are unambiguously attributable to a regulation and should
                thus be quantified, and distinguishing them from responses that would
                be anticipated to occur in its absence is inherently part of the
                rulemaking process, and inevitably requires agencies considering new
                regulations to apply careful judgment in making those distinctions.
                 To that end, the agencies felt it was appropriate to offset
                rebound-related safety costs because of the benefit rebound miles
                confer to society. As described in more detail in Section 1.b)(6),
                additional driving that occurs as a consequence of the fuel economy
                rebound effect is undertaken voluntarily, and the agencies can infer
                from the fact that it is freely chosen that the mobility benefits it
                provides necessarily exceed the additional operating costs and
                increased exposure to safety risks it entails. Since reducing the
                standards has the ancillary effect of reducing rebound miles, the
                agencies concluded that including safety costs associated with rebound
                driving would cause the agencies to underestimate the lost value of
                rebound driving; therefore, it was appropriate to offset rebound safety
                costs to account for the lost benefits.\2039\ Thus, the significance of
                the terms freely chosen and voluntary was to signal that consumers'
                actions were motivated in part by benefits that may not have been not
                explicitly identified or accounted for, rather than to act as a
                prohibitive characteristic.
                ---------------------------------------------------------------------------
                 \2039\ Arguably rebound fatalities and non-fatal injuries should
                be included in today's analysis as a cost without an offset. While a
                perfectly rational driver would fully and accurately internalize the
                costs associated with driving on a per-mile basis and would only
                drive if the expected benefits at least offset the expected costs,
                it is difficult to ascertain how much of the risk a real person
                internalizes. If not for the reduced standards, fatalities would
                increase due to rebound driving.
                ---------------------------------------------------------------------------
                 When considering commenters' suggestion to offset fleet turnover
                fatalities (as well as injury and ancillary costs), the agencies
                attempted to identify specific benefits whose loss would be logically
                attributable to the changes in standards this rule adopts, and were not
                accounted for elsewhere in
                [[Page 24799]]
                their analysis. The agencies considered whether accelerated turnover of
                the car and light truck fleet could cause mobility losses analogous to
                those resulting from the rebound effect, but determined that on
                balance, increasing the pace at which new vehicles replace older models
                that are retired from use provides additional mobility and other
                benefits.\2040\ In addition, the agencies considered whether consumers
                experience some previously unidentified loss in welfare when they
                purchase new vehicles, particularly when they do so to replace an older
                model. As explained in in Section 1.b)(6) and 1.b)(8), the agencies
                instead concluded that purchasers instead experience gains in welfare
                as a result, but that the resulting benefits are already accounted for
                elsewhere in their analysis.
                ---------------------------------------------------------------------------
                 \2040\ This occurs because newer vehicles are not only more
                fuel-efficient on average than the older models they replace, but
                also provide more reliable, comfortable, and otherwise higher-
                quality transportation service, so they tend to be driven more than
                those they replace.
                ---------------------------------------------------------------------------
                 Finally, the agencies contemplated whether--as commenters
                contended-- owners of older vehicles derive some heretofore
                unaccounted-for benefit from continuing to use them, which might be
                reduced when the rule encourages more rapid retirement of older models.
                Applying the same logic used to explain additional driving in response
                to the rebound effect, an older vehicle will continue to be maintained
                in working condition and driven when the benefits provided to the owner
                is sufficient to offset the costs of maintenance and operation,
                including the economic costs associated with additional exposure to
                safety risks. Therefore, there is a benefit to driving an older
                vehicle. But the relevant question is not whether a benefit exists but
                how this rule might affect those benefits. With the very limited
                exception of classic cars, it is unlikely that the benefit of driving
                an older vehicle confers a greater benefit than driving a newer
                vehicle.\2041\ Normally, when a vehicle is scrapped, it is replaced
                with a newer vehicle. Hence mobility is not lost, but rather
                transferred between vehicles--and with it, the associated
                benefits.\2042\ In the limited instances where a retired vehicle is not
                replaced with a newer vehicle, that action is freely taken and the
                agencies can infer from that decision that the benefit derived from
                scrapping the vehicle outweighed any possible loss, including lost
                mobility. Offsetting the reduction in scrappage safety costs--realized
                because of the standards--without a complementary benefit would be
                directionally inconsistent.\2043\
                ---------------------------------------------------------------------------
                 \2041\ If the benefit of driving an older vehicle was higher
                than the benefit of driving a newer vehicle, we would anticipate
                consumers to forgo replacing older vehicles with newer vehicles.
                 \2042\ Since driving newer vehicles, including newer used
                vehicles, likely confers greater benefits than would-be scrapped
                vehicle, the agencies are likely underestimating the value of
                increased scrappage.
                 \2043\ A similar argument could be made that consumers
                `internalize' additional fuel costs, and therefore pre-tax fuel
                savings should also be offset. However, this would also ignore that
                benefits are remaining constant while the costs to obtain those
                benefits is increasing.
                ---------------------------------------------------------------------------
                 The agencies reaffirm that off-setting safety costs attributable to
                the sales and scrappage effects is inappropriate. Commenters' arguments
                relied exclusively on the premise that driving older vehicles is freely
                chosen and thus must have associated benefits, without considering the
                impact of accelerating their retirement on the rule's overall net
                safety and mobility benefits. Furthermore, the agencies remain
                concerned that potential buyers may be ``frozen out'' of the new
                vehicle market by prohibitively high prices; in which case enabling
                access to newer, safer vehicles provides measurable safety benefits
                that should be considered by the analysis.
                 However, in an abundance of caution, the agencies performed a
                sensitivity analysis that applies the same safety offset to sales/
                scrappage safety impacts that was applied to the rebound effect safety
                impacts. The results are provided in Table VI-244 below. As might be
                expected, this adjustment reduces net benefits in all scenarios, but
                does not substantially shift the relative scope among alternatives.
                BILLING CODE 4910-59-P
                [[Page 24800]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.454
                BILLING CODE 4910-59-C
                 Again, the agencies feel that this offset is inappropriate. The
                sensitivity case disregards many of the tangible gains in safety
                expected from increased sales and
                [[Page 24801]]
                scrappage. Furthermore, the agencies note that--even if they replaced
                the central analysis' assumptions with this sensitivity case--the
                anticipated changes in net benefits would not be enough to change their
                decision.
                (2) Revised Sales-Scrappage Safety Model
                 In response to the comments, the agencies have taken several steps
                to revise the sales-scrappage safety model. First, the agencies
                developed a revised statistical model to explain historical
                improvements in the lifetime safety performance of each successive new
                vintage of cars and light trucks, and used the results of this improved
                model to project the future trend in the overall fatality rates. While
                the revised historical trend model itself is more complex than the one
                utilized in the proposal, the overall procedure is simpler; the
                agencies have collapsed the two piecemeal components discussed above
                into one model and eliminated the need to `reconcile' differences
                between competing future projections. Next, the agencies applied
                detailed empirical estimates of the market uptake and improving
                effectiveness of crash avoidance technologies to estimate their effect
                on the fleet-wide fatality rate, including explicitly incorporating
                both the direct effect of those technologies on the crash involvement
                rates of new vehicles equipped with them, as well as the ``spillover''
                effect of those technologies on improving the safety of occupants of
                vehicles that are not equipped with these technologies.
                (a) Crash Avoidance
                 In the NPRM, the agencies took a very generalized approach to
                estimating the pace of future safety trends. For reasons discussed
                above, the agencies noted that there was uncertainty regarding actual
                trends in fatality rates. This issue was addressed by numerous
                commenters who took opposing positions. Among them, IPI stated that
                ``[t]he agencies have not provided an adequate explanation for why past
                safety trends are likely to continue until the mid-2020s.'' IPI further
                noted that ``crash avoidance technology may not be adopted as easily or
                readily as crash mitigation technologies have been.'' \2044\ In
                response, the agencies note that the trend the agencies adopted for the
                NPRM was not a direct continuation of past trends. Rather, it was a
                simple average of several possible models the agencies had examined,
                accepting each as an illustration of different and conflicting possible
                future scenarios.
                ---------------------------------------------------------------------------
                 \2044\ IPI, Appendix, NHTSA-2018-0067-12213, at 98.
                ---------------------------------------------------------------------------
                 By contrast, States and Cities asserted that fatality rates may be
                lower in the future than the agencies estimated, noting that the NPRM
                analysis did not ``account for safety benefits that new safety
                technologies in future vehicles will have on the agencies predicted
                outcome.'' \2045\ While the agencies agree that the NPRM analysis did
                not analyze individual safety benefits of new technologies, the trends
                included in the NPRM were intended, in part, as a proxy estimate of the
                impact of these technologies. As discussed in the NPRM, these
                technologies were cited as a justification for assuming a continued
                downward trend in the fatality rate through roughly 2035.
                ---------------------------------------------------------------------------
                 \2045\ States and Cities, Detailed Comments, NHTSA-2018-0067-
                11735, at 80.
                ---------------------------------------------------------------------------
                 Nonetheless, the agencies believe that further analysis of these
                potential trends can now be ascertained for several explicit
                technologies. In response to comments suggesting that the agencies
                account more directly for new safety technologies, the agencies
                augmented the sales-scrappage safety analysis for the final rule with
                recent research into the effectiveness of specific advanced crash
                avoidance safety technologies (also known as ADAS or advanced driver
                assistance systems) that are expected to drive future safety
                improvement to estimate the impacts of crash avoidance technologies.
                The analysis analyzes six crash avoidance technologies that are
                currently being produced and commercially deployed in the new vehicle
                fleet. These include Frontal Collision Warning (FCW), Automatic
                Emergency Braking (AEB), Lane Departure Warning (LDW), Lane Keep Assist
                (LKA), Blind Spot Detection (BSD), and Lane Change Alert (LCA).\2046\
                These are the principal technologies that are being developed and
                adopted in new vehicle fleets and will likely drive vehicle-based
                safety improvements for the coming decade. These technologies are being
                installed in more and more new vehicles; in fact, 12 manufacturers
                recently reported that they voluntarily installed AEB systems in more
                than 75 percent of their new vehicles sold in the year ending August
                31, 2019.\2047\ The agencies note that the terminology and the detailed
                characteristics of these systems may differ across manufacturers, but
                the basic system functions are common across all.
                ---------------------------------------------------------------------------
                 \2046\ A full description of these technologies and several
                other technologies referenced below may be found in the
                corresponding FRIA safety impacts discussion.
                 \2047\ NHTSA Announces Update to Historic AEB Commitment by 20
                Automakers, NHTSA press release December 17, 2019. https://www.nhtsa.gov/press-releases/nhtsa-announces-update-historic-aeb-commitment-20-automakers.
                ---------------------------------------------------------------------------
                 These six technologies address three basic crash scenarios through
                warnings to the driver or alternately, through dynamic vehicle control:
                 1. Forward collisions, typically involving a crash into the rear of
                a stopped vehicle;
                 2. Lane departure crashes, typically involving inadvertent drifting
                across or into another traffic lane; and
                 3. Blind spot crashes, typically involving intentional lane changes
                into unseen vehicles driving in or approaching the driver's blind spot.
                 Unlike traditional safety features where the bulk of the safety
                improvements were attributable to improved protection when a crash
                occurs (crash worthiness), the impact of advanced crash avoidance
                technologies (ADAS or advanced driver assistance systems) will have on
                fatality and injury rates is a direct function of their effectiveness
                in preventing or reducing the severity of the crashes they are designed
                to mitigate. This effectiveness is typically measured using real world
                data comparing vehicles with these technologies to similar vehicles
                without them. While these technologies are actively being deployed in
                new vehicles, their penetration in the larger on-road vehicle fleet has
                been at a low, but growing level. This limits the precision of
                statistical regression analyses, at least until the technologies become
                more common in the on-road fleet.
                 Our approach in the final rule is to derive effectiveness rates for
                these advanced crash-avoidance technologies from safety technology
                literature. The agencies then apply these effectiveness rates to
                specific crash target populations for which the crash avoidance
                technology is designed to mitigate and adjusted to reflect the current
                pace of adoption of the technology, including the public commitment by
                manufactures to install these technologies. The products of these
                factors, combined across all 6 advanced technologies, produce a
                fatality rate reduction percentage that is applied to the fatality rate
                trend model discussed below, which projects both vehicle and non-
                vehicle safety trends. The combined model produces a projection of
                impacts of changes in vehicle safety technology as well as behavioral
                and infrastructural trends.
                [[Page 24802]]
                (i) Technology Effectiveness Rates
                (a) Forward Crash Collision Technologies
                 For forward collisions, manufacturers are currently equipping
                vehicles with FCW, which warns drivers of impending collisions, as well
                as AEB, which incorporates the sensor systems from FCW together with
                dynamic brake support (DBS) and crash imminent braking (CIB) to help
                avoid crashes or mitigate their severity. Manufacturers have committed
                voluntarily to install some form of AEB on all light vehicles by the
                2023 model year (September 2022).\2048\
                ---------------------------------------------------------------------------
                 \2048\ See https://www.nhtsa.gov/press-releases/nhtsa-iihs-announcement-aeb.
                ---------------------------------------------------------------------------
                 Table VI-245 summarizes studies which have measured effectiveness
                for various forms of FCW and AEB over the past 13 years. Most studies
                focused on crash reduction rather than injury reduction. This is a
                function of limited injury data in the on-road fleet, especially during
                the early years of deployment of these technologies. In addition, it
                reflects engineering limitations in the technologies themselves.
                Initial designs of AEB systems were basically incapable of detecting
                stationary objects at speeds higher than 30 mph, making them
                potentially ineffective in higher speed crashes that are more likely to
                result in fatalities or serious injury. For example, Wiacek et al. (2-
                15) conducted a review of rear-end crashes involving a fatal occupant
                in the 2003-2012 NASS-CDS data-bases to determine the factors that
                contribute to fatal rear-end crashes.\2049\ They found that the speed
                of the striking vehicle was the primary factor in 71 percent of the
                cases they examined. The average Delta-V of the striking vehicle in
                these cases was 46 km/h (28.5 mph), implying pre-crash travel speeds in
                excess of this speed. While Table VI-245 includes studies going back to
                2005, the agencies focus our discussion on more recent studies
                conducted after 2012 in order to reflect more current safety systems
                and vehicle designs.2050 2051 2052 2053 2054 2055 2056 2057
                ---------------------------------------------------------------------------
                 \2049\ Wiacek, C., Bean, J., Sharma, D., Real World Analysis of
                Fatal Rear-End Crashes, National Highway Traffic Safety
                Administration, 24th Enhanced Safety of Vehicles Conference, 150270,
                2015.
                 \2050\ Sugimoto, Y., and Sauer, C., (2005). Effectiveness
                Estimation Method for Advanced Driver Assistance System and its
                Application to Collision Mitigation Brake systems, paper number 05-
                148, 19th International Technical Conference on the Enhanced safety
                of Vehicles (ESV), Washington DC, June 6-9, 2005.
                 \2051\ Page, Y., Foret-Bruno, J., & Cuny, S. (2005). Are
                expected and observed effectiveness of emergency brake assist in
                preventing road injury accidents consistent?, 19th ESV Conference,
                Washington DC.
                 \2052\ Najm, W.G., Stearns, M.D., Howarth, H., Koopman, J. &
                Hitz, J., (2006). Evaluation of an Automotive Rear-End Collision
                Avoidance System (technical report DOT HS 810 569), Cambridge, MA:
                John A. Volpe National Transportation System Center, U.S. Department
                of Transportation.
                 \2053\ Breuer, JJ., Faulhaber, A., Frank, P. and Gleissner, S.
                (2007). Real world Safety Benefits of Brake Assistance Systems,
                Proceedings of the 20th International Technical Conference of the
                Enhanced Safety of Vehicles (ESV) in Lyon, France June 18-21, 2007.
                 \2054\ Keuhn, M., Hummel, T., and Bende J., Benefit estimation
                of advanced driver assistance systems for cars derived from real-
                world accidents, Paper No. 09-0317, 21st International Technical
                Conference on the Enhanced Safety of Vehicles (ESV)--International
                Congress Centre, Stuttgart, Germany, June 15-18, 2009.
                 \2055\ Grover, C., Knight, I., Okoro, F., Simmons I., Couper,
                G., Massie, P., and Smith, B. (2008). Automated Emergency Brake
                Systems: Technical requirements, Costs and Benefits, PPR227, TRL
                Limited, DG Enterprise, European Commission, April 2008.
                 \2056\ Kusano, K.G., and Gabler, H.C. (2015). Comparison of
                Expected Crash Injury and Injury Reduction from Production Forward
                Collision and Lane Departure Warning Systems, Traffic Injury
                Prevention 2015; Suppl. 2: S109-14.
                 \2057\ HLDI (2011). Volvo's City Safety prevents low-speed
                crashes and cuts insurance costs, Status Report, Vol. 46, No. 6,
                July 19,2011.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.455
                [[Page 24803]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.456
                 Doecke et al. (2012) created
                2058 2059 2060 2061 2062 2063 simulations of 103 real world
                crashes and applied AEB system models with differing specifications to
                determine the change in impact speed that various AEB interventions
                might produce. Their modeling found significant rear-end crash speed
                reductions with various AEB performance assumptions. In addition, they
                estimated a 29 percent reduction in rear-end crashes and that 25
                percent of crashes over 10 km/h were reduced to 10 km/h or less.
                ---------------------------------------------------------------------------
                 \2058\ Docke, S.D., Anderson, R.W.G., Mackenzie, J.R.R., Ponte,
                G. (2012). The potential of autonomous emergency braking systems to
                mitigate passenger vehicle crashes. Australian Road Safety Research
                Policing and Education Conference, October 4-6, 2012, Wellington,
                New Zealand.
                 \2059\ Chauvel, C., Page, Y., Files, B.N., and Lahausse, J.
                (2013). Automatic emergency braking for pedestrians effective target
                population and expected safety benefits, Paper No. 13-0008, 23rd
                International Technical Conference on the Enhanced Safety of
                Vehicles (ESV), Seoul, Republic of Korea, May 27-30, 2013.
                 \2060\ Fildes B., Keall M., Bos A., Lie A., Page, Y., Pastor,
                C., Pennisi, L., Rizzi, M., Thomas, P., and Tingvall, C.
                Effectiveness of Low Speed Autonomous Emergency Braking in Real-
                World Rear-End Crashes. Accident Analysis and Prevention, AAP-D-14-
                00692R2.
                 \2061\ Cicchino, J.B. (2017). Effectiveness of forward collision
                warning and autonomous emergency braking systems in reducing front-
                to-rear crash rates. Accident Analysis and Prevention, V. 99, Part
                A, February 2017, Pages 142-52.
                 \2062\ Kusano, K.D., and Gabler H.C. (2012). Safety Benefits of
                Forward Collision Warning, Brake Assist, and Autonomous Braking
                Systems in Rear-End Collisions, Intelligent Transportation Systems,
                IEEE Transactions, Volume 13 (4).
                 \2063\ Leslie, A, Kiefer, R., Meitzner, M, and Flannagan, C.
                (2019). Analysis of the Field Effectiveness of General Motors
                Production Active Safety and Advanced headlighting Systems.
                University of Michigan Transportation Research Institute, UMTRI-
                2019-6, September, 2019.
                ---------------------------------------------------------------------------
                 Cicchino (2016) analyzed the effectiveness of a variety of forward
                collision mitigation systems including both FCW and AEB systems.
                Cicchino used a Poisson regression to compare rates of police-reported
                crashes per insured vehicle year between vehicles with these systems
                and the same models that did not elect to install them. The analysis
                was based on crashes occurring during 2010 to 2014 in 22 States and
                controlled for other factors that affected crash risk. Cicchino found
                that FCW reduced all rear-end striking crashes by 27 percent and rear-
                end striking injury crashes by 20 percent, and that AEB functional at
                high-speeds reduced these crashes by 50 and 56 percent, respectively.
                She also found that low speed AEB without driver warning reduced all
                crashes by 43 percent and injury crashes by 45 percent. She also found
                that even low-speed AEB could impact crashes at higher speed limits.
                Reductions were found of 53 percent, 59 percent, and 58 percent for all
                rear-end striking crash rates, rear-end striking injury crash rates,
                and rear-end third party injury crash rates, respectively, at speed
                limits of 40-45 mph. For speed limits of 35 mph or less, reductions of
                40 percent, 40 percent, and 43 percent were found. For speed limits of
                50 mph or greater, reductions of 31 percent, 30 percent, and 28
                percent, were found. Further, Cicchino (2016) found significant
                reductions (30 percent) in rear-end injury crashes even in crashes on
                roadways where speed limits exceeded 50 mph.
                 Kusano and Gabler (2012) examined the effectiveness of various
                levels of forward collision technologies including FCW and AEB based on
                simulations of 1,396 real world rear end crashes from 1993-2008 NASS
                CDS data-bases. The authors developed a probability-based framework to
                account for variable driver responses to the warning systems. Kusano
                and Gabler found FCW systems could reduce rear-end crashes by 3.2
                percent and driver injuries in rear-end crashes by 29 percent. They
                also found that full AEB systems with FCW, pre-crash brake assist, and
                autonomous pre-crash braking could reduce rear-end crashes by 7.7
                percent and reduce moderate to fatal driver injuries in rear-end
                crashes by 50 percent.
                 Fildes et al. (2015) performed meta-analyses to evaluate the
                effectiveness of low-speed AEB technology in passenger vehicles based
                on real-world crash experience across six different predominantly
                European countries. Data from these countries was pooled into a
                standard analysis format and induced exposure methods were used to
                control for extraneous effects. The study found a 38 percent overall
                reduction in rear-end crashes for vehicles with AEB compared to similar
                vehicles without this technology. The study also found no statistical
                evidence for any difference in effectiveness between urban roads with
                speed limits less than or equal to 60 km/h, and rural roads with speed
                limits greater than 60 km/h. Fildes et al. (2015) found no statistical
                difference in the performance of AEBs on lower speed urban or higher
                speed rural roadways.
                 Kusano and Gabler (2015) simulated rear-end crashes based on a
                sample of 1,042 crashes in the 2012 NASS-CDS. Modelling was based on 54
                model year 2010-2014 vehicles that were evaluated in NHTSA's New Car
                Assessment Program (NCAP). Kusano and Gabler found FCW systems could
                prevent 0-67 percent of rear-end crashes and 2-69 percent of serious to
                fatal driver injuries.
                 Leslie et al. (2019) analyzed the relative crash performance of
                123,377 General Motors (GM) MY 2013 to 2017 vehicles linked to State
                police-reported crashes by Vehicle Identification numbers (VIN). GM
                provided VIN-linked safety content information for these vehicles to
                enable precise identification of safety technology content. The authors
                analyzed the effectiveness of a variety of crash avoidance technologies
                including both FCW and AEB separately. They estimated effectiveness
                comparing system-relevant crashes to baseline
                [[Page 24804]]
                (control group) crashes using a quasi-induced exposure method in which
                rear-end struck crashes are used as the control group. Leslie et al.
                found that FCW reduced rear-end striking crashes of all severities by
                21 percent, and that AEB (which includes FCW) reduced these crashes by
                46 percent.\2064\
                ---------------------------------------------------------------------------
                 \2064\ The agencies note that UMTRI, the sponsoring organization
                for the Leslie et al. study, published a previous version of this
                same study utilizing the same methods in March of 2018 (Flannagan,
                C. and Leslie, A, Crash Avoidance Technology Evaluation Using real-
                World crashes, University of Michigan Transportation research
                Institute, March 22, 2018). The agencies focused on the more recent
                2019 study because its sample size is significantly larger and it
                represents more recent model year vehicles. The revised (2019) study
                uses the same basic techniques but incorporated a larger data-base
                of system-relevant and control cases (123,377 cases in the 2019
                study vs. 35,401 in the 2018 study). Relative to the Flannagan and
                Leslie (2018) findings, the results of the 2019 study varied by
                technology. The revised study found effectiveness rates of 21% for
                FCW and 46% for AEB, compared to 16% and 45% in the 2018 study. The
                revised study found effectiveness rates of 10% for LDW and 20% for
                LKA, compared to 3% and 30% for these technologies in the 2018
                study. The revised study found effectiveness rates of 3% for BSD and
                26-37% for LCA systems, compared to 8% and 19-32% for these
                technologies in the 2018 study. Thus, some system effectiveness
                estimates increased while others decreased.
                ---------------------------------------------------------------------------
                 For this analysis, the agencies based their projections on Leslie
                et al. because they are the most recent study, and thus reflect the
                most current versions of these systems in the largest number of
                vehicles, and also because they arguably have the most precise
                identification of the presence of the specific technologies in the
                vehicle fleet. Furthermore, Leslie et al. was the only study to report
                estimates for each of the six crash avoidance technologies analyzed for
                the final rule, hence providing a certain level of consistency amongst
                estimates. The agencies recognize that there is uncertainty in
                estimates of these technologies effectiveness, especially at this early
                stage of deployment. For this reason, the agencies examine a range of
                effectiveness rates to estimate boundary outcomes in a sensitivity
                analysis.
                 Leslie et al. measured effectiveness against all categories of
                crashes, but did not specify effectiveness against crashes that result
                in fatalities or injuries. The agencies examined a range of
                effectiveness rates against fatal crashes using a central case based on
                boundary assumptions of no effectiveness and full effectiveness across
                all crash types. Our central case is thus a simple average of these two
                extremes. Sensitivity cases were based on the 95th percent confidence
                intervals calculated from this central case. Leslie et al. found
                effectiveness rates of 21 percent for FCW and 46 percent for AEB. Our
                central fatality effectiveness estimates will thus be 10.5 percent for
                FCW and 23 percent for AEB. The calculated 95th percentile confidence
                limits range is 8.11 to 12.58 percent effective for FCW and 20.85 to
                25.27 for AEB. The agencies note that our central estimate is
                conservative compared to averages of those studies that did
                specifically examine fatality impacts; that is, the analysis assumes
                reduced future fatalities less than most of, or the average of, those
                studies, and thus minimizes the estimate of lives saved under
                alternatives to the augural standards. Furthermore, the agencies note
                that the estimates against fatal crashes is higher in the recent
                studies in Table VI-245, which reflects the agencies' understanding
                that earlier iterations of AEB and FCW may have been less effective
                against crashes that result in fatalties than newer and improved
                versions.\2065\
                ---------------------------------------------------------------------------
                 \2065\ As an example of improvements, the agencies note that the
                Mercedes system described in their 2015 owner's manual specified
                that for stationary objects the system would only work in crashes
                below 31 mph, but that in their manual for the 2019 model, the
                systems are specified to work in these crashes up to 50 mph.
                ---------------------------------------------------------------------------
                (b) Lane Departure Crash Technologies
                 For lane departure crashes, manufacturers are currently equipping
                vehicles with lane departure warning (LDW), which monitors lane
                markings on the road and alerts the driver when their vehicle is about
                to drift beyond a delineated edge line of their current travel lane, as
                well as lane keep assist (LKA), which provides gentle steering
                adjustments to help drivers avoid unintentional lane crossing. Table
                VI-246 summarizes studies which have measured effectiveness for LDW and
                LKA.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.457
                 Cicchino (2018) examined crash involvement rates per insured
                vehicle 2066 2067 2068 2069 2070 year for
                [[Page 24805]]
                vehicles that offered LDW as an option and compared crash rates for
                those that had the option installed to those that did not. The study
                focused on single-vehicle, sideswipe, and head-on crashes as the
                relevant target population for LDW effectiveness rates. The study
                examined 5,433 relevant crashes of all severities found in 2009-2015
                police-reported data from 25 States. The study was limited to crashes
                on roadways with 40 mph or greater speed limits not covered in ice or
                snow since lower travel speeds would be more likely to fall outside of
                the LDW systems' minimum operational threshold. Cicchino found an
                overall reduction in relevant crashes of 11 percent for vehicles that
                were equipped with LDW. She also found a 21 percent reduction in injury
                crashes. The result for all crashes was statistically significant,
                while that for injury crashes approached significance (p2071 2072 2073
                ---------------------------------------------------------------------------
                 \2071\ Cicchino, J.B. (2017b). Effects of blind spot monitoring
                systems on police-reported lane-change crashes. Insurance Institute
                for Highway Safety, August 2017.
                 \2072\ Leslie et al., supra note 2063.
                 \2073\ Isaksson-Hellman, I., Lindman, M., An evaluation of the
                real-world safety effect of a lane change driver support system and
                characteristics of lane change crashes based on insurance claims.
                Traffic Injury Prevention, February 28, 2018: 19 (supp. 1).
                ---------------------------------------------------------------------------
                [[Page 24806]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.458
                [GRAPHIC] [TIFF OMITTED] TR30AP20.459
                 Cicchino (2017) used Poisson regression to compare crash
                involvement rates per insured vehicle year in police-reported lane-
                change crashes in 26[thinsp]U.S. States during 2009-2015 between
                vehicles with blind spot monitoring and the same vehicle models without
                the optional system, controlling for other factors that can affect
                crash risk. Systems designs across the 10 different manufacturers
                included in the study varied regarding the extent to which the size of
                the adjacent lane zone that they covered exceeded the blind spot area,
                speed differentials at which vehicles could be detected, and their
                ability to detect rapidly approaching vehicles, but these different
                systems were not examined separately. The study examined 4,620 lane
                change crashes, including 568 injury crashes. Cicchino found an overall
                reduction of 14 percent in blind spot related crashes of all
                severities, with a non-significant 23 percent reduction in injury
                crashes.
                 Leslie et al. (2019) analyzed the relative crash performance of
                123,377 2013-2017 General Motors (GM) vehicles linked to State police-
                reported crashes by Vehicle Identification numbers (VIN). GM provided
                VIN-linked safety content information for these vehicles to enable
                precise identification of safety technology content. The authors
                analyzed the effectiveness of a variety of crash avoidance technologies
                including both BSD and LCA separately. They estimated effectiveness
                comparing system-relevant crashes to baseline (control group) crashes
                using a quasi-induced exposure method in which rear-end struck crashes
                are used as the control group. Flannagan and Leslie found that BSD
                reduced lane departure crashes of all severities by 3 percent (non-
                significant), and that LCA (which includes BSD) reduced these crashes
                by 26 percent.
                 Isaksson-Hellman and Lindman (2018) evaluated the effect of the
                Volvo Blind Spot Information System (BLIS) on lane change crashes.
                Volvo's BLIS functions as an LCA, detecting vehicles approaching the
                blind spot as well as those already in it. The authors analyzed crash
                rate differences in lane change situations for cars with and without
                the BLIS system based on a population of 380,000 insured vehicle years.
                The authors found the BLIS system did not significantly reduce the
                overall number of lane change crashes of all severities, but they did
                find a significant 31 percent reduction in crashes with a repair cost
                exceeding $1250, and a 30 percent lower claim cost across all lane
                change crashes, indicating a reduced crash severity effect.
                 Like lane change technologies, blind spot technologies are
                operational at travel speeds where fatalities are likely to occur. The
                agencies therefore assume the Leslie et al. crash reduction estimates
                are generally applicable to all crash severities, including fatal
                crashes. Our central effectiveness estimates are thus 3 percent for BSD
                and 26 percent for LCA. For sensitivity analysis, the agencies adopt
                the 95 percent confidence intervals from Flannagan & Leslie. For LCA
                this range is 16.59-33.74 percent. For BSD, the upper range was 14.72
                percent, but the findings were not statistically significant. The
                agencies therefore limit the range to 0-14.72 percent.
                 Table VI-248 summarizes the effectiveness rates calculated in
                Leslie et al. and used in this analysis. Differences between the rates
                listed as ``Used in CAFE Fatality Analysis'' and those computed from
                Leslie et al. are explained in the above discussion.
                [[Page 24807]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.460
                (ii) Target Populations for Crash Avoidance Technologies
                 The impact on fatality rates that will occur due to these
                technologies will be a function of both their effectiveness rate and
                the portion of occupant fatalities that occur under circumstances that
                are relevant to the technologies function. The agencies base our target
                population estimates on a recent study that examined these portions
                specifically for a variety of crash avoidance technologies including
                those analyzed here. Wang (2019) documented target populations for five
                groups of collision avoidance technologies in passenger vehicles
                including forward collisions, lane keeping, blind zone detection,
                forward pedestrian impact, and backing collision avoidance. The first
                three of these affect the light occupant target population examined in
                this analysis. Wang separately examined crash populations stratified by
                severity including fatal injuries, non-fatal injuries, and property
                damaged only (PDO) vehicles. She based her analysis on 2011-2015 data
                from NHTSA's Fatality Analysis Reporting System (FARS), National
                Automotive Sampling System (NASS), and General Estimates System (GES).
                FARS data was the basis for fatal crashes while nonfatal injuries and
                PDOs were derived from the NASS and GES.
                 Wang followed the pre-crash typology concept initially developed by
                the Volpe National Transportation Systems Center (Volpe). Under this
                concept, crashes are categorized into mutually exclusive and distinct
                scenarios based on vehicle movements and critical events occurring just
                prior to the crash. Table VI-249 summarizes the portion of total annual
                crashes and injuries for each crash severity category that is relevant
                to the three crash scenarios examined.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.461
                 The relevant proportions vary significantly depending on the
                severity of the crash. The rear-end crashes that are addressed by FCW
                and AEB technologies tend to be low-speed crashes and thus account for
                a larger portion of non-fatal injury and PDO crashes than for
                fatalities. Only 4 percent of fatal crashes occur in front-to-rear
                crashes, but over 30 percent of nonfatal crashes are this type. By
                contrast, fatal crashes are highly likely to involve inadvertent lane
                departure, 44 percent of all light vehicle occupant fatalities occur in
                crashes that involve lane departure, but only 17 percent of non-fatal
                injuries and 12 percent of PDOs involve this crash scenario. Blind spot
                crashes account for only about 2 percent of fatalities, 7 percent of
                MAIS1-5 injuries, and 12 percent of PDOs.
                 The target population of this analysis is occupants of the light
                vehicles subject to CAFE. The values in Table VI-249 are portions of
                all crashes that occur annually. These include crashes of motor
                vehicles not subject to the current CAFE rulemaking such as medium and
                large trucks, buses, motorcycles, bicycles, etc. To adjust for this,
                the values in Wang were normalized to represent their portion of all
                light passenger vehicle (PV) crashes, rather than all crashes of any
                type. Wang provides total PV fatalities consistent with her technology
                numbers which are used as a baseline for this process. Based on 2011-
                2015 FARS data, Wang
                [[Page 24808]]
                found an average of 29,170 PV occupant fatalities occurred annually.
                 A second adjustment to Wang's results was made to make them
                compatible with the effectiveness estimates found in Leslie et al. In
                her target population estimate for lane departure warning, Wang
                included both head-on collisions and rollovers, but Leslie et al. did
                not. The Leslie et al. effectiveness rate is thus applicable to a
                smaller target population than that examined by Wang. To make these
                numbers more compatible, counts for these crash types were removed from
                Wang's lane departure totals.
                 Electronic Stability Control (ESC) has been standard equipment in
                all light vehicles in the U.S. since the 2012 model year. ESC is highly
                effective in reducing roadway departure and traction loss crashes, and
                although it will be present in all future model year vehicles, it was
                present in only about 30 percent of the 2011-2015 on-road fleet
                examined by Wang. To reflect the impact of ESC on future on-road fleets
                therefore, the agencies further adjusted Wang's numbers to reflect a
                100 percent ESC presence in the on-road fleet. The agencies allocated
                the reduced roadway departure fatalities to the LDW target population,
                and the reduced traction loss fatalities to the AEB target population.
                This has the effect of reducing the total fatalities in both groups as
                well as in the total projected fatalities baseline.
                 Table VI-250 summarizes the revised incidence counts and re-
                calculated proportions of total PV occupant crash/injury. Revised
                totals are derived from original totals referenced in Table 1-3 in Wang
                (2019).
                [GRAPHIC] [TIFF OMITTED] TR30AP20.462
                (iii) Fleet Penetration Schedules
                 The third element of the rule's safety projections is the fleet
                technology penetration schedules. Advanced safety technologies (ADAS)
                will only influence the safety of future MY fleets to the extent that
                they are installed and used in those fleets. These technologies are
                already being installed on some vehicles to varying degrees, but the
                agencies expect that over time, they will become standard equipment due
                to some combination of market pressure and/or safety regulation. The
                agencies adopt this assumption based on the history of most previous
                vehicle safety technologies, which are now standard equipment on all
                new vehicles sold in the U.S.
                 The pace of technology adoption is estimated based on a variety of
                factors, but the most fundamental is the current pace of adoption in
                recent years. These published data were obtained from Ward's Automotive
                Reports for each technology.\2074\ Since these technologies are
                relatively recent, only a few years of data--typically 2 or 3 years--
                were available from which to derive a trend. This makes these
                projections uncertain, but under these circumstances, a continuation of
                the known trend is the baseline assumption, which the agencies modify
                only when there is a rationale to justify it.
                ---------------------------------------------------------------------------
                 \2074\ Derived from Ward's Automotive Yearbooks, 2014 through
                2018, % Factory Installed Electronic ADAS Equipment tables,
                weighting domestic and imported passenger cars and light trucks by
                sales volume.
                ---------------------------------------------------------------------------
                 The technologies were examined in pairs reflecting their mutual
                target populations. Both FCW and AEB affect the same target
                population--frontal collisions. Both systems have been installed in
                some current MY vehicles, but their relative paces are expected to
                diverge significantly due to a formal agreement brokered by NHTSA and
                IIHS involving nearly all auto manufacturers, to have AEB installed in
                100 percent of their vehicles by September 2022 (MY 2023).\2075\ Wards
                first published installation rates for FCW and AEB for the 2016 model
                year and as of this analysis the 2017 MY is the latest data they have
                published. The agencies thus have data indicating that FCW was
                installed in 17.6 percent of MY 2016 vehicles and 30.5 percent of MY
                2017 vehicles. AEB was installed in 12.0 percent of MY 2016 vehicles
                and 27.0 percent of MY 2017 vehicles. AEB was installed in 12.0 percent
                of MY 2016 vehicles and 27.0 percent of MY 2017 vehicles. More recent
                reports submitted by manufacturers to the Federal Register indicate
                that installation rates accelerated in MY 2018 and 2019
                [[Page 24809]]
                vehicles. Four manufacturers, Tesla, Volvo, Audi, and Mercedes, have
                already met their voluntary commitment of 100 percent installation 3
                years ahead of schedule. During the period September 1, 2018 through
                August 31, 2019, 12 of the 20 manufacturers equipped more than 75
                percent of their new passenger vehicles with AEB, and overall
                manufacturers equipped more than 9.5 million new passenger vehicles
                with AEB.\2076\
                ---------------------------------------------------------------------------
                 \2075\ See https://www.nhtsa.gov/press-releases/nhtsa-iihs-announcement-aeb.
                 \2076\ See NHTSA Announces Update to Historic AEB Commitment by
                20 Automakers. December 17, 2019. https://www.nhtsa.gov/press-releases/nhtsa-announces-update-historic-aeb-commitment-20-automakers.
                ---------------------------------------------------------------------------
                 Because of the NHTSA/IIHS agreement, the agencies assume that AEB
                will be in 100 percent of light vehicles by the 2023 MY. To derive
                installation rates for MYs 2018 through 2022, the agencies interpolate
                between the MY 2017 rate of 27 percent and the MY 2023 rate of 100
                percent. To derive a MY 2015 estimate, the agencies modelled the
                results for MYs 2016-2023 and calculated a value for year x=0,
                essentially extending the model results back one year on the same
                trendline.
                 For FCW, the agencies used the same interpolation/modeling method
                as was used for AEB to derive an initial baseline trend. However, while
                both systems are available on some portion of the current MY fleet, the
                agencies anticipate that by MY 2023, all vehicles will have AEB systems
                that essentially encompass both FCW and AEB functions. The agencies
                therefore project a gradual increase in both systems until the sum of
                both systems penetration rates exceeds 100 percent. At that point, the
                agencies project a gradual decrease in FCW only installations until FCW
                only systems are completely replaced by AEB systems in MY 2023.
                 For LDW, Wards penetration data were available as far back as MY
                2013, giving a total of 5 data points through MY 2017. The projection
                for LDW was derived by modelling these data points. The data indicate a
                near linear trend and our initial projections of future years were
                derived directly from this model. Wards did not report any of the more
                advanced LKA systems until MY 2016, leaving only 2 data points. The
                agencies modelled a simple trendline through these data points to
                estimate the pace of future LKA installations. As with Frontal crashes,
                the agencies assume a gradual phase-in of the most effective
                technology, LKA, will eventually replace the lesser technology, LDW,
                and the agencies allow gradual increases in both systems penetration
                until their sum exceeds 100 percent, at which point LDW penetration
                begins to decline to zero while LKA penetration climbs to 100 percent.
                 For blind spot crashes, Wards data was available for MYs 2013-2017
                for BSD, but no data was available to distinguish LCA systems. LCA
                systems were available as optional equipment on at least 10 MY 2016
                vehicles.\2077\ In addition, Flannagan and Leslie found numerous cases
                in State data-bases involving vehicles with LCA. Because LCA data is
                not specifically identified, the agencies will estimate its frequency
                based on the samples found in Flannagan & Leslie. In that study, 62
                percent of vehicles with blind spot technologies has BSD alone, while
                38 percent had LCA (which includes BSD). The agencies employ this ratio
                to establish the relative frequency of these technologies in our
                projections. As with frontal and lane change technologies, the agencies
                assume a gradual phase-in of the most effective technology, LCA, will
                eventually replace the lesser technology, BSD, and the agencies allow
                gradual increases in both systems penetration until their sum exceeds
                100 percent, at which point BSD penetration begins to decline to zero
                while LCA penetration climbs to 100 percent.
                ---------------------------------------------------------------------------
                 \2077\ See, e.g. https://www.autobytel.com/car-buying-guides/features/10-cars-with-lane-change-assist-using-cameras-or-sensors-130847.
                ---------------------------------------------------------------------------
                (iv) Impact Calculations
                 Table VI-251, Table VI-252, and Table VI-253 summarize the
                resulting estimates of impacts on fatality rates for frontal crash
                technologies, lane change technologies, and blind spot technologies
                respectively for MYs 2016-2035. All previously discussed inputs are
                shown in the tables. The effect of each technology is the product of
                its effectiveness, it's percent installation in the MY fleet, and the
                portion of the total light vehicle occupant target population that each
                technology might address. Since installation rates for each technology
                apply to different portions of the vehicle fleet (i.e., vehicles have
                either the more basic or more advanced version of the technology), the
                effect of the two technologies combined is a simple sum of the two
                effects. Likewise, since each crash type addresses a unique target
                population, there is no overlap among the three crash types and the sum
                of the normalized crash impacts across all three crash types represents
                the total impact on fatality rates from these 6 technologies for each
                model year. These cumulative results are shown in the last column of
                Table VI-253. As technologies phase in to newer MY fleets,\2078\ their
                impact on the light vehicle occupant fatality rate increases
                proportionally to roughly 8.5 percent before levelling off. That is,
                eventually, by approximately MY 2026, these technologies are expected
                to reduce fatalities and fatality rates for new vehicles by roughly 8.5
                percent below their initial baseline levels.
                ---------------------------------------------------------------------------
                 \2078\ While it is technically possible to retrofit these
                systems into the on-road fleet, such retrofits would be
                significantly more expensive than OEM installations. The agencies
                thus assume all on-road fleet penetration of these technologies will
                come through new vehicle sales.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24810]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.463
                [GRAPHIC] [TIFF OMITTED] TR30AP20.464
                [[Page 24811]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.465
                BILLING CODE 4910-59-C
                [[Page 24812]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.466
                (b) Fatality Trend Model
                 The revised fatality trend model differs from the model employed in
                the NPRM in four main respects:
                 The fatality rates for individual model years and ages
                were re-calculated to correct the counts of fatalities to occupants of
                light-duty vehicles and to reflect the revised VMT estimates, the
                latter of which incorporate revisions to both vehicle registration
                counts and the estimated relationship between vehicle age and annual
                use; \2079\
                ---------------------------------------------------------------------------
                 \2079\ These revised estimates of the number of miles traveled
                by vehicles of each model year during past calendar years were
                developed from the expanded sample of vehicles' odometer readings
                obtained by NHTSA.
                ---------------------------------------------------------------------------
                 In response to comments on the version used in the NPRM, t
                model adds controls for changes to factors (such as driver demographics
                and behavior, and geographic patterns of travel) that can affect
                fatality rates for vehicles of all model years and ages;
                 The revised analysis clusters past model years into
                ``safety cohorts,'' which are groups of successive model years that
                exhibit similar fatality rates during their first years of use, in
                order to represent the actual historical pattern of safety improvements
                more realistically; and
                 The model employs a slightly less complex mathematical
                relationship between a model year's age and its fatality rate
                (fatalities per mile driven), which still describes the observed
                relationship accurately.
                 Similar to the fatality trend model employed in the proposal, the
                revised estimates of annual travel were combined with tabulations of
                annual fatalities occurring among occupants of light-duty vehicles of
                each model year during past calendar years, tabulated from NHTSA's FARS
                data. Fatalities occurring in vehicles produced during each model year
                making up a calendar year's light-duty vehicle fleet are divided by the
                estimated number of miles they were driven during that calendar year to
                calculate historical fatality rates by model year and calendar year,
                measured as fatalities per billion miles traveled. These data represent
                the dependent variable in the revised statistical model of fatality
                rates.
                 Longitudinal or time-series analyses such as the model of
                historical variation in fatality rates for individual model years need
                to incorporate three separate effects to account for all potential
                sources of variation. First, they need to employ model year in some
                form as an explanatory variable, to account for improvements in the
                safety of vehicles produced during successive model years that persist
                throughout their lifetimes in the vehicle fleet. This is an example of
                a ``cohort effect'' in the age-period-cohort framework that is widely
                used to of analysis of population-wide behavior.\2080\ Second, such a
                model must account for the effect of age on the safety of each
                individual model year as it grows older, accumulates mileage, and in
                most cases changes ownership one or more times during its expected
                service lifetime (the ``aging effect'' in age-period-cohort analysis).
                ---------------------------------------------------------------------------
                 \2080\ For a detailed explanation of the rationale and methods
                for age-period-cohort analysis, see for example Columbia University
                Mailman School of Public Health, Population Health Methods: Age-
                Period-Cohort Analysis, available at https://www.mailman.columbia.edu/research/population-health-methods/age-period-cohort-analysis (accessed February 12, 2020); and Kupper,
                Lawrence L. et al., ``Statistical age-period-cohort analysis: A
                review and critique,'' Journal of Chronic Diseases 38:10 (1985), at
                811-830, available at https://www.sciencedirect.com/science/article/abs/pii/0021968185901055#! (accessed February 12, 2020).
                ---------------------------------------------------------------------------
                 Finally, most longitudinal analyses, including the historical
                safety model developed here, need to account explicitly for factors
                that vary over time--in this case, calendar years. By doing so, they
                can affect the safety of vehicles of all model years and ages making up
                the fleet during successive calendar years, or change the composition
                of total travel by vehicles of different model years and ages. In
                either case, such time-related factors--often referred to as ``period
                effects''--can change the overall safety performance of the entire
                fleet from one calendar year to the next, independently of and in
                addition to the changes that would result from the combination of new
                model years entering the fleet while older ones are retired from
                service (the cohort effect), and the aging of all model years making up
                the fleet. For example, an increase in seat belt use among all drivers
                during a calendar year would be expected to reduce the fatality rates
                of vehicles of all model years and ages in use during that year, while
                an economic recession may change the composition of drivers and
                vehicles on the road during a calendar year. In either case, one result
                will be a change in the fleet-wide composite fatality rate for that
                calendar year.
                 Figure VI-83 below illustrates the contributions of cohort, aging,
                and time-period effects to changes over time in population-wide
                behavior. As the figure indicates, these effects are conceptually
                independent, but interact in ways that combine to produce the observed
                historical evolution of the fleet-wide fatality rate for light-duty
                vehicle occupants. Again, calendar year or time-period factors can
                affect the safety performance of the entire fleet independently of the
                effect that would result from the combination of changes in the
                specific model years making up the fleet and the advancing ages of all
                model years, and any ``period effect'' effect attributable to factors
                that vary over time is in addition to cohort and aging effects.
                [[Page 24813]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.467
                 To introduce such period effects into the fatality trend model,
                which were absent from the NPRM analysis, the agencies obtained
                historical data on factors that varied by calendar year, and were
                expected to be responsible for such effects. As indicated previously,
                these included the following:
                 Seat belt use, as measured by the fraction of drivers
                observed to be wearing lap and shoulder belts, estimated by NHTSA's
                National Occupant Protection Survey (NOPUS);
                 Driving under the influence of alcohol or drugs, measured
                by the fraction of drivers reporting having recently done so in surveys
                conducted by the U.S. Centers for Disease Control (CDC); \2081\
                ---------------------------------------------------------------------------
                 \2081\ The agencies also experimented with measures of drivers
                appearing to be under the influence of alcohol or drugs included in
                NHTSA's NOPUS, available at https://crashstats.nhtsa.dot.gov/#/PublicationList/18.
                ---------------------------------------------------------------------------
                 Use of hand-held electronic devices, measured by the
                fraction of drivers visually observed to be doing so in NHTSA's NOPUS;
                 The fraction of licensed drivers who are male and under
                the age of 25 (historically the riskiest cohort of drivers), as
                reported by the FHWA's annual Highway Statistics publication; \2082\
                ---------------------------------------------------------------------------
                 \2082\ Federal Highway Administration, Highway Statistics,
                various years, Table DL-20, available at https://www.fhwa.dot.gov/policyinformation/statistics.cfm.
                ---------------------------------------------------------------------------
                 The fraction of miles traveled in rural areas, also as
                reported by FHWA; \2083\ and
                ---------------------------------------------------------------------------
                 \2083\ Federal Highway Administration, Highway Statistics,
                various years, Table VM-1, available at https://www.fhwa.dot.gov/policyinformation/statistics.cfm.
                ---------------------------------------------------------------------------
                 The overall performance of the U.S. economy, as measured
                by the annual rate of unemployment.\2084\
                ---------------------------------------------------------------------------
                 \2084\ See Bureau of Labor Statistics, historical data series
                LNS14000000, available at https://data.bls.gov/cgi-bin/surveymost?ln.
                ---------------------------------------------------------------------------
                 The agencies were unable to obtain useful measures of roadway
                design parameters or road conditions that would be expected to affect
                safety. Although such measures exist, they tend to be reported for
                individual road and highway segments or routes, and it is difficult to
                combine these data into meaningful, aggregate measures that describe
                overall driving conditions that are likely to vary by calendar year.
                Nor could they identify satisfactory measures of incident response time
                or the effectiveness of emergency medical treatment in reducing the
                consequences of injuries occurring in motor vehicle crashes.
                 An important challenge to incorporating these time-period effects
                into the fatality trend model arose from the fact that their patterns
                of variation over the historical period the agencies analyzed (which
                extended from calendar year 1995 to 2017) were extremely closely
                correlated, making it virtually impossible to distinguish their
                independent contributions to improvements in fleet-wide safety over
                time. Table VI-254 below reports the pairwise correlation coefficients
                among the potential measures of period effects listed above. As it
                suggests, patterns of variation about their respective mean values over
                the period analyzed were very similar (with the exception of the
                unemployment rate), and the resulting high statistical correlations (or
                ``collinearity'') among them made it nearly impossible to identify
                their independent effects on variation in safety over time, even when
                controlling for the effects of model year and vehicle age.
                [[Page 24814]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.468
                 To address this difficulty, the agencies substituted a time trend--
                that is, a variable that takes the value of one in the first calendar
                year and increases by one in each successive calendar year--in an
                effort to capture the joint movements in the variables that were
                intended to measure time-period effects on safety. The agencies
                experimented with both linear and more complex time trends to capture
                the apparently declining rate of improvement in fleet-wide safety over
                time, but found that the linear trend captured the combined effects
                most reliably. Because the model's dependent variable is the natural
                logarithm of model year and age-specific fatality rates, using a linear
                time trend corresponds to assuming a constant percentage decline in
                fatality rates each year (rather than a constant absolute decline each
                year), and this pattern appeared to provide the best fit to the
                observed historical pattern of safety improvements. Finally, after
                noting that the linear time trend did not fully capture the effects on
                fleet-wide safety associated with the economic recessions in 2001 and
                2007-11, the agencies supplemented the time trend with indicator (or
                ``dummy'') variables for these years, finding that only those for 2008,
                2009, and 2010 improved its explanatory power significantly.
                 Another significant improvement to the NPRM analysis was to group
                model years into ``safety cohorts'' on the basis of similarity in their
                fatality rates when new (that is, during their first year in service),
                rather than treating each model year as a separate cohort. Groupings
                were created through a combination of identifying years when new safety
                regulations initially took effect or were phased in, examining of
                first-year fatality rates, and limited statistical experimentation.
                Grouping successive model years reduces the number of cohorts
                significantly, since similar fatality rates were typically observed for
                at least five, and sometimes as many as ten, consecutive model years
                over the historical period the agencies examined. Grouping model years
                into a smaller number of cohorts rather than treating each model year
                as a separate cohort offers the advantage of introducing some variation
                in the ages of vehicles making up the same cohort during a calendar
                year, which improves the statistical reliability with which the
                independent effect of age itself can be estimated.
                 Figure VI-84 below shows historical variation in the fatality rates
                of past model years when each one was newly-introduced (i.e., during
                its first year in use).\2085\ It clearly displays the significant
                improvement in the safety of new vehicles over time in response to
                improvements in safety features, including those required by NHTSA's
                safety regulations. The figure also clearly documents the natural
                clustering of fatality rates for successive model years that was used
                to identify and define the safety cohorts used in the revised model. In
                the panel structure of the model, which combines time-series and cross-
                section variation in fatality rates for individual model years as their
                ages vary across calendar years, the clustering of first-year fatality
                rates for successive model years is captured by using separate ``fixed
                effects'' for each safety cohorts illustrated in the figure. Some
                judgment is inevitably required to distinguish between successive
                cohorts and identify when the fatality rate for new model years has
                changed significantly; the agencies experimented with using from five
                to eight cohorts, ultimately finding that the agencies could
                distinguish most reliably among the fatality rates for five cohorts.
                ---------------------------------------------------------------------------
                 \2085\ For simplicity, the figure assumes that each model year's
                first year of use was the calendar year identical to its designated
                model year; for example, the first full year of use for model year
                2000 was assumed to be calendar year 2000. In fact, new vehicles
                frequently become available for purchase during the calendar year
                preceding their designated model year and continue to be sold
                through the calendar year following it, although most sales occur
                during the calendar year matching their designated model year.
                ---------------------------------------------------------------------------
                [[Page 24815]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.469
                 A final revision to the NPRM model was to employ a slightly less
                complex mathematical relationship between a model year's age and its
                fatality rate than had been used in the NPRM version. Specifically, the
                revised model relates fatality rates to age itself as well as the
                second and third powers of age (that is, age squared and age cubed),
                but omits the fourth power of age, which was included in the model
                developed for the NPRM. This slightly simpler relationship proved
                adequate to capture fully the complex--but strongly recurring--pattern
                of fatality rates for past model years as they aged. Specifically, as
                Figure VI-85 below shows, fatality rates have tended to remain
                approximately constant for the first few years of most recent model
                years' lifetimes, before increasing steadily through age 15-20 and then
                declining gradually over the remainder of their lifetimes.
                 As discussed previously, the increase in fatality rates through
                approximately age 20 is generally thought to result primarily from the
                fact that used vehicles are commonly purchased and driven by members of
                households whose demographic characteristics, driving behavior, and
                geographic locations are associated with more risky driving behavior
                and thus more frequent or severe crashes. Of course, increased
                frequency of mechanical failures as vehicles age and accumulate mileage
                also seems likely to contribute to this pattern. In contrast, the
                consistent tendency for fatality rates to decline after about age 20 is
                less well understood, but may owe partly to the demographic
                characteristics and driving behavior of owners of very old vehicles.
                Whatever its source, the number of vehicles remaining in service past
                age 20 is so small and their use typically so limited that their
                contribution to the fleet-wide fatality rate is minimal.
                [[Page 24816]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.470
                 Figure VI-85 documents the relationship between age and fatality
                rate for selected past model years.\2086\ As it shows, fatality rates
                for recent model years follow a complex but strikingly similar pattern
                of increase and subsequent decline with increasing age, although the
                figure also shows that the earliest model years included in the sample
                (1975-1980) tended not to display increasing fatality rates in the
                first half of their lifetimes. At the same time, the figure illustrates
                the gradual downward shift in fatality rates at all ages for successive
                past model years, although there is considerable variation in the
                extent of this shift for individual model years, particularly when they
                are examined at specific ages. That is, the downward shift in fatality
                rates for successive model years is not necessarily ``monotonic,''
                particularly when it is examined at specific individual ages.
                ---------------------------------------------------------------------------
                 \2086\ For a color version, see the corresponding safety
                discussion in the accompanying FRIA.
                ---------------------------------------------------------------------------
                 The agencies believe that the increase in fatality rates for cars
                and light trucks produced during recent model years through
                approximately age 20 reflects the fact that as aging vehicles change
                ownership via the used car market, they are often purchased and driven
                by households whose demographic characteristics and locations are
                associated with riskier driving behavior and conditions. The decline in
                vehicles' fatality rates after this age is not well understood, but
                seems likely to reflect the fact that the relatively small fraction of
                those originally produced in a model year that survive beyond age 20-25
                are owned and driven by households that maintain them carefully, are
                likely to reside in areas where driving conditions are safest, and
                whose members engage in less risky driving behavior.
                 After examining the information summarized in Figure VI-85, the
                agencies conclude that the effect of increasing age on vehicle safety
                appears to be largely independent of the improvement in new cars'
                fatality rates over successive model years, and appears to operate
                similarly for all except the earliest model years in our historical
                sample (which includes model years 1975-2017).\2087\ As a formal
                statistical test, the agencies experimented with allowing the aging
                effect to change across model years when the agencies estimated the
                revised model, anticipating that newer safety technologies and vehicle
                designs might ``flatten'' the relationship between fatality rates and
                age--that is, reduce the degree to which fatality rates increased over
                the 5-20 year range of vehicle ages--for newer model years. However,
                the agencies found no evidence that the effect of age on safety changed
                significantly for more recent model years compared to older ones, so
                the agencies retained the assumption of identical aging effects for all
                model
                [[Page 24817]]
                years in the revised model.\2088\ Thus the revised model shows
                progressively lower fatality rates for more recent model years when
                they are new, but fatality rates for all model years increase with age
                and subsequently decline according to the same non-linear pattern
                displayed in Figure VI-85. On a related question, the agencies also
                found that including the squared and cubed values of age in addition to
                age itself as explanatory variables in the model, while excluding the
                fourth power of age, which had been included in the NPRM model, proved
                adequate to capture the pattern of variation in fatality rates with
                increasing age that most past model years have exhibited. Table VI-255
                below reports the estimated parameter values for alternative
                specifications of the model, together with various goodness-of-fit and
                other diagnostic measures. The analysis described in the following
                section uses the estimated time trend from Model 2 in the table, which
                implies annual reduction in fatality rates for all model years of 2.14
                percent.
                ---------------------------------------------------------------------------
                 \2087\ Of course, the agencies cannot observe the safety
                performance of all model years included in the agencies' data sample
                over their entire lifetimes, because the data the agencies use to
                estimate the model start in calendar year 1990, by which time all
                model years before 1990 were no longer new--for example, MY1975 cars
                are already 15 years old by then--while the newest model years in
                the agencies' sample are still very ``young'' when the agencies'
                data ends in calendar year 2017. Thus, the agencies have only
                incomplete information about the relationship of fatality rates to
                age over the entire lifetimes of these model years, so it is
                possible that this relationship differs at particularly early or
                advanced ages for the oldest and newest model years in the agencies'
                sample.
                 \2088\ Specifically, the agencies tested for interactions
                between the age and model year variables, which would reveal changes
                in the relationship between fatality rates and age for more recent
                model years, but found that such interaction effects were generally
                not statistically significant. Allowing for interactions between age
                and the indicator variables for safety cohorts (recall that these
                represent groupings of successive model years) produced this same
                result--few of the interaction effects were statistically
                significant.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24818]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.471
                BILLING CODE 4910-59-C
                BILLING CODE 4910-59-P
                [[Page 24819]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.472
                [[Page 24820]]
                BILLING CODE 4910-59-C
                Using the Model and Technology Analysis to Forecast Fatality Rates
                 The newest safety cohort includes model years from 2009 to 2017, so
                in effect the agencies estimate that all those model years have
                essentially the same fatality rate in their first year of use. The
                agencies apply the estimated effectiveness of crash avoidance
                technologies in reducing fatal crashes to the observed fatality rate
                for model years 2009 to 2017 vehicles during their first year in use to
                estimate fatality rates for future model years during the first year
                each one is introduced. Figure VI-86 below shows the result of this
                process; as it indicates, fatality rates for new model years decline
                gradually through 2035 and then stabilize, reflecting the fact that the
                agencies are only able to project the effectiveness of emerging crash
                avoidance technologies on the safety of new vehicles through that year.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.473
                 The next step in constructing the forecast of fleet-wide fatality
                rates is to apply the age-related increases in the fatality rate for
                each model year making up the previous calendar year's fleet. For
                example, the agencies assume that the fatality rates for all model
                years comprising the light-duty vehicle fleet in 2017 increase with age
                according to the relationship captured by the estimated coefficients on
                the age variables in the preferred model specification shown in Table
                VI-255. The same assumption is applied to all new model years
                introduced in subsequent years. Finally, the agencies also assume that
                the historical decline in fatality rates observed over past calendar
                years (the ``period effect'' captured by the time trend variable) will
                continue into the future. This implies that fatality rates for all
                model years and ages will decline by an additional 2.41 percent in each
                successive future calendar year from the rates that would have resulted
                from the combined effects of continuing improvements in the safety of
                newly-introduced model years and the effect of increasing age.\2089\
                ---------------------------------------------------------------------------
                 \2089\ The agencies do not apply this trend reduction to the
                fatality rates for the newest model year in each calendar year's
                fleet, because it is assumed to be independent of both the decline
                in new-car fatality rates and the aging effect.
                ---------------------------------------------------------------------------
                 This process produces an estimate of the fatality rate for each
                model year making up the fleet during each future calendar year. That
                estimate reflects the combination of (1) reductions in fatality rates
                for new cars, reflecting the continued improvements in their safety due
                to crash avoidance technologies (through MY2035); (2) increases in the
                fatality rates for each model year in the fleet from the previous
                calendar year, which represent the effect of age estimated by the
                historical model; and (3) the continuing downward trend in fatality
                rates for all vehicles except the newest model year in each calendar
                year's fleet, which is derived from the historical model.
                 The agencies then weight the fatality rate for each model year
                making up a future year's fleet by the fraction of total fleet-wide VMT
                it accounts for, and sum
                [[Page 24821]]
                the results to produce an estimate of the fleet-wide fatality rate. The
                CAFE model does not actually use this fleet-wide fatality rate, because
                all of the fatality calculations are performed separately for each
                individual model year making up the fleet, which are then aggregated;
                nevertheless, the agencies provide the fleet-wide rate as a useful
                check on the reasonableness of our fatality rate forecasts for
                individual model years as they enter the fleet and age over their
                respective lifetimes. Figure VI-87 displays the projected fleet-wide
                fatality rates for future calendar years, as well as the trend in their
                recent historical values.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.474
                (d) Impact of Advanced Technologies on Older Vehicle Fatality Rates
                 In the NPRM, the agencies calculated the potential safety impacts
                of delayed purchases of vehicles with new safety technology that might
                result from higher vehicles prices associated with more stringent CAFE
                standards. A number of commenters noted that since these improvements
                will be driven by crash avoidance technologies, they will also benefit
                older vehicles and reduce their fatality rates as well. For example,
                CARB noted that ``safety improvements generally provide systematic
                safety benefits to all vehicles in the on-road fleet, not only to new
                vehicles. However, NHTSA's safety model assigns safety coefficients to
                vehicles solely based on their model year and it fails to incorporate
                the effect that new safety designs and technologies will have on
                systematically improving fleet-wide on-road safety.'' IPI similarly
                noted that should ``new safety technologies be adopted, the predicted
                fatalities for all the older vehicle vintages will have to be lowered
                as well because effective crash avoidance technologies will lower all
                vehicles' fatality costs.''
                 The agencies agree that the users of older vehicles will also
                benefit from crash avoidance technologies on newer vehicles. In
                response, the agencies have modified our methodology to reflect lower
                fatality rates on older vehicles resulting from the new crash avoidance
                technologies. Crash avoidance technologies prevent crashes from
                happening and thus benefit both the vehicle with the technology and any
                other vehicles that it might have collided with. However, the scope of
                these impacts on older vehicle's fatality rates are somewhat limited
                due to several factors:
                 Single vehicle crashes, which make up about half of all fatal
                crashes, will not be affected. Only multi-vehicle crashes involving a
                newer vehicle with the advanced technology and an older vehicle will be
                affected. Multi-vehicle crashes account for roughly half of all light
                vehicle occupant fatalities.
                 For a new safety technology to benefit an older vehicle in
                a multi-vehicle crash, the vehicle with the technology must have been
                in a position to control, or prevent the crash. For example, in front-
                to-rear crashes which can be addressed by FCW and AEB, the older
                vehicle would only benefit if it was the vehicle struck from behind. If
                the struck vehicle were the newer vehicle, its AEB technology would not
                prevent the crash. Logically this would occur in roughly half of two-
                vehicle crashes and a third of all three-vehicle crashes. Since most
                multi-vehicle crashes involve only two vehicles, roughly half of all
                multi-vehicle crashes might qualify.
                 The benefits experienced by older vehicles are
                proportional to the probability that the vehicles they collide with are
                newer vehicles with advanced crash avoidance technology. The
                [[Page 24822]]
                agencies estimate that the probability that this would occur is a
                function of the relative exposure of vehicles by age, measured by the
                portion of total VMT driven by vehicles of that age. Based on VMT
                schedules (see CY 2016 example in Table VI-256), new (current MY)
                vehicles account for about 9.6 percent of annual fleet VMT. The
                relevant portion would increase over time as additional MY vehicles are
                produced with advanced technologies. However, the portion of older
                vehicle crashes that might be affected by newer technologies is
                initially very small--only about 2 percent (.5*.5*.096) of older
                vehicles involved in crashes might benefit from advanced crash
                avoidance technologies in other vehicles in the first year.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.475
                [[Page 24823]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.476
                 To reflect this safety benefit for older vehicles, the agencies
                calculated a revised fatality rate for each older MY vehicle on the
                road based on its interaction with each new MY starting with MY 2021
                vehicles based on the following relationship:
                Revised fatality rate = Fm-((x-y)mnp) + F(1-m)
                Where: F = initial fatality rate for each MY
                x = baseline MY fatality rate
                y = current MY fatality rate
                m = proportion of occupant fatalities that occur in multi-vehicle
                crashes (52 percent)
                n = probability that crash is with a new MY vehicle containing
                advanced technologies
                p = probability that new vehicle is ``striking'' vehicle
                 The initial fatality rate for each vehicle MY (F) was derived by
                combining fatality counts from NHTSA's Fatality Analysis Reporting
                System (FARS) with VMT data from IHS/Polk.
                 The baseline MY fatality rate (x) represents the baseline rate over
                which the impact of new crash avoidance technologies should be measured
                It establishes the baseline rate for each MY that will be compared to
                the most current MY rate to determine the change in fatality rate (FR)
                for each MY. The relative effectiveness of new crash-avoidance
                technologies in modifying the fatality rate of older model vehicles is
                measured differently depending on the age of the older vehicle. The
                fatality rate is a historical measure that reflects safety differences
                due to both crashworthiness technologies such as air bags and crash
                avoidance technologies such as electronic stability control, but up
                through MY 2017, crashworthiness standards are the predominant cause of
                these differences.
                 The most recent significant crashworthiness safety standard, which
                upgraded roof strength standards which was effective in all new
                passenger vehicles in MY 2017. Crashworthiness standards would not have
                secondary benefits for older MY vehicles. Post MY 2017, the agencies
                believe crash avoidance technologies will drive safety improvements. To
                isolate the added crash avoidance safety expected in newer vehicles,
                the marginal impact of the difference between the MY 2017 fatality rate
                and the most current MY fatality rate represents the added marginal
                effectiveness of new crash-avoidance technologies of each subsequent MY
                for MYs 2017 and earlier. Beginning with MY 2018, the difference
                between the older MY fatality rate and most current MY rate determines
                the potential safety benefit for the older vehicles.
                [[Page 24824]]
                 The current MY fatality rate (y), represents the projected fatality
                rate of future MY vehicles after adjustment for the impacts of the
                advanced crash avoidance technologies and projected improvements in
                non-technology factors examined in this analysis. This process was
                discussed in detail in the previous section.
                 The proportion of passenger vehicle occupant fatalities that occur
                in multi-vehicle crashes (m), was derived from an analysis of occupants
                of fatal passenger vehicle crashes from 2002-2017 FARS. The analysis
                indicated that 47.8 percent of fatal crash occupants were in single
                vehicle crashes, 40.2 percent were in two vehicle crashes, and 12
                percent were in crashes involving 3 or more vehicles. Overall, 52.2
                percent were in multi-vehicle crashes.
                 The portion of older vehicle crashes involving newer vehicles
                containing advanced crash avoidance technologies (n), is assumed to be
                equal to the cumulative risk exposure of vehicles that have these
                technologies. This exposure is measured by the product of annual VMT by
                vehicle age and registrations of vehicles of that age. The CAFE model
                calculates this dynamically, but as an example, based on 2016
                registration data (see Table VI-256 above), the most current MY would
                represent 9.6 percent of all VMT in a calendar year, implying a 9.6
                percent probability that the vehicle encountered would be from the most
                current MY. This percentage would increase for each CY as more MY
                vehicles adopt advanced crashworthiness technologies. The agencies note
                that other factors such as uneven concentrations of newer vs. older
                vehicles or improved crash avoidance in the younger vehicles already on
                the road that are the basis for the agencies' VMT proportion table
                might disrupt this assumption, but it is likely that this would only
                serve to slow the probability of these encounters, making this a
                conservative assumption in that it maximizes the probability that older
                vehicles might benefit from newer technologies.
                 The probability that the vehicle with advanced crash avoidance
                technology is the controlling or striking vehicle (p), was calculated
                using the relative frequency of fatal crash occupants in multi-vehicle
                crashes. As noted previously, 40.2 percent were in two vehicle crashes,
                and 12 percent were in crashes involving 3 or more vehicles. The
                agencies assume a probability of 50 percent for two vehicle crashes and
                33 percent for crashes with 3 or more vehicles. Weighted together the
                agencies estimate a 46.1 percent probability that, given a multi-
                vehicle crash involving a vehicle with advanced technologies and an
                older vehicle without them, the newer vehicle will be the striking
                vehicle or in a position where its crash avoidance technologies might
                influence the outcome of the crash with the older vehicle.
                 This process is illustrated in Table VI-257 below for adjustments
                due to improvements in MY 2021 vehicles back through MY 1995. In Table
                VI-257, the actual model year fatality rate is shown in the second
                column. As noted above, the base fatality rate, shown in column 3, is
                the MY 2017 rate for all MYs prior to 2018, after which it becomes the
                actual MY rate. Column 4 shows the difference between the fatality rate
                for MY 2021 and the base rate for each MY. Column 5 shows the resulting
                revised fatality rate that would be used for each older MY, and column
                6 and 7 list the change in that rate. The various factors noted in the
                above formula are applied in column 5. The results indicate a 0.006
                decrease in pre-2018 MY vehicles fatality rates, with declining impacts
                going forward to MY 2021. In subsequent years, this impact would grow
                to reflect the both the increased probability that an older vehicle
                would crash with vehicles containing advanced technology, as well as
                the increased technology levels in progressively newer vehicles. This
                table was created using NPRM inputs and is provided for explanatory
                purposes only. The actual impacts are dynamically calculated within the
                Volpe model and reflect revised fatality rate trends going forward and
                cover even older model years.
                [[Page 24825]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.477
                (e) Dynamic Fleet Composition
                 As described in the sales discussion in Section Dynamic Fleet Share
                (DFS), the standards may impact the distribution of cars and trucks
                purchased. As light trucks, SUVs and passenger cars respond differently
                to technology applied to meet the standards--namely mass reduction--
                fleets with different compositions of body styles will have varying
                amounts of fatalities. Since mass-safety fatalities are calculated by
                multiplying mass point-estimates by VMT, which implicitly captures the
                impact of the dynamic fleet share model, the estimates of mass-safety
                fatalities in the previous section include the impact of vehicle prices
                on fleet composition.
                (c) Impact of Rebound Effect on Fatalities
                 The ``rebound effect'' is a measure of the additional driving that
                occurs when the cost of driving declines. More stringent standards
                reduce vehicle operating costs, and in response, some consumers may
                choose to drive more. Driving more increases exposure to risks
                associated with on-road transportation, and this added exposure
                translates into higher fatalities. The agencies have calculated this
                impact by estimating the change in VMT that results from alternative
                standards.
                 As noted previously, rebound miles are not imposed on consumers by
                regulation. They are a freely chosen activity resulting from reduced
                vehicle operational costs. As such, the agencies believe a large
                portion of the safety risks associated with additional driving are
                [[Page 24826]]
                offset by the benefits drivers gain from added driving. For the
                proposal, the agencies assumed that, in deciding to drive more, drivers
                internalize the full cost to themselves and others, including the cost
                of accidents, associated with their additional driving.
                 In response to the NPRM, EDF noted that consumers may not fully
                value the added safety risk, such as risk to other drivers.\2090\ In
                making this point, EDF suggested a value of 50 percent would be
                conservative, but did not provide supporting evidence for that value.
                The agencies agree that the level of risk internalized by drivers is
                uncertain, and for the final rule have revised the portion of the added
                monetized safety risk that consumers internalize to 90 percent, which
                mostly offsets the societal impact of any added fatalities from this
                voluntary consumer choice.
                ---------------------------------------------------------------------------
                 \2090\ EDF, Appendix B, NHTSA-2018-0067-12108, at 101.
                ---------------------------------------------------------------------------
                 The actual portion of risk from crashes that drivers internalize is
                unknown. The agencies suspect that drivers are more likely to
                internalize serious crash consequences than minor ones, and some
                drivers may not perfectly internalize injury consequences to other
                individuals, especially occupants of other vehicles and pedestrians.
                However, legal consequences from crash liability, both criminal and
                civil, should also act as a caution for drivers considering added crash
                risk exposure. The agencies considered several approaches to estimating
                internalized crash risk. The first assumes that drivers value harm to
                themselves as well as legal liability for causing harm to others. It
                considers that all fatalities in single vehicle crashes are fully
                valued, that there is roughly a 50 percent chance that each driver
                would be the one killed in multi-vehicle crashes, and that there is
                roughly a 50 percent chance that each driver would be at-fault in a
                multi-vehicle crash that they survived. This produces an estimate of
                roughly 87 percent. Another approach assumes that drivers fully value
                all damage in single vehicle crashes, and only discount property damage
                incidents in multi-vehicle crashes. Based on data in Blincoe, et al.
                (2015),\2091\ multi-vehicle property-damage-only crashes account for
                about 7 percent of all societal crash costs, leaving 93 percent
                recognized under this approach. Yet another approach would assume
                drivers value injury crashes, but discount non-injury related costs
                such as property damage and traffic congestion. This approach results
                in roughly an 88 percent estimate of costs internalized. Overall, while
                the agencies recognize this proportion is uncertain, the agencies
                believe it is reasonable to assume that drivers internalize 90 percent
                of the crash risk that results from added driving.
                ---------------------------------------------------------------------------
                 \2091\ Blincoe, L., Miller, T.R., Zaloshnja, E., Lawrence, B.A.,
                (May 2015, Revised) The Economic and Societal Impact of Motor
                Vehicle Crashes, 2010, (DOT HS 812 012), National Highway Traffic
                Safety Administration, Washington, DC.
                ---------------------------------------------------------------------------
                 IPI commented that additional mileage attributable to the scrappage
                and dynamic fleet model is ``inexplicably and unjustifiably not offset
                by countervailing mobility benefits in the benefit cost analysis--and
                the agencies inappropriately claim that these traffic fatalities--which
                comprise the other half of the 12,700 projection--also justify the roll
                back.'' \2092\ In this comment, IPI has erroneously conflated the
                rebound effect and the scrappage effect. The agencies have
                appropriately accounted for the additional value consumers get out of
                increases in fuel efficiency, which manifest in two ways: Reductions in
                fuel costs, and the additional driving resulting from the reductions in
                per-mile fuel costs. The agency cannot appropriately consider one
                without the other, as the two effects trade off, one against the other,
                according to consumer preferences between the two.
                ---------------------------------------------------------------------------
                 \2092\ IPI, Appendix, NHTSA-2018-0067-12213, at 12 (internal
                citation omitted).
                ---------------------------------------------------------------------------
                 The scrappage effect represents the behavior of consumers when
                their choices are restricted by more stringent fuel economy standards.
                For instance, the consumer loses lower-price and less fuel-efficient
                bundles of vehicle attributes that would be available in the absence of
                more stringent alternatives. If anything, these consumers experience an
                un-estimated cost regarding the lost utility from being priced out of
                the new car market and being forced to drive an older, less safe--and
                likely less fuel efficient--vehicle. That the agencies have assessed
                the benefits of the rebound effect by assuming they are at least as
                great as 90 percent of the additional safety costs of rebound driving,
                does not mean that other channels of safety effects must be offset.
                However, the agencies did evaluate whether the sales, scrappage, and
                dynamic fleet share model could lead to changes in fuel economy in the
                legacy fleet that may result in significant changes in VMT and/or fuel
                economy. Upon further review, the agencies determined that such an
                effect--if it were to exist--would be very small and would not impact
                the analysis meaningfully, so the agencies declined to include this
                effect in the final rule's analysis.
                d) Fatalities by Source
                 For the NPRM, the agencies calculated rebound fatalities by running
                the model with a 20 percent rebound assumption and again with a 0
                percent rebound assumption. The following difference was assumed to
                assign the change in fatalities of the rule due to rebound:
                Rebound Fatalities = (FatalitiesAlt,20 -
                FatalitiesAlt,0) -
                (FatalitiesAug,20 -
                FatalitiesAug,0)
                 Similarly, the agencies calculated mass reduction fatalities by
                running the model using the central assumptions about coefficients on
                delta curb weight and again setting these coefficients to 0, so that a
                change in mass reduction would not affect the fatality rate of a
                vehicle. The following difference assigned the change in fatalities of
                the rule due to changes in mass reduction levels:
                [Delta]CW Fatalities = (FatalitiesAlt,MR - FatalitiesAlt,NoMR) -
                (FatalitiesAug,MR) - (FatalitiesAug,NoMR)
                Where ``Alt'' represents the alternative being estimated, ``Aug'' is
                the augural or baseline, ``MR'' stands for mass reduction, and
                ``NOMR'' means no mass reduction or mass reduction equaling zero.
                 The NPRM modeling then assumed that the remaining incremental
                fatalities were due to changes in sales, scrappage, and the dynamic
                fleet share. This can be represented by the following:
                Sales/Scrap Fatalities = (FatalitiesAlt - FatalitiesAug) - Rebound
                Fatalities - [Delta]CW Fatalities
                 The changes to the VMT model (mainly the constraint that fixes
                total non-rebound VMT to be constant across alternatives) necessitated
                revising how fatalities are partitioned by source. The number of
                vehicles of each regulatory class and age changes in each regulatory
                alternative. Because of this, taking the increment of the rebound
                fatalities solved in each scenario as described above would capture
                changes both to the usage per vehicle from rebound, but also
                differences in the number of vehicles. This would wrongly attribute
                some of the sales and scrappage fatalities to rebound. Similarly,
                taking the increment of the mass reduction fatalities solved in each
                scenario as described above would capture the changes both to the
                fatality rate for vehicles (from mass reduction) and the difference in
                the number of vehicles across alternatives. This would likewise have
                the potential of wrongly attributing the source of sales and scrappage
                fatalities to mass reduction.
                [[Page 24827]]
                 Instead of computing the fatalities due to rebound in each scenario
                and then taking the incremental values across alternatives, the
                agencies compute rebound fatalities by taking the difference in per
                vehicle rebound miles in the regulatory alternative and the augural
                case multiplied by the augural fatality rate per mile and augural
                vehicle count. Holding the number of vehicles constant addresses the
                concern about the NPRM fatality allocation method wrongly attributing
                rebound fatalities to the sales and scrappage models. Fatalities due to
                rebound are computed as follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.478
                Where ``RVMT'' is VMT including rebound miles, ``NRVMT'' is VMT
                excluding rebound miles, ``Veh'' is the quantity of vehicles, and
                ``Alt'' and ``Aug'' have the same meaning described above. The
                rebound fatalities will show as zero for the augural scenario, and
                all alternatives will show fatalities due to rebound miles using the
                augural vehicle counts.
                 The fatalities due to mass reduction will use the augural vehicle
                counts, augural per vehicle VMT including rebound--this simplifies to
                total VMT including rebound, as shown below. Using a constant vehicle
                count addresses the concern of the NPRM method wrongly assigning some
                mass reduction fatalities to the sales and scrappage models. As with
                the fatalities attributable to rebound, the fatalities attributable to
                changes in mass reduction are calculated inherently as incremental
                values, relative to the augural standards (the values will appear as
                zero for augural standards in the outputs). The equation used to
                calculate the fatalities due to curb weight changes is as follows:
                [Delta]CW FatalitiesAlt = (Fatality RateAlt - Fatality RateAug) * R
                VMTAug
                 The agencies then computed the sales/scrappage fatalities as the
                remainder, as was done in the NPRM.
                Sales/Scrap Fatalities = (FatalitiesAlt-FatalitiesAug)-Rebound
                Fatalities-[Delta]CW Fatalities
                (e) Adjustment for Non-Fatal Crashes
                 Fatalities are valued as a societal cost within the CAFE models'
                cost and benefit accounting. Their value is based on the comprehensive
                value of a fatality, which includes lost quality of life and is
                quantified in the value of a statistical life (VSL) as well as economic
                consequences such as medical and emergency care, insurance
                administrative costs, legal costs, and other economic impacts not
                captured in the VSL alone. These values were derived from data in
                Blincoe et al. (2015), adjusted to 2018 economics, and updated to
                reflect the official DOT guidance on the value of a statistical life.
                This gives a societal value of $10.4 million for each fatality, which
                is an update to the value used in the NPRM.\2093\ The CAFE safety model
                estimates traffic fatalities but does not directly estimate the
                corresponding non-fatal injuries and property damage that would result
                from the same factors that influence fatalities. To address this, the
                agencies developed an adjustment factor applied to fatality costs that
                accounts for these crashes and related costs. The agencies' approach to
                estimating non-fatal costs remains relatively unchanged from the
                proposal, however the agencies have made one minor adjustment to
                account for advance crash technologies as advocated by commenters.
                ---------------------------------------------------------------------------
                 \2093\ The NPRM used a societal value of $9,900,000 in 2016
                dollars.
                ---------------------------------------------------------------------------
                 In the proposal, development of this factor was premised on the
                assumption that non-fatal crashes would be affected by the standards in
                proportion to their current nationwide rate of incidence and severity.
                The agencies assumed the injury profile--the relative number of crashes
                of each injury severity level that occur nationwide--would increase or
                decrease congruent with changes in fatalities, meaning that the ratio
                between fatal and non-fatal costs remained constant across
                alternatives. The agencies recognized that this may not be the case,
                but did not have data to support individual injury estimates across
                injury severities. The agencies provided several explanations as to why
                a proportionality assumption may be an oversimplification.\2094\ For
                example, the agencies reviewed NHTSA's separate analysis of traffic
                crash data showing that older model year vehicles are generally less
                safe than newer vehicles, meaning fatalities would comprise a larger
                portion of the total injury picture for older vehicles. This would
                imply lower ratios across the non-fatal injury and property damage only
                (PDO) crash profiles and would imply the adjustment overstates total
                societal impacts.
                ---------------------------------------------------------------------------
                 \2094\ See 83 FR 43146 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 As noted previously, in response to requests by commenters, the
                agencies have added the estimated impact of six advanced crash
                avoidance technologies that are currently being deployed commercially
                to their analysis of future fatality rates. The same data and methods
                described previously in this section to compute the impact of advanced
                crash avoidance technologies on fatalities can also be used to examine
                the effectiveness of these technologies against non-fatal and PDO
                crashes. The inputs and results are summarized for nonfatal injuries in
                Table VI-258 through Table VI-260, and for PDOs in Table VI-261 through
                Table VI-263.\2095\
                ---------------------------------------------------------------------------
                 \2095\ See previous discussion in this section for the studies
                and methodology used to create these estimates.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24828]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.479
                [GRAPHIC] [TIFF OMITTED] TR30AP20.480
                [[Page 24829]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.481
                [GRAPHIC] [TIFF OMITTED] TR30AP20.482
                [[Page 24830]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.483
                [GRAPHIC] [TIFF OMITTED] TR30AP20.484
                [[Page 24831]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.485
                [GRAPHIC] [TIFF OMITTED] TR30AP20.486
                [[Page 24832]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.487
                BILLING CODE 4910-59-C
                 Based on a comparison of the combined average effectiveness impacts
                for the three crash severity groups (fatalities, non-fatal injuries,
                and property damage), it is apparent that these advanced crash
                avoidance technologies would reduce non-fatal injuries and property
                damage crashes by even more than they would fatalities.\2096\ To
                explore the scope of this impact, the agencies developed an adjustment
                factor that reflects the ratio of the decline in the rate of non-fatal
                crashes to that of fatal crashes. This factor would hypothetically
                affect the portion of safety improvement that is attributable to safety
                technologies. The adjustments were based on the cumulative fatality
                rates (for all three technology groups) by model year, noted in Table
                VI-251 (Phased Impact of Crashworthiness Technologies on Fatality
                Rates, Forward Collision Crashes) for fatalities, Table VI-260 for non-
                fatal injuries, and Table VI-263 for PDOs, which are listed by MY in
                the last column of Table VI-260 and Table VI-263. These factors would
                modify the original non-fatal impacts--which were derived using an
                assumption that they were proportional to fatal impacts--to reflect the
                higher effectiveness of these technologies against non-fatal crashes.
                ---------------------------------------------------------------------------
                 \2096\ For example, for MY 2035, the combined effectiveness for
                PDO crashes is .224784, as shown in the second to last column of
                Table VI-6, which is 2.613 times the .0860 combined effectiveness
                for fatalities, as seen in the final table from the Crash Avoidance
                discussion above, which shows the disproportional impact of crash
                avoidance technologies on non-fatal accidents.
                ---------------------------------------------------------------------------
                 The agencies considered including this additional adjustment factor
                to account for the additional cost savings attributable to advance
                crash avoidance technologies. The impact of such a factor would
                decrease the incidence and severity, and thus the costs of nonfatal
                crashes in regulatory alternatives where new vehicle sales increase,
                including the preferred alternative. The agencies ultimately erred on
                the side of caution for this rulemaking and have excluded this factor.
                Therefore, today's analysis assumes that advance crash avoidance
                technologies impact non-fatal and PDO crashes to the same extent as
                fatal crashes. The agencies will consider including an adjustment for
                non-fatal and PDO crashes in future rulemakings.
                [[Page 24833]]
                 The original proportionality-based adjustment factor, which is
                described in detail in the following paragraphs, was derived from
                Tables 1-8 and I-3 in Blincoe et al. (2015). Incidence in Table I-3 in
                Blincoe et al. reflects the Abbreviated Injury Scale (AIS), which ranks
                nonfatal injury severity based on an ascending 5 level scale with the
                most severe injuries ranked as level 5.\2097\
                ---------------------------------------------------------------------------
                 \2097\ More information on the basis for these classifications
                is available from the Association for the Advancement of Automotive
                Medicine at https://www.aaam.org/abbreviated-injury-scale-ais/.
                ---------------------------------------------------------------------------
                 Table 1-3 in Blincoe et al. lists injured persons with their
                highest (maximum) injury determining the AIS level. This scale is
                represented in terms of maximum abbreviated injury scale (MAIS) level.
                MAIS0 refers to uninjured occupants in injury vehicles, MAIS1 injuries
                are generally considered minor (e.g., a superficial laceration) with no
                probability of death, MAIS2 injuries are generally considered moderate
                (e.g., a fractured sternum) with a 1-2 percent probability of death,
                MAIS3 injuries are serious (e.g., open fracture of the humerus) with an
                8-10 percent probability of death, MAIS4 injuries are severe (e.g.,
                perforated trachea) with a 5-50 percent probability of death, and MAIS5
                injuries are critical (e.g., rupture liver with tissue loss) with a 5-
                50 percent probability of death. Counts for PDO's refer to vehicles in
                which no one was injured. From Table VI-264, ratios of injury
                incidence/fatality are derived for each injury severity level as
                follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.488
                [GRAPHIC] [TIFF OMITTED] TR30AP20.489
                 For each fatality that occurs nationwide in traffic crashes, there
                are 561 vehicles involved in PDOs, 139 uninjured occupants in crashes
                which resulted in at least one injury,\2098\ 105 minor injuries, 10
                moderate injuries, 3 serious injuries, and fractional numbers of the
                most serious categories which include severe and critical nonfatal
                injuries. For each fatality ascribed to the standards, it is assumed
                there will be non-fatal crashes in these same ratios.
                ---------------------------------------------------------------------------
                 \2098\ Uninjured passengers incur a cost despite being
                uninjured. For example, they are often transported to emergency care
                even tough uninjured resulting in lost time and productivity;
                furthermore, their vehicle might be damaged even though they are
                uninjured.
                ---------------------------------------------------------------------------
                 Property damage costs associated with delayed fleet turnover must
                be treated differently than rebound- and mass-related costs because
                crashes that involve vehicles that are retained longer due to the
                standards involve damage to older, used vehicles instead of newer
                vehicles.\2099\ Used vehicles are worth less and will cost less to
                repair, if they are repaired at all. The consumer's property damage
                loss is thus reduced by longer retention of these vehicles. To estimate
                this loss, average new and used vehicle prices were compared. New
                vehicle transaction prices were estimated from a study published by
                Kelley Blue Book.\2100\ Based on this data, the average new vehicle
                transaction price in January 2017 was $34,968. Used vehicle transaction
                prices were obtained from Edmonds Used Vehicle Market Report published
                in February of 2017.\2101\ Edmonds data indicate the average used
                vehicle transaction price was $19,189 in 2016. There is a minor timing
                discrepancy in these data because the new vehicle data represent
                January 2017, and the used vehicle price is for the average over 2016.
                The agencies were unable to locate exact matching data, but believe the
                difference is minor and negligible.
                ---------------------------------------------------------------------------
                 \2099\ The agencies note that property damage costs are the
                costs realized given an accident has occurred. The disparity of
                incidence rates between new and older vehicles is accounted for
                above in the fatality calculations.
                 \2100\ Press Release, ``New-Car Transaction Prices Remain High,
                Up More Than 3 Percent Year-Over-Year in January 2017, According to
                Kelley Blue Book,'' February 1, 2017, available at https://mediaroom.kbb.com/2017-02-01-New-Car-Transaction-Prices-Remain-High-Up-More-Than-3-Percent-Year-Over-Year-In-January-2017-According-To-Kelley-Blue-Book.
                 \2101\ Edmonds Used Vehicle Market Report, February 2017.
                Available at https://dealers.edmunds.com/static/assets/articles/2017_Feb_Used_Market_Report.pdf.
                ---------------------------------------------------------------------------
                 Based on these data, new vehicles are on average worth 82 percent
                more than used vehicles. To estimate the effect of higher property
                damage costs for newer vehicles in crashes, the per unit property
                damage costs from Table I-9 in Blincoe et al. (2015) were multiplied by
                this factor.\2102\ Results are illustrated in Table VI-265.
                ---------------------------------------------------------------------------
                 \2102\ The original unit costs were derived from vehicles
                involved in crashes, which are predominately used vehicles. While
                not precise, we assume this average cost is a reasonable proxy for
                the property damage to a used vehicle.
                ---------------------------------------------------------------------------
                [[Page 24834]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.490
                 The total property damage cost reduction was then calculated as a
                function of the number of increased fatalities due to stricter CAFE and
                CO2 standards as follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.491
                Where:
                 S = total property damage reductions from retaining used
                vehicles longer
                 F = increase in fatalities estimated due to used vehicles
                being retained longer because of stricter standards
                 r = ratio of non-fatal injuries or PDO vehicles to
                fatalities
                 p = value of property damage prevented by retaining older
                vehicle
                 n = the 8 injury severity categories
                 The number of fatalities ascribed to the standards because of
                slower fleet turnover was multiplied by the unit cost per fatality from
                Table I-9 in Blincoe et al. (2015) to determine the societal impact of
                fatalities.\2103\ After subtracting the total reductions in property
                damage from this value, the agencies divided the fatality cost by it to
                estimate that overall, fatalities account for 39 percent of the total
                costs that would result from older vehicle retention.
                ---------------------------------------------------------------------------
                 \2103\ Note--These calculations used the original values in the
                Blincoe et al. (2015) tables without adjusting for economics. These
                calculations produce ratios and are thus not sensitive to
                adjustments for inflation.
                ---------------------------------------------------------------------------
                 These calculations are summarized as follows:
                SV = Fv/x-s
                Where:
                 SV = Value of societal impacts of all crashes resulting
                from changes to fleet turnover
                 F = Increase in fatalities estimated due to retaining used
                vehicles longer because of stricter standards
                 v = Comprehensive societal value of preventing 1 fatality
                 x = Percent of total societal loss from crashes
                attributable to fatalities
                 S = total property damage reductions from retaining used
                vehicles longer
                 For the fatalities that occur because of mass effects or to the
                rebound effect, the calculation was more direct, a simple application
                of the ratio of the portion of costs produced by fatalities to the
                change in fatalities; there is no need to adjust for property damage
                because all impacts were derived from the mix of vehicles in the on-
                road fleet. Again, from Table I-8 in Blincoe et al. (2015), the
                agencies derived this ratio based on all cost factors including
                property damage to be 36 percent.
                 For purposes of application in the CAFE model, these two factors
                (the factor for sales/scrappage, and the factor for mass and rebound)
                were combined based on the relative contribution to total fatalities of
                different factors. As noted previously, although a safety impact from
                the rebound effect is calculated, these impacts are considered to be
                freely chosen rather than imposed by the standards and imply personal
                benefits at least equal to the sum of their added operational costs and
                the portion of safety consequences internalized. However, the agencies
                still calculate and report the impacts of the rebound effect to provide
                a comprehensive view of the impacts of the standards. There are two
                different factors depending on which metric is considered (total
                impacts or CAFE imposed impacts). The agencies created these two
                adjustment factors by weighting components by the relative contribution
                to changes in fatalities associated with each component. This process
                and results are shown in Table VI-266. Note that due to programming
                constraints, the agencies applied the average weighted factor to all
                fatalities. This will tend to overstate costs slightly because of sales
                and scrappage and to understate costs associated with mass and rebound.
                [[Page 24835]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.492
                f) Summary of Safety Impacts
                 Table VI-267 through Table VI-270 summarize the safety effects of
                CAFE standards across the various alternatives under the 3 percent and
                7 percent discount rates.
                 Table VI-271 through Table VI-274 summarize these impacts for
                CO2 standards. As noted in Section VI.D.2.e), societal
                impacts are valued using a $10.4 million value per statistical life
                (VSL). Note that fatalities in these tables are undiscounted--only the
                monetized societal impact is discounted.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.493
                [GRAPHIC] [TIFF OMITTED] TR30AP20.494
                [[Page 24836]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.495
                [GRAPHIC] [TIFF OMITTED] TR30AP20.496
                [[Page 24837]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.497
                [[Page 24838]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.498
                [[Page 24839]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.499
                [[Page 24840]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.500
                [[Page 24841]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.501
                [[Page 24842]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.502
                [[Page 24843]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.503
                BILLING CODE 4910-59-C
                 These tables present aggregations or averages of results for
                calendar years through 2050. Underlying model output files provide
                results for each model year in each calendar year.\2104\ These results
                can be used for more detailed review and analysis of estimated trends.
                For example, for each calendar year through 2050, the following two
                tables--one for CAFE standards and one for CO2 standards--
                show (a) the number of light-duty vehicles in service, (b) the travel
                accumulated by those vehicles,
                [[Page 24844]]
                and (c) the total number fatalities among the types included in today's
                analysis.
                ---------------------------------------------------------------------------
                 \2104\ FOOTNOTE 2104???
                ---------------------------------------------------------------------------
                 The analysis shows the annual number of fatalities for the final
                standards growing more slowly than under the baseline standards,
                reflecting the combined effects of fleet turnover, mass reduction, and
                shifts between passenger cars and light trucks in the new vehicle
                fleet.
                 Table VI-274 summarizes the non-fatal safety impacts under
                alternative CAFE and CO2 standards:
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.504
                [[Page 24845]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.505
                BILLING CODE 4910-59-C
                 The Pennsylvania Department of Environmental Protection commented
                that the agencies did not fully account for safety improvements
                associated with the augural standards.\2105\ The agencies note that the
                analysis accounts for the safety impacts of mass reduction, sales and
                scrappage, rebound, vehicle model year and vehicle age for each of the
                alternatives relative to the augural baseline. The commenter did not
                provide any specific items that were omitted from the analysis. The
                agencies believe the analysis thoroughly assesses the safety effects of
                all the alternatives.
                ---------------------------------------------------------------------------
                 \2105\ NOT ON MANUSCRIPT.
                ---------------------------------------------------------------------------
                Simulating Environmental Impacts of Regulatory Alternatives
                 This final rulemaking predominantly addresses fuel economy of the
                light-duty vehicle fleet in the United States through different
                technologies to improve efficiency. Inherently, these technologies will
                reduce the fuel consumed and therefore impact CO2 and other
                greenhouse gases foremost. Certain technologies will also impact air
                quality through changes to criteria pollutants and air toxics emitted
                at the tailpipe as well as upstream of the fuel source. Upstream
                emissions for conventional fuels occur during crude oil extraction,
                transportation, refining, and the transportation, storage, and
                distribution of the finished fuel. For electricity, upstream emissions
                are dependent on the mix of feedstocks such as coal, natural gas,
                nuclear, and renewable sources for power generation. Similarly,
                specific hydrogen production pathways such as natural gas reforming or
                electrolysis of water molecules will determine the upstream emissions
                of hydrogen fuel. Emission impacts are described in greater detail in
                the following sections.\2106\
                ---------------------------------------------------------------------------
                 \2106\ NHTSA also uses the results of the CAFE model to analyze
                the potential environmental impacts of the regulatory alternatives
                in its Environmental Impact Statement (EIS). That EIS informs the
                agency's decision-making process.
                ---------------------------------------------------------------------------
                 The impacts of both greenhouse gases (GHGs) and criteria pollutant
                emissions that result from changes in vehicle usage and fuel
                consumption were estimated and considered as part of this analysis.
                GHGs are gaseous constituents in the atmosphere, both natural and
                anthropogenic, and absorb infrared radiation. Primary GHGs in the
                atmosphere are water vapor, CO2, nitrous oxide
                (N2O), methane (CH4), and ozone. Criteria air
                pollutants include carbon monoxide (CO), nitrogen dioxide
                (NO2) (one of several oxides of nitrogen), ozone, sulfur
                dioxides (SO2), particulate matter (including fine
                particulate matter, or PM2.5), and lead. Vehicles do not
                directly emit ozone, but ozone impacts are evaluated based on emissions
                of the ozone precursor pollutants nitrogen oxides (NOX) and
                volatile organic compounds (usually referred to as VOC). These
                pollutants are emitted during vehicle storage and use, as well as
                throughout the fuel production and distribution system. While increases
                in domestic fuel refining, storage, and distribution that result from
                higher fuel consumption will increase emissions of these pollutants,
                reduced vehicle use associated with the fuel economy rebound effect
                will decrease their emissions. The net effect of CAFE and
                CO2 standards on total emissions of each criteria pollutant
                depends on the relative magnitudes of increases in its emissions during
                fuel refining and distribution, and decreases in its emissions
                resulting from vehicle use. Because the relationship between emissions
                in fuel refining and vehicle use is different for each criteria
                pollutant, the net effect of fuel consumption on total emissions of
                each pollutant differs between regulatory alternatives.
                Climate Change and CO2 Emissions Considered in This Rule
                 The NPRM described how both agencies consider climate change and
                GHG emissions under their respective programs for fuel economy and
                CO2. As noted in the NPRM, ``In 1988, NHTSA included climate
                change concepts in its CAFE notices and prepared its first
                environmental assessment addressing that subject.'' \2107\
                Additionally, NHTSA ``cited concerns about climate change as one of its
                reasons for limiting the extent of its reduction of the CAFE standard
                for MY 1989 passenger cars.'' \2108\ As stated in the NPRM, ``Since
                then, NHTSA has considered the effects of reducing tailpipe emissions
                of CO2 in its fuel economy rulemakings pursuant to the need
                of the United States to conserve energy by reducing petroleum
                consumption.\2109\
                ---------------------------------------------------------------------------
                 \2107\ 83 FR 43211 (citing 53 FR 33080, 33096 (Aug. 29, 1988)).
                 \2108\ Id. (citing 53 FR 39275, 39302 (Oct. 6, 1988)).
                 \2109\ 83 FR 43211.
                ---------------------------------------------------------------------------
                 Similarly, in the NPRM, EPA described that ``the primary purpose of
                Title II of the Clean Air Act is the protection of public health and
                welfare. EPA's light-duty vehicle GHG standards serve this purpose, as
                the GHG emissions from light-duty vehicles have been found by EPA to
                endanger public health and welfare (see EPA's 2009 Endangerment Finding
                for on-highway motor vehicles), and the goal of these standards is to
                reduce these emissions that contribute to climate change.'' \2110\ In
                the NPRM, EPA summarized its purpose for establishing CO2
                standards as follows:
                ---------------------------------------------------------------------------
                 \2110\ 83 FR 4228 (citing 74 FR 66496 (Dec. 15, 2009)).
                 Section 202(a)(1) of the Clean Air Act (CAA) states that ``the
                Administrator shall by regulation prescribe (and from time to time
                revise) . . . standards applicable to the emission of any air
                pollutant from any class or classes of new motor vehicles . . . ,
                which in his judgment cause, or contribute to, air
                [[Page 24846]]
                pollution which may reasonably be anticipated to endanger public
                health or welfare.'' If EPA makes the appropriate endangerment and
                cause or contribute findings, then section 202(a) authorizes EPA to
                issue standards applicable to emissions of those pollutants. Indeed,
                EPA's obligation to do so is mandatory: Coalition for Responsible
                Regulation, 684 F.3d at 114; Massachusetts v. EPA, 549 U.S. at
                533.\2111\
                ---------------------------------------------------------------------------
                 \2111\ 83 FR 43228.
                 The agencies modeled the estimated physical changes in quantity of
                CO2, CH4, and NO2 emissions in the
                NPRM analysis, and conducted additional modeling of climate-related
                impacts, including sea-level rise, global temperate increases, and
                ocean pH changes in the Draft EIS accompanying the NPRM. The Draft EIS
                also included a comprehensive discussion of climate change impacts,
                drawing from various Intergovernmental Panel on Climate Change (IPCC)
                reports, the U.S. Global Change Research Program (USGCRP) National
                Climate Assessment (NCA) reports, and other peer-reviewed reports and
                assessment reports. The agencies also considered the increase in
                climate damages from an increase in CO2 emissions,\2112\
                also known as the social cost of carbon and discussed previously in
                Section VI.D.1, above.
                ---------------------------------------------------------------------------
                 \2112\ 83 FR 43106.
                ---------------------------------------------------------------------------
                 Many commenters expressed a desire for more information on the
                rule's potential climate impacts, so the discussion has been expanded
                here and in the Final EIS. Specifically, commenters stated that the
                agencies failed to address climate change in the proposal, and that the
                proposal ignored ``scores of studies and reports'' on climate change
                published since EPA's 2009 Endangerment Finding and promulgation of the
                existing CO2 and CAFE standards.\2113\ Several commenters
                presented summaries of climate impacts, citing IPCC, USGCRP, and other
                reports explicitly relied on in the DEIS, on temperature increases,
                increases in extreme weather events, ocean warming, acidification, and
                sea level rise, impacts on the United States' water supply, human
                health impacts, impacts to crop productivity and global food security,
                potential increases in the spread of infectious disease, national
                security impacts, and impacts to animal and plant species, including
                Federally protected species, among other impacts.\2114\
                ---------------------------------------------------------------------------
                 \2113\ NHTSA-2018-0067-12088.
                 \2114\ NHTSA-2018-0067-11735; NHTSA-2018-0067-11926; NHTSA-2018-
                0067-11972; NHTSA-2018-0067-12088; NHTSA-2018-0067-12127; NHTSA-
                2018-0067-12303; NHTSA-2018-0067-12378; NHTSA-2018-0067-12436.
                ---------------------------------------------------------------------------
                 In addition to comments stating the agencies had presented too
                little information on climate change in the NPRM, some commenters
                disagreed with how the agencies framed the impact of the rule on
                climate change. Many commenters cited IPCC and USGCRP to reinforce
                their understanding that human activities are the dominant cause of
                global warming since the mid-20th century. NHTSA considered both the
                IPCC and USGCRP reports in the DEIS accompanying the NPRM and in this
                final rule, and did not dispute those findings. Commenters also cited
                IPCC and the National Climate Assessments, among other reports, as
                support to their understanding that regardless of the perceived
                magnitude of the rule on total CO2 emissions, any additional
                actions taken now to reduce CO2 emissions would affect the
                degree of climate impacts in the future. Further discussion of these
                comments occurs in Section VIII.
                 Just as NHTSA does with both the draft and final EIS, and as EPA
                did for its Endangerment and Cause or Contribute Findings for
                Greenhouse Gases under the Clean Air Act, for this rule, both agencies
                relied on existing studies and reports to summarize the current state
                of climate science and provide a framework for the analysis of impacts.
                The agencies drew primarily on panel-reviewed synthesis and assessment
                reports from the Intergovernmental Panel on Climate Change (IPCC) and
                the U.S. Global Change Research Program (GCRP), supplemented with past
                reports from the U.S. Climate Change Science Program (CCSP), the
                National Research Council, and the Arctic Council and EPA's Technical
                Support Document for Endangerment and Cause or Contribute Findings for
                Greenhouse Gases under the Clean Air Act,\2115\ which, as stated above,
                relied on past major international or national scientific assessment
                reports.
                ---------------------------------------------------------------------------
                 \2115\ EPA Technical Support Document for Endangerment and Cause
                or Contribute Findings for Greenhouse Gases under Section 202(a) of
                the Clean Air Act. December 7, 2009. U.S. Environmental Protection
                Agency, Office of Atmospheric Programs, Climate Change Division:
                Washington, DC. Available at: https://www.epa.gov/sites/production/files/2016-08/documents/endangerment_tsd.pdf.
                ---------------------------------------------------------------------------
                 Assessment reports assess numerous individual studies to draw
                general conclusions about the potential impacts of climate change. Even
                where assessment reports include consensus conclusions of expert
                authors, uncertainty still exists, as with all assessments of
                environmental impacts. Given the global nature of climate change and
                the need to communicate uncertainty to a variety of decision-makers,
                IPCC has focused considerable attention on developing a systematic
                approach to characterize and communicate this information. The IPCC is
                a United Nations panel, founded in 1988, which evaluates climate
                science by assessing research on climate change and synthesizing
                relevant research into major assessment reports. The IPCC provides
                regular assessments on climate impacts and future risks, and options
                for adaptation and risk mitigation. The agencies used the system
                developed by IPCC to describe uncertainty associated with various
                climate change impacts.
                 The IPCC reports communicate uncertainty and confidence bounds
                using commonly understood but carefully defined words in italics to
                represent likelihood of occurrence. The referenced IPCC documents
                provide a full understanding of the meaning of those uncertainty terms
                in the context of the IPCC findings. The IPCC notes that there are two
                primary uncertainties with climate modeling: Model uncertainties and
                scenario uncertainties: \2116\
                ---------------------------------------------------------------------------
                 \2116\ IPCC. Climate Change 2013: The Physical Science Basis.
                Contribution of Working Group I to the Fifth Assessment Report of
                the Intergovernmental Panel on Climate Change. Stocker, T.F., D.
                Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels,
                Y. Xia, V. Bex and P.M. Midgley (Eds.). Cambridge University Press:
                Cambridge, United Kingdom and New York, NY, USA. pp. 1535. Available
                at: http://www.ipcc.ch/report/ar5/wg1/. [hereinafter IPCC 2013].
                ---------------------------------------------------------------------------
                 Model uncertainties. These uncertainties occur when a
                climate model might not accurately represent complex phenomena in the
                climate system. For some processes, the scientific understanding could
                be limited regarding how to use a climate model to ``simulate''
                processes in the climate system.
                 Scenario uncertainties. These uncertainties arise because
                of uncertainty in projecting future GHG emissions, concentrations, and
                forcings (e.g., from solar activity).
                 According to IPCC, these types of uncertainties are described by
                using two metrics for communicating the degree of certainty: Confidence
                in the validity of findings, expressed qualitatively, and quantified
                measures of uncertainties, expressed probabilistically.\2117\ The
                confidence levels synthesize the judgments about the validity of the
                findings, determined through evaluation of the evidence and the degree
                of scientific agreement. The qualitative expression of confidence
                ranges are described, in italics, from very low to very high, with
                higher confidence levels assigned to findings that are supported by
                high scientific agreement. The quantitative expression of confidence
                ranges from exceptionally unlikely to
                [[Page 24847]]
                virtually certain, with higher confidence representing findings
                supported by robust evidence. Table VI-276 shows that the degree of
                confidence increases as evidence becomes more robust and agreement is
                greater.
                ---------------------------------------------------------------------------
                 \2117\ IPCC 2013.
                 [GRAPHIC] [TIFF OMITTED] TR30AP20.506
                
                 As described in more detail in the Final EIS, the process known as
                the greenhouse effect is responsible for trapping a portion of a
                planet's heat in the planet's atmosphere, rather than allowing all of
                that heat to be radiated into space. GHGs trap heat in the lower
                atmosphere (the atmosphere extending from Earth's surface to
                approximately 4 to 12 miles above the surface), absorb heat energy
                emitted by Earth's surface and lower atmosphere, and reradiate much of
                it back to Earth's surface, thereby causing warming. Human activities,
                particularly fossil-fuel combustion, lead to the presence of increased
                concentrations of GHGs in the atmosphere; this buildup of GHGs is
                changing the Earth's energy balance. IPCC states the warming
                experienced over the past century is due to the combination of natural
                climatic forcers (e.g., natural GHGs, solar activity) and human-made
                climate forcers.\2118\ IPCC concluded, ``[h]uman influence has been
                detected in warming of the atmosphere and the ocean, in changes in the
                global water cycle, in reductions in snow and ice, in global mean sea-
                level rise, and in changes in some climate extremes. . . . This
                evidence for human influence has grown since [the IPCC Working Group 1
                (WG1) Fourth Assessment Report (AR4)]. IPCC reports that it is
                extremely likely that human influence has been the dominant cause of
                the observed warming since the mid-20th century.'' \2119\
                ---------------------------------------------------------------------------
                 \2118\ IPCC 2013.
                 \2119\ IPCC 2013.
                ---------------------------------------------------------------------------
                 Although the climate system is complex, IPCC has identified the
                following drivers of climate change:
                 GHGs. Primary GHGs in the atmosphere are water vapor,
                atmospheric CO2, N2O (nitrous oxide),
                CH4 (methane), and ozone.\2120\
                ---------------------------------------------------------------------------
                 \2120\ IPCC 2013.
                ---------------------------------------------------------------------------
                 Aerosols. Aerosols are natural (e.g., from volcanoes) and
                human-made particles in the atmosphere that scatter incoming sunlight
                back to space, causing cooling. Some aerosols are hygroscopic (i.e.,
                attract water) and can affect the formation and lifetime of clouds.
                Large aerosols (more than 2.5 micrometers in size) modify the amount of
                outgoing long-wave radiation.\2121\ Other particles, such as black
                carbon, can absorb outgoing terrestrial radiation, causing warming.
                Natural aerosols have had a negligible cumulative impact on climate
                change since the start of the industrial era.\2122\ Further discussion
                of black carbon and other aerosols is located in Chapter 4 of the FEIS.
                ---------------------------------------------------------------------------
                 \2121\ IPCC 2013.
                 \2122\ IPCC 2013.
                ---------------------------------------------------------------------------
                 Clouds. Depending on cloud height, cloud interactions with
                terrestrial and solar radiation can vary. Small changes in the
                properties of clouds can have important implications for both the
                transfer of radiative energy and weather.\2123\
                ---------------------------------------------------------------------------
                 \2123\ IPCC 2013.
                ---------------------------------------------------------------------------
                 Ozone. Ozone is created through photochemical reactions
                from natural
                [[Page 24848]]
                and human-made gases. In the troposphere, ozone absorbs and reemits
                long-wave radiation. In the stratosphere, the ozone layer absorbs
                incoming short-wave radiation.\2124\
                ---------------------------------------------------------------------------
                 \2124\ IPCC 2013.
                ---------------------------------------------------------------------------
                 Solar radiation. Solar radiation, the amount of solar
                energy that reaches the top of Earth's atmosphere, varies over time.
                Solar radiation has had a negligible impact on climate change since the
                start of the industrial era compared to other main drivers.\2125\
                ---------------------------------------------------------------------------
                 \2125\ IPCC 2013.
                ---------------------------------------------------------------------------
                 Surface changes. Changes in vegetation or land surface
                properties, ice or snow cover, and ocean color can affect surface
                albedo.\2126\ The changes are driven by natural seasonal and diurnal
                changes (e.g., snow cover) as well as human influences (e.g., changes
                in vegetation type).\2127\
                ---------------------------------------------------------------------------
                 \2126\ Surfaces on Earth (including land, oceans, and clouds)
                reflect solar radiation back to space. This reflective
                characteristic, known as albedo, indicates the proportion of
                incoming solar radiation the surface reflects. High albedo has a
                cooling effect because the surface reflects rather than absorbs most
                solar radiation.
                 \2127\ IPCC 2013.
                ---------------------------------------------------------------------------
                 Effects of emissions and the corresponding processes that affect
                climate are highly complex and variable, which complicates the
                measurement and detection of change. However, IPCC indicates that an
                increasing number of studies conclude that anthropogenic GHG emissions
                are affecting climate in detectable and quantifiable
                ways.2128 2129 GHGs occur naturally and because of human
                activity. Other GHGs, such as the fluorinated gases,\2130\ are
                primarily anthropogenic in origin and are used in commercial
                applications such as refrigeration and air conditioning and industrial
                processes such as aluminum production.
                ---------------------------------------------------------------------------
                 \2128\ IPCC. Summary for Policymakers. In: Change 2013: The
                Physical Science Basis. Contribution of Working Group I to the Fifth
                Assessment Report of the Intergovernmental Panel on Climate Change.
                Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J.
                Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (Eds.).
                Cambridge University Press: Cambridge, United Kingdom and New York,
                NY, USA. 1535 pp. Available at: http://www.ipcc.ch/pdf/assessment-report/ar5/wg1/WG1AR5_SPM_FINAL.pdf.
                 \2129\ GCRP. 2017. Climate Science Special Report: Fourth
                National Climate Assessment. U.S. Global Change Research Program.
                [Wuebbles, D.J., D.W. Fahey, K.A. Hibbard, D.J. Dokken, B.C.
                Stewart, and T.K. Maycock (Eds.)]. U.S. Government Printing Office:
                Washington, DC 477 pp. doi:10.7930/J0J964J6. Available at: https://science2017.globalchange.gov/downloads/CSSR2017_FullReport.pdf.
                [hereinafter GCRP 2017].
                 \2130\ Fluorinated GHGs or gases include PFCs, HFCs,
                SF6, and NF3.
                ---------------------------------------------------------------------------
                 In its most recent assessment of climate change (IPCC WG1 AR5),
                IPCC states that, ``Warming of the climate system is unequivocal, and
                since the 1950s, many of the observed changes are unprecedented over
                decades to millennia. The atmosphere and ocean have warmed, the amounts
                of snow and ice have diminished, sea level has risen, and the
                concentrations of greenhouse gases have increased.'' \2131\ IPCC
                concludes that, at continental and global scales, numerous long-term
                changes in climate have been observed. To be more specific, IPCC and
                the GCRP include the following trends observed over the 20th century as
                further supporting the evidence of climate-induced changes:
                ---------------------------------------------------------------------------
                 \2131\ IPCC 2013.
                ---------------------------------------------------------------------------
                 Most land areas have very likely experienced warmer and/or
                fewer cold days and nights along with warmer and/or more frequent hot
                days and nights.2132 2133 From 1880 to 2016, the global mean
                surface temperature rose by about 0.9 [deg]C (1.6 [deg]F).\2134\ Air
                temperatures are warming more rapidly over land than over
                oceans.2135 2136 Similar to the global trend, the U.S.
                average temperature is about 1.8 [deg]F warmer than it was in 1895, and
                this rate of warming is increasing--most of the warming has occurred
                since 1970.\2137\ IPCC projects a continuing increase in surface
                temperature between 2081 and 2100, with a likely range between 0.3
                [deg]C (0.5 [deg]F) and 4.8 [deg]C (8.6 [deg]F), compared with 1986
                through 2005, where the lower value corresponds to substantial future
                mitigation of carbon emissions.\2138\
                ---------------------------------------------------------------------------
                 \2132\ IPCC Climate Change 2014: Impacts, Adaptation, and
                Vulnerability. Part A: Global and Sectoral Aspects. Contribution of
                Working Group II to the Fifth Assessment Report of the
                Intergovernmental Panel on Climate Change. Field, C.B., V.R. Barros,
                D.J. Dokken, K.J. Mach, M.D. Mastrandrea, T.E. Bilir, M. Chatterjee,
                K.L. Ebi, Y.O. Estrada, R.C. Genova, B. Girma, E.S. Kissel, A.N.
                Levy, S. MacCracken, P.R. Mastrandrea, and L.L. White (Eds.).
                Cambridge University Press: Cambridge, United Kingdom and New York,
                NY, USA, 1132 pp. Available at: http://ipcc-wg2.gov/AR5/report/.
                [hereinafter IPCC 2014].
                 \2133\ GCRP 2017.
                 \2134\ GCRP 2017.
                 \2135\ IPCC 2013.
                 \2136\ GCRP 2017.
                 \2137\ GCRP 2017.
                 \2138\ IPCC 2013.
                ---------------------------------------------------------------------------
                 Cold-dependent habitats are shifting to higher altitudes
                and latitudes, and growing seasons are becoming
                longer.2139 2140 According to the IPCC, ``it is virtually
                certain that there will be more frequent hot and fewer cold temperature
                extremes over most land areas on daily and seasonal timescales'' and it
                is very likely that heat wave frequency and duration will also
                increase.\2141\
                ---------------------------------------------------------------------------
                 \2139\ IPCC 2014.
                 \2140\ GCRP 2017.
                 \2141\ IPCC 2014.
                ---------------------------------------------------------------------------
                 Sea level is rising, caused by thermal expansion of the
                ocean and melting of snowcaps and ice sheets.2142 2143
                Between 1971 and 2010, global ocean temperature warmed by approximately
                0.25 [deg]C (0.45 [deg]F) in the top 200 meters (approximately 660
                feet).\2144\ IPCC concludes that mountain glaciers, ice caps, and snow
                cover have declined on average, further contributing to sea-level rise.
                Losses from the Greenland and Antarctic ice sheets very likely
                contributed to sea-level rise from 1993 to 2010, and satellite
                observations confirm that they have contributed to sea-level rise in
                subsequent years.\2145\ IPCC projects that the global temperature
                increase will continue to affect sea level, causing a likely rise of
                0.26 meter (0.85 foot) to 0.82 meter (2.7 feet) in the next
                century.\2146\
                ---------------------------------------------------------------------------
                 \2142\ IPCC 2013.
                 \2143\ GCRP 2017.
                 \2144\ IPCC 2013.
                 \2145\ IPCC 2013.
                 \2146\ IPCC 2013.
                ---------------------------------------------------------------------------
                 More frequent weather extremes such as droughts, floods,
                severe storms, and heat waves have been observed.2147 2148
                Average atmospheric water vapor content has increased since at least
                the 1970s over land and the oceans, and in the upper troposphere,
                largely consistent with air temperature increases.\2149\ Because of
                changes in climate, including increased moisture content in the
                atmosphere, heavy precipitation events have increased in frequency over
                most land areas.2150 2151 Observations of increased dryness
                since the 1950s suggest that some regions of the world have experienced
                longer, more intense droughts caused by higher temperatures and
                decreased precipitation, particularly in the tropics and
                subtropics.\2152\ Heavy precipitation events have increased globally
                since 1951, with some regional and subregional variability.\2153\ A
                warmer atmosphere holds more moisture and increases the energy
                available for convection, causing stronger storms and heavier
                precipitation.2154 2155
                ---------------------------------------------------------------------------
                 \2147\ IPCC 2013.
                 \2148\ GCRP 2017.
                 \2149\ IPCC 2013.
                 \2150\ IPCC 2013.
                 \2151\ Min, S.-K., Zhang, X., Zwiers, F.W., & Hegerl, G.C. 2011.
                Human contribution to more-intense precipitation extremes. Nature,
                470(7334), pp. 378-81. Available at: https://doi.org/10.1038/nature09763.
                 \2152\ IPCC 2013.
                 \2153\ IPCC 2013.
                 \2154\ GCRP 2017.
                 \2155\ Gertlet, C., O'Gorman, P. 2019. Changing available energy
                for extratropical cyclones and associated convection in the Northern
                Hemisphere summer, PNAS 116(10):4105-4110.
                ---------------------------------------------------------------------------
                [[Page 24849]]
                 Many commenters urged the agencies to consider more stringent
                standards to address GHG emissions. The Northeast States for
                Coordinated Air Use Management (NESCAUM) stated that ``effectively
                combatting climate change requires GHG reductions on a national and
                international scale. Maintaining an aggressive downward trend in
                transportation sector GHG emissions will not occur in the absence of
                strong national GHG emission reductions.'' \2168\ Similarly, the Center
                for Biological Diversity et al. stated ``the scientific record is now
                overwhelming that climate change poses grave harm to public health and
                welfare; that its hazards have become even more severe and urgent than
                previously understood; and that avoiding devastating harm requires
                substantial reductions in greenhouse gas emissions, including from the
                critically important transport sector, within the next decade.'' \2169\
                Minnesota Pollution Control Agency (MPCA), the Minnesota Department of
                Transportation (MnDOT), and the Minnesota Department of Health (MDH)
                stated ``Tackling climate change will require aggressive and immediate
                action on reducing emissions from the transportation sector. The
                existing GHG and CAFE standards are a critical piece to the
                multifaceted and global effort to reduce GHG emissions.'' \2170\
                ---------------------------------------------------------------------------
                 \2168\ NESCAUM, NHTSA-2018-0067-11691.
                 \2169\ Center for Biological Diversity et al., NHTSA-2018-0067-
                12000.
                 \2170\ MPCA, MnDOT, and MDH, NHTSA-2018-0067-11706.
                ---------------------------------------------------------------------------
                 Commenters also expressed concerns that the agencies did not
                accurately consider the effects of climate change resulting from the
                rulemaking. Pennsylvania Department of Environmental Protection (PA
                DEP) stated ``the Proposed Rule does not fully consider the potential
                effects of global climate change resulting from these forgone
                reductions or the interests of states in preventing or mitigating the
                impacts of climate change on their citizens and environment.'' \2171\
                The Center for Biological Diversity et al. stated ``the agencies
                callously disregard the demonstrated need to reduce emissions sharply
                over the next decade if severe impacts of a destabilized climate are to
                be avoided.'' \2172\ Similarly, the Joint Submission from the States of
                California et al. and the Cities of Oakland et al. stated ``discussion
                of the effect of the Proposed Rollback on GHG emissions significantly
                understates the outcome,'' and ``the overwhelming scientific consensus
                is that immediate and continual progress toward a near-zero GHG-
                emission economy by mid-century is necessary to avoid truly
                catastrophic climate change impacts.'' \2173\
                ---------------------------------------------------------------------------
                 \2171\ PA DEP, NHTSA-2018-0067-11956.
                 \2172\ Center for Biological Diversity et al., NHTSA-2018-0067-
                12000.
                 \2173\ Joint Submission from the States of California et al. and
                the Cities of Oakland et al., NHTSA-2018-0067-11735.
                ---------------------------------------------------------------------------
                 The agencies have carefully considered these comments in the
                context of the information on climate change summarized in the NPRM and
                DEIS, and have updated information for this final rule. The agencies
                drew upon updates to climate science and impacts for the analysis from
                reports and studies that were updated or released since the NPRM,
                including IPCC's Global Warming of 1.5 degrees C report, Volume 2 of
                the 4th National Climate Assessment, and IPCC's Special Report on
                Climate Change and Land, and the IPCC's Special Report on the Ocean and
                Cryosphere in a Changing Climate.
                 The following sections also provide additional context about
                climate impacts from this final rule; the results of the agencies'
                quantitative analysis presented in Section VII shows estimated
                CO2, CH4, and N2O emissions resulting
                from the rule, and the discussion of how each agency balanced climate
                change as a factor considered in decision-making is presented in
                Section VIII. The Final EIS accompanying today's rule also includes a
                comprehensive discussion of climate impacts, and additional climate
                modeling that estimates climate-related effects. As discussed in more
                detail in the FEIS and following sections, but relevant for placing the
                following discussion in context, climate modeling performed for this
                final rule shows the following impacts as a result of the final
                standards selected: CO2 Concentrations of 789.80 ppm in
                2100, compared with 789.11 ppm under the augural standards; global mean
                surface temperature increases of 3.487 [deg]C in 2100, compared with
                3.484 [deg]C under the augural standards; sea-level rise increases of
                76.34 cm in 2100, compared with 76.28 cm under the augural standards;
                and ocean pH of 8.2172 in 2100, compared with 8.2176 under the augural
                standards. These equal differences of 0.69 ppm, 0.003 [deg]C, 0.06 cm,
                and -0.0004, respectively. Additionally, the agencies valued
                anticipated climate-related economic effects in accordance with E.O.
                13783, as discussed in Section VI.D.1.
                (1) Global Greenhouse Gas Emissions
                 According to NOAA and IPCC, Global atmospheric CO2
                concentrations have increased 46.4 percent, from approximately 278
                parts per million (ppm) in 1750 \2174\ to approximately 407 ppm in
                2018.\2175\ According to IPCC and WRI, in 2014, CO2
                emissions \2176\ accounted for 76 percent of global GHG emissions on a
                global warming potential (GWP)-weighted basis,\2177\ followed by
                CH4 (16 percent), N2O (6 percent), and
                fluorinated gases (2 percent).2178 2179 IPCC notes that
                atmospheric concentrations of CH4 and N2O
                increased approximately 150 and 20 percent, respectively, over roughly
                the same period.\2180\
                ---------------------------------------------------------------------------
                 \2174\ IPCC 2013.
                 \2175\ NOAA. Globally Averaged Marine Surface Annual Mean
                CO2 Data. Available at: ftp://aftp.cmdl.noaa.gov/products/trends/co2/co2_annmean_gl.txt.
                 \2176\ These global GHG estimates do not include contributions
                from land-use change and forestry or international bunker fuels.
                 \2177\ Each GHG has a different radiative efficiency (the
                ability to absorb infrared radiation) and atmospheric lifetime. To
                compare their relative contributions, GHG emission quantities are
                converted to carbon dioxide equivalent (CO2e) using the
                100-year time horizon global warming potential (GWP) as reported in
                IPCC's Second Assessment Report (AR2): The Science of Climate Change
                in Sections B.7 Summary of Radiative Forcing and B.8 Global Warming
                Potential.
                 \2178\ IPCC. 1996. Second Assessment: Climate Change 1995.
                Inventories. Available at: https://www.ipcc.ch/site/assets/uploads/2018/06/2nd-assessment-en.pdf.
                 \2179\ WRI (World Resources Institute). 2018. Climate Analysis
                Indicators Tool (CAIT) 2.0: WRI's Climate Data Explorer. Available
                at: http://cait.wri.org/. [hereinafter WRI 2018].
                 \2180\ IPCC 2013.
                ---------------------------------------------------------------------------
                 According to WRI, developed countries, including the United States,
                have been responsible for the majority of historical GHG emissions
                since the mid-1800s and still have some of the highest GHG emissions
                per capita.\2181\ While annual emissions from developed countries have
                been relatively flat over the last few decades, world population
                growth, industrialization, and increases in living standards in
                developing countries are expected to cause global fossil-fuel use and
                resulting GHG emissions to grow substantially. According to IPCC,
                global GHG emissions since 2000 have been increasing nearly three times
                faster than in the 1990s.\2182\ This is further illustrated in Figure
                VI-88 showing carbon dioxide emissions since 1990 by world region:
                \2183\
                ---------------------------------------------------------------------------
                 \2181\ WRI 2018.
                 \2182\ IPCC 2013.
                 \2183\ EPA's Climate Change Indicators in the United States,
                2016: www.epa.gov/climate-indicators. Data source: WRI, 2015.
                ---------------------------------------------------------------------------
                [[Page 24850]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.507
                 GHGs are emitted from a wide variety of sectors, including energy,
                industrial processes, waste, agriculture, and forestry. According to
                WRI, the energy sector is the largest contributor of global GHG
                emissions, accounting for 72 percent of global emissions in 2014; other
                major contributors of GHG emissions are agriculture (10 percent) and
                industrial processes (6 percent).\2184\ Transportation CO2
                emissions--from the combustion of petroleum-based fuels--account for
                roughly 15 percent of total global GHG emissions, and have increased by
                64 percent from 1990 to 2014.2185 2186
                ---------------------------------------------------------------------------
                 \2184\ WRI 2018.
                 \2185\ The energy sector is largely composed of emissions from
                fuels consumed in the electric power, transportation, industrial,
                commercial, and residential sectors. The 15 percent value for
                transportation is therefore included in the 72 percent value for
                energy.
                 \2186\ WRI 2018.
                ---------------------------------------------------------------------------
                 In general, global GHG emissions continue to increase, although
                annual increases vary according to factors such as weather, energy
                prices, and economics. Comparing observed carbon emissions to projected
                emissions, the current global trajectory is similar to the most fossil
                fuel-intensive emissions scenario (A1Fi) in the IPCC Special Report on
                Emissions Scenarios (2000) and the highest emissions scenario (RCP8.5)
                represented by the more recent Representative Concentration Pathways
                (RCP).2187 2188
                ---------------------------------------------------------------------------
                 \2187\ The Representative Concentration Pathways (RCPs) were
                developed for the IPCC AR5 report. They define specific pathways to
                emission concentrations and radiative forcing in 2100. The RCPs
                established four potential emission concentration futures, a
                business-as-usual pathway (RCP8.5), two stabilization pathways
                (RCP6.0, 4.5), and an aggressive reduction pathway (RCP2.6).
                 \2188\ IPCC 2013.
                ---------------------------------------------------------------------------
                (2) U.S. Greenhouse Gas Emissions and the Transportation Sector
                 Most GHG emissions in the United States are from the energy sector,
                with the majority of those being CO2 emissions coming from
                the combustion of fossil fuels. Fossil fuel combustion CO2
                emissions alone account for 76 percent of total U.S.GWP-weighted
                emissions, with the remaining 24 percent contributed by other sources
                such as industrial processes and product use, agriculture and forestry,
                and waste.\2189\ CO2 emissions due to combustion of fossil
                fuels are from fuels consumed in the transportation (37 percent of
                fossil fuel combustion CO2 emissions), electric power (35
                percent), industrial (16 percent), residential (6 percent), and
                commercial (5 percent) sectors.\2190\ In 2017, U.S. GHG emissions were
                estimated to be 6,456.7 MMTCO2e,\2191\ or approximately 14
                percent of global GHG emissions.2192 2193
                ---------------------------------------------------------------------------
                 \2189\ EPA Inventory of U.S. Greenhouse Gas Emissions and Sinks:
                1990-2017. EPA 430-R-19-001. U.S. Environmental Protection Agency.
                Washington DC Available at: https://www.epa.gov/sites/production/files/2019-04/documents/us-ghg-inventory-2019-main-text.pdf.
                [hereinafter EPA 2019].
                 \2190\ EPA 2019.
                 \2191\ Most recent year for which an official EPA estimate is
                available. EPA 2019.
                 \2192\ Based on global and U.S. estimates for 2014, the most
                recent year for which a global estimate is available. Excluding
                emissions and sinks from land-use change and forestry and
                international bunker fuels.
                 \2193\ WRI 2018.
                ---------------------------------------------------------------------------
                 Similar to the global trend, CO2 is by far the primary
                GHG emitted in the U.S.,
                [[Page 24851]]
                representing 82 percent of U.S. GHG emissions in 2017 (on a GWP-
                weighted basis),\2194\ and accounting for 15 percent of total global
                CO2 emissions.2195 2196 Although CO2
                is the GHG with the largest contribution to warming, methane accounts
                for 10.2 percent of U.S. GHGs on a GWP-weighted basis, followed by
                N2O (5.6 percent) and the fluorinated gases (2.6
                percent).\2197\
                ---------------------------------------------------------------------------
                 \2194\ EPA 2019.
                 \2195\ The estimate for global emissions from the World
                Resources Institute is for 2014, the most recent year with available
                data for all GHGs. It excludes emissions and sinks from land use
                change and forestry.
                 \2196\ WRI 2018.
                 \2197\ EPA 2019.
                ---------------------------------------------------------------------------
                 When U.S. CO2 emissions are apportioned by end use,
                transportation is the single leading source of U.S. emissions from
                fossil fuels, causing over one-third of total CO2 emissions
                from fossil fuels.\2198\ Passenger cars and light trucks account for 59
                percent of total U.S. CO2 emissions from transportation, an
                increase of 14 percent since 1990.\2199\ This increase in emissions is
                attributed to about 50 percent increase in vehicle miles traveled (VMT)
                because of population growth and expansion, economic growth, and low
                fuel prices. Additionally, the rising popularity of sport utility
                vehicles and other light trucks with lower fuel economy than passenger
                cars has contributed to higher emissions.2200 2201 Although
                emissions typically increased over this period, emissions declined from
                2008 to 2009 because of decreased economic activity associated with the
                most recent recession.\2202\
                ---------------------------------------------------------------------------
                 \2198\ Apportioning by end use allocates emissions associated
                with electricity generation to the sectors (residential, commercial,
                industrial, and transportation) where it is used. EPA 2019.
                 \2199\ EPA 2019.
                 \2200\ EPA 2019.
                 \2201\ DOT. 2016. Table 4-23: Average Fuel Efficiency of U.S.
                Light Duty Vehicles. U.S. Department of Transportation, Bureau of
                Transportation Statistics. Available at: https://www.rita.dot.gov/bts/sites/rita.dot.gov.bts/files/publications/national_transportation_statistics/html/table_04_23.html.
                 \2202\ EPA 2019.
                ---------------------------------------------------------------------------
                 Today's rule addresses light-duty vehicle fuel economy and
                CO2 emissions from new-model passenger cars and light
                trucks. Several commenters observed that the transportation sector
                accounted for a large, if not the largest, portion of the United States
                greenhouse gas emissions, and that light-duty vehicle emissions
                contributed to a large fraction of that portion.\2203\ Many commenters
                referenced the IPCC Report from 2018 on Global Warming of 1.5 Degrees
                Celsius, which considered transportation sector greenhouse gas
                emissions in describing pathways to limit climate impacts.
                ---------------------------------------------------------------------------
                 \2203\ NHTSA-2018-0067-11284; NHTSA-2018-0067-10966; NHTSA-2018-
                0067-11691; NHTSA-2018-0067-11735; NHTSA-2018-0067-11765; NHTSA-
                2018-0067-11921; NHTSA-2018-0067-12000; NHTSA-2018-0067-12021;
                NHTSA-2018-0067-12022; NHTSA-2018-0067-12088; NHTSA-2018-0067-12303;
                NHTSA-2018-0067-4159.
                ---------------------------------------------------------------------------
                 Graphically, historical trends in U.S. GHG emissions reported by
                EPA appear as follows.\2204\
                ---------------------------------------------------------------------------
                 \2204\ Historical data from https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks. The asterisk
                indicates that the chart does not include reported emissions changes
                attributable to land use, land use change, and forestry (LULUCF).
                ---------------------------------------------------------------------------
                [[Page 24852]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.508
                 Notably, light-duty vehicle CO2 emissions outweigh other
                GHG emissions from light-duty vehicles, and light-duty vehicle
                CO2 emissions have been relatively stable over a nearly 30-
                year period during which highway vehicles miles traveled has increased
                by about 50 percent.\2205\ Without fuel economy increases that have
                accumulated since EPCA's passage in 1975, recent light-duty vehicle
                CO2 emissions would have been 50 percent greater than shown
                above.\2206\
                ---------------------------------------------------------------------------
                 \2205\ https://www.fhwa.dot.gov/policyinformation/travel_monitoring/historicvmt.pdf.
                 \2206\ DOT reports fuel economy levels of the historical on-road
                fleet at https://www.bts.gov/content/average-fuel-efficiency-us-light-duty-vehicles.
                ---------------------------------------------------------------------------
                 For fuel combustion, EIA's National Energy Modeling System (NEMS),
                which EIA uses to produce its Annual Energy Outlook (AEO) forecasts of
                U.S. energy consumption and supply, provides corresponding estimates of
                CO2 emissions. For the final rule, modeling conducted by the
                agencies using the AEO2019 version of NEMS shows the following levels
                of future CO2 emissions from sectors other than light-duty
                vehicles (which this rule impacts directly) and refineries (which this
                rule is estimated to impact through changes in fuel consumption):
                [[Page 24853]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.509
                 As this chart indicates, EIA's representation of laws and
                regulations current as of AEO2019 shows aggregate emissions from these
                sectors remaining remarkably stable through 2050, despite projected
                growth in the U.S. population and economy.
                 The agencies agree with commenters that the transportation sector,
                and specifically light-duty vehicle emissions, contribute to the
                largest portion of the United States' greenhouse gas emissions.\2207\
                However, the fuel economy and CO2 of vehicles, regulated in
                this rulemaking, is not the only determining factor for whether the
                light-duty transportation sector would see a rise or decline in
                CO2 emissions. As discussed elsewhere in this rule, the
                standards from the final rule affect only new vehicles, which are
                responsible for approximately 3.5 percent of on-road VMT in any year.
                The agencies recognize that the revised standards result in additional
                CO2 emissions, and these emissions are accounted for in the
                analysis. It is worthwhile to note that the difference between the
                augural standard and the new standard is a small change to a small
                fraction of total VMT, and it is important to consider in context the
                different mechanisms that contribute to transportation sector
                greenhouse gas emissions. These mechanisms are considered in the 2018
                IPCC special report cited by commenters as well; in addition to vehicle
                fuel efficiency, IPCC considers preventing (or reducing) the need for
                transport,\2208\ as ``increasingly efficient fleets of vehicles over
                time . . . does not necessarily limit the driven distance.'' (internal
                citations omitted).\2209\
                ---------------------------------------------------------------------------
                 \2207\ See U.S. Energy Information Administration available at
                https://www.eia.gov/todayinenergy/detail.php?id=29612 and EPA,
                Sources of Greenhouse Gas Emissions available at https://www.epa.gov/ghgemissions/sources-greenhouse-gas-emissions.
                 \2208\ IPCC 2018 at 349 (citing Gota et al., 2018).
                 \2209\ IPCC 2018 at 377 (citing Ajanovic and Haas, 2017; Sen et
                al., 2017).
                ---------------------------------------------------------------------------
                b) Air Quality
                 This section discusses the health and environmental effects
                associated with exposure to some of the criteria and air toxic
                pollutants impacted by the proposed vehicle standards. The agencies
                note that these impacts are, compared to the impacts on vehicular fuel
                consumption and CO2 emissions, small and mixed. CAFE and
                CO2 standards directly impact vehicular fuel consumption and
                CO2 emissions. Notwithstanding modest indirect impacts, such
                as impacts on vehicle
                [[Page 24854]]
                sales, retention, and mileage accumulation, one can ``draw a direct
                line'' between CAFE/CO2 standards and resultant changes in
                overall fuel consumption and CO2 emissions, and these follow
                the expected trends.
                 Changes in emissions of criteria pollutants due to these rules will
                impact air quality. The Clean Air Act (CAA) is the primary federal
                statute that addresses air quality. Pursuant to its CAA authority, the
                EPA has established National Ambient Air Quality Standards (NAAQS) for
                six criteria pollutants: CO, NO2, ozone, SO2,
                particulate matter (PM), and lead. Vehicles do not directly emit ozone,
                but ozone impacts are evaluated based on emissions of the ozone
                precursor pollutants nitrogen oxides (NOX) and volatile
                organic compounds (VOC). When the measured concentrations of a criteria
                pollutant in a geographic region are less than those permitted by
                NAAQS, EPA designates the region as an attainment area for that
                pollutant; regions where concentrations of criteria pollutants exceed
                Federal standards are called nonattainment areas. Former nonattainment
                areas that are now in compliance with NAAQS are designated as
                attainment areas and are commonly referred to as maintenance areas.
                Each state with a nonattainment area is required to develop and
                implement a State Implementation Plan (SIP) documenting how the region
                will reach attainment levels within periods specified in the CAA. For
                maintenance areas, the SIP must document how the State intends to
                maintain compliance with NAAQS. When EPA changes a NAAQS, each State
                must revise its SIP to address how it plans to attain the new standard.
                In addition to analyzing criteria pollutants, the agencies considered
                hazardous air pollutants emitted from vehicles that are known or
                suspected to cause cancer or other serious health and environmental
                impacts and are referred to as mobile source air toxics, as further
                discussed in this section. Table VI-277 below provides an overview of
                criteria pollutants and mobile source air toxics with a high level
                overview of health effects. See further within this section for details
                on the pollutants and toxics.
                BILLING CODE 4910-59-P
                [[Page 24855]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.510
                [[Page 24856]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.511
                BILLING CODE 4910-59-C
                 The CAA requires the EPA to review periodically the NAAQS and the
                supporting science, and to revise the standards as appropriate.\2210\
                Schedules for recently completed and ongoing reviews are summarized
                here. In February 2019, the EPA issued a decision to retain the
                existing primary NAAQS for SO2.\2211\ For the ongoing
                reviews of the NAAQS for PM and ozone, the EPA intends to issue
                proposed decisions in early 2020 and final decisions in late 2020.
                ---------------------------------------------------------------------------
                 \2210\ https://www.epa.gov/criteria-air-pollutants/naaqs-table.
                 \2211\ 84 FR 9866 (March 18, 2019).
                ---------------------------------------------------------------------------
                 Nationally, levels of PM2.5, ozone, NO2,
                SO2, CO and air toxics have declined significantly in the
                last 30 years. However, as of January 31, 2020, more than 130 million
                people lived in counties designated nonattainment for one or more of
                the NAAQS, and this figure does not include the people living in areas
                with a risk of exceeding a NAAQS in the future. Many Americans continue
                to be exposed to ambient concentrations of air toxics at levels which
                have the potential to cause adverse health effects. In addition,
                populations who live, work, or attend school near major roads
                experience elevated exposure concentrations to a wide range of air
                pollutants. As discussed in the FEIS, concentrations of many air
                pollutants are elevated near high-traffic roadways. If minority
                populations and low-income populations disproportionately live near
                such roads, then an issue of environmental justice (EJ) may be present.
                Comments were received from multiple entities expressing concern about
                emissions and EJ communities. The agencies considered EJ when
                considering the effects of this rule; EJ considerations and EJ-related
                comments received on the NPRM and DEIS are discussed in Section X and
                the FEIS.
                 Total emissions from on-road mobile sources (highway vehicles) have
                declined dramatically since 1970 because of pollution controls on
                vehicles and regulation of the chemical content of fuels, despite
                continuing increases in vehicle miles traveled (VMT). From 1970 to
                2016, emissions from on-road mobile sources declined 89 percent for CO,
                71 percent for NOX, 59 percent for PM2.5, 40
                percent for PM10, 93 percent for SO2, and 90
                percent for VOCs.\2212\ The figure below further shows the highway
                vehicle emissions trends that indicate reduced pollutants regulated
                under NAAQS.
                ---------------------------------------------------------------------------
                 \2212\ See https://www.epa.gov/transportation-air-pollution-and-climate-change/accomplishments-and-success-air-pollution-transportation https://gispub.epa.gov/air/trendsreport/2019/#home.
                ---------------------------------------------------------------------------
                [[Page 24857]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.410
                 Many commenters expressed concerns about the increase of emissions
                leading to regions in nonattainment for ozone and particulate matter
                and concerns regarding the inability to meet the NAAQS. The Center for
                Biological Diversity et al., and a number of State and local
                governments and government agencies asserted that State and local
                jurisdictions would be at jeopardy of becoming nonattainment areas
                under the proposed rule.\2213\ CARB and the joint submission from the
                States of California and Cities of Oakland stated that the proposed
                rule would result in ``increases in emissions [which] will undermine
                state implementation plans'' and the proposed rule ``would create an
                additional 1.24 tons per day of NOX emissions in the South
                Coast basin.'' \2214\ The South Coast Air Quality Management District
                (SCAQMD) stated ``[a]s a regional air quality district, we have limited
                authority to control emissions from mobile sources, and rely on the
                Federal government to take action,'' and they expressed concern about
                meeting the NAAQS under the proposed rule because, to meet that
                standard, the Basin would have to ``reduce NOX emissions by
                45% beyond existing requirements.'' \2215\
                ---------------------------------------------------------------------------
                 \2213\ Center for Biological Diversity, et al., NHTSA-2018-0067-
                12123.
                 \2214\ CARB, NHTSA-2018-0067-11873, Joint Submission from States
                of California and Cities of Oakland, NHTSA-2018-0067-11735.
                 \2215\ SCAQMD, NHTSA-2018-0067-11813.
                ---------------------------------------------------------------------------
                 In particular, commenters including PA DEP, the Regional Air
                Pollution Control Agency (RAPCA), and CARB, expressed the importance of
                existing CAFE standards in meeting the NAAQS.\2216\ The Northeast
                States for Coordinated Air Use Management (NESCAUM) also asserted that
                regulation and reduction of GHG was necessary to meet the NAAQS, and
                ``[o]ur states recognize the urgent need to reduce GHG emissions across
                all sectors of our economy.'' \2217\ Similarly, the agencies from
                Minnesota stated that ``[t]he existing standards are critical for
                states to attain and maintain the NAAQS because vehicles account for
                about 24% of Minnesota's overall air pollution emissions.'' \2218\ The
                Pima County Department of Environmental quality stated that
                ``[f]reezing emission reductions for six years could put this region in
                jeopardy of being designated as non-attainment of the ozone standard
                and impact the health of many of our most vulnerable residents.''
                \2219\ The Washington State Department of Ecology stated that increases
                in NOX and VOC would increase ozone levels in two areas at
                rise of ozone nonattainment in the Puget Sound and the Tri-Cities.''
                \2220\ The Pennsylvania Department of Environmental Protection stated
                ``[r]emoving currently realized emissions reductions and forgoing
                future achievable emissions reductions may make it more difficult for
                areas to attain and maintain the NAAQS. PADEP relies on emission
                reductions from mobile sources as part of its SIP planning to attain
                and maintain the
                [[Page 24858]]
                NAAQS.'' \2221\ The North Carolina Department of Environmental Quality
                asserted that based on modeling analysis conducted by NCDEQ, ``we
                believe that the fleet changes predicted by the CAFE modeling would
                lead to emissions increases that would interfere with the ability of
                some ozone maintenance areas to meet transportation conformity budgets
                and maintain compliance with the NAAQS.'' \2222\
                ---------------------------------------------------------------------------
                 \2216\ PA DEP, NHTSA-2018-0067-11956, RAPCA NHTSA-2018-0067-
                11620, and CARB NHTSA-2018-0067-11873.
                 \2217\ NESAUM, NHTSA-2018-0067-11691.
                 \2218\ Minnesota Pollution Control Agency(MPCA), the Minnesota
                Department of Transportation (MnDOT), and the Minnesota Department
                of Health(MDH), NHTSA-2018-0067-11706.
                 \2219\ Pima County Department of Environmental Quality, NHTSA-
                2018-0067-11876.
                 \2220\ Washington State Department of Ecology, NHTSA-2018-0067-
                11926.
                 \2221\ PA DEP, NHTSA-2018-0067-11956.
                 \2222\ North Carolina Department of Environmental Quality,
                NHTSA-2018-0067-12025.
                ---------------------------------------------------------------------------
                 Many State commenters also expressed concern about their ability to
                conform with their State Implementation Plan (SIP) after this rule, as
                the Federal vehicle emissions standards previously set were
                incorporated into the SIPs and a rollback could result in further
                increased emissions.\2223\ CARB stated that its ``2016 SIP calls for
                reducing NOX emissions by approximately 6 tons per day,''
                and according to CARB, the proposed rule would not allow California to
                achieve its South Coast SIP commitments without dramatic
                countermeasures to reduce emissions elsewhere.\2224\ Similarly, other
                agencies expressed concern about SIP requirements, such as PA DEP, who
                stated that ``[b]y flatlining emissions standards at the MY 2020 level,
                the agencies' Proposed Rule increases vehicle emissions. The Proposed
                Rule would interfere with Pennsylvania's SIP planning requirements.''
                \2225\
                ---------------------------------------------------------------------------
                 \2223\ CARB NHTSA-2018-0067-11873, SCAQMD NHTSA-2018-0067-11813,
                NESCAUM NHTSA-2018-0067-11691, Joint Submission from Colorado local
                governments NHTSA-2018-0067-11929, PA DEP NHTSA-2018-0067-11956, and
                Joint Submission from the States of California et al. and the Cities
                of Oakland et al. NHTSA-2018-0067-11735.
                 \2224\ CARB NHTSA-2018-0067-11873.
                 \2225\ PA DEP NHTSA-2018-0067-11956.
                ---------------------------------------------------------------------------
                 The commenters expressed concerns that this final rule will present
                challenges in fulfilling existing SIP requirements and in attaining or
                maintaining the NAAQS, resulting in the need for emission reductions to
                offset increases due to this rule. This final rulemaking predominantly
                addresses fuel economy and CO2 emissions of the light-duty
                vehicle fleet. It does not affect EPA's Tier 3 vehicle and gasoline
                (Tier 3) standards or California's low emission vehicle III (LEV III)
                emission standards. Tier 3 and LEV III regulations are predominantly
                responsible for regulating criteria pollutant emissions (e.g.
                NOX, VOCs, and carbon monoxide) from light-duty vehicles.
                While this final rulemaking will result in increases in the amount of
                gasoline produced, the number of vehicle re-fueling events and
                emissions of certain criteria pollutants and precursors the emissions
                impact will vary from area to area depending on factors such as the
                composition of the local vehicle fleet and the amount of gasoline
                produced in the area. The agencies expect that states will evaluate any
                adverse emissions or air quality impacts that result from the
                finalization of this rule in the context of state implementation plan
                development for relevant NAAQS, such as the relevant ozone and
                PM2.5 NAAQS.
                 CARB, the joint submission from the States of California and Cities
                of Oakland, and other commenters also stated that the rulemaking
                ``fails to meet the general conformity requirements under the Clean Air
                Act.'' \2226\ Similarly, the Center for Biological Diversity, et al.,
                stated ``it is highly unlikely that the Proposal would not violate
                general conformity.'' \2227\ The states and cities expressed that the
                General Conformity rule applies to this action because ``[f]irst, an
                increase in criteria pollutants is reasonably foreseeable as the
                agencies quantified those emissions as part of this rulemaking. Second,
                the agencies can practically control those emissions as they possess
                ultimate regulatory authority over standards that govern vehicle
                operation.'' \2228\ CARB stated ``NHTSA's determination regarding its
                own conformity obligations . . . does not address conformity-related
                obligations EPA may have that flow from the joint rulemaking.'' \2229\
                SCAQMD similarly stated that ``EPA counts as a federal agency that must
                comply with general conformity requirements. The proposal leaves
                unclear whether EPA also determined its actions comply with the general
                conformity requirements under 40 CFR 93.150 and general conformity SIP
                revisions allowed under 40 CFR 51.851.'' \2230\ SCAQMD concluded that
                EPA must make its own conformity determination, ``and it is not clear
                that EPA can rely on NHTSA's analysis given its dissimilar position in
                having continuing program responsibility over mobile source
                emissions.'' \2231\
                ---------------------------------------------------------------------------
                 \2226\ CARB, NHTSA-2018-0067-11873, Joint Submission from States
                of California and Cities of Oakland, NHTSA-2018-0067-11735.
                 \2227\ Center for Biological Diversity, et al., NHTSA-2018-0067-
                12123.
                 \2228\ Joint Submission from States of California and Cities of
                Oakland, NHTSA-2018-0067-11735.
                 \2229\ CARB, NHTSA-2018-0067-11873.
                 \2230\ SCAQMD, NHTSA_2018-0067-11813.
                 \2231\ SCAQMD, NHTSA_2018-0067-11813.
                ---------------------------------------------------------------------------
                 EPA and NHTSA disagree with the commenters that this rule is
                subject to the CAA section 176(c) conformity requirement and the
                General Conformity regulations. A General Conformity evaluation is
                required for a general Federal action proposed to occur within specific
                nonattainment or maintenance areas. For a General Conformity evaluation
                to be necessary, the action must cause emissions of the criteria and
                precursor pollutants for which the areas are nonattainment or
                maintenance, and the emissions must originate within those areas.
                Further, the evaluation would require a demonstration that the action
                conforms to a specific State Implementation Plan's strategy for air
                pollution prevention and control applicable to the nonattainment and
                maintenance areas. In addition, any mitigation or offsets required to
                demonstrate conformity may require written commitments that must be
                fulfilled, and offsets must occur during the same calendar year as the
                emission increases from the action.
                 While the EPA established the framework of methods and procedures
                that Federal agencies must follow when General Conformity applies to
                their actions, it is the responsibility of each Federal agency to
                prepare its own General Conformity evaluation for actions the agency
                supports, funds, permits or approves. When the EPA functions as a lead
                agency for actions that are subject to General Conformity, such as
                water projects, and the agency may issue permits or approve actions
                that require a General Conformity evaluation, EPA is responsible for
                and sometimes is required to prepare its own General Conformity
                evaluation. For the reasons specified here and in Section X.E.2, a
                General Conformity evaluation is not necessary for either agency.
                 As stated in section 4.1.1.4 of the DEIS and in section 4.1.1.4 of
                the FEIS, the agencies do not believe the proposed rule would result in
                either direct or indirect emissions as defined for General Conformity
                at 40 CFR 93.152 or as required for applicability of the rule under
                section 93.153(b). Furthermore, as described in the proposal, emissions
                from operation of vehicles produced during the model years covered by
                this rule, while reasonably foreseeable, cannot be quantified with any
                certainty in any particular nonattainment or maintenance area. In
                addition, while the emissions rates from MY 2021-2026 vehicles are
                projected for future years in this rule, neither NHTSA nor EPA has
                control over where, when or how many of the vehicles will operate
                during a given future year or within a certain geographical area.
                Therefore, the emissions are not quantifiable. Furthermore, the General
                Conformity
                [[Page 24859]]
                applicability analysis requires an analytical comparison of the
                emissions from MY 2021-2026 vehicles in some specific nonattainment or
                maintenance area in a specific future year, to the emissions projected
                from the operation of vehicles produced in other model years that would
                otherwise operate in that same area in the same future year. Without
                the identity of the future year vehicle fleet by type/make/model (which
                depends on a specific nonattainment or maintenance location and year),
                the net emissions, or total of direct and indirect emissions, cannot be
                quantified. Thus, this rule, in and of itself, is not subject to a
                General Conformity evaluation.
                 CARB stated that this rulemaking would, if finalized, invalidate
                the model underlying California's SIPs (the EMFAC 2014 model), which
                would result in the SIPs being disapproved by EPA.\2232\ CARB expressed
                further concern that as a result of the Clean Air Act's conformity
                requirements, this disapproval would put significant limits on new
                RTPs, TIPS, or regionally significant transportation projects being
                adopted or approved in California.\2233\
                ---------------------------------------------------------------------------
                 \2232\ CARB, NHTSA-2018-0067-11873.
                 \2233\ CARB, NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 The commenter expressed the opinion that if this rule is finalized,
                EPA would disapprove its SIPs because its on-road emission factor model
                (EMFAC) would be invalidated. The commenter also opined that such
                disapprovals would limit the ability of metropolitan planning
                organizations in California to make transportation conformity
                determinations for metropolitan transportation plans, transportation
                improvement programs and certain transportation projects. It is
                premature to assume that EPA will disapprove SIPs because they are
                based on EMFAC2014 or EMFAC2017. EPA will evaluate and address, as
                appropriate, the impact of the SAFE action on future SIP approval
                actions EMFAC2014 and EMFAC2017 remain approved emission factor models
                for SIPs and transportation conformity analyses in California. EPA is
                aware that California released adjustment factors to be applied to
                EMFAC2014 and EMFAC2017 model results to account for impacts of the
                SAFE Part 1 rule for on-road criteria pollutant emissions from light-
                duty vehicles. EPA will work with CARB and DOT on the appropriate
                implementation of federal requirements based on current and available
                information.
                 Because passenger cars and light trucks are subject to gram-per-
                mile emissions standards for criteria pollutants, more fuel-efficient
                (and, correspondingly, less CO2-intensive) vehicles are not,
                from the standpoint of air quality, ``cleaner'' vehicles. Therefore, to
                the extent that CAFE/CO2 standards lead to changes in
                overall quantities of vehicular emissions that impact air quality,
                these are dominated by induced changes in highway travel. Changes in
                overall fuel consumption do lead to changes in emissions from
                ``upstream'' processes involved in supplying fuel to vehicles.
                Depending on how total vehicular emissions and total upstream emissions
                change in response to less stringent standards, overall emissions could
                increase or decrease. While small in magnitude, net impacts could also
                vary considerably among different geographic areas. In other words,
                CAFE and CO2 standards impact fuel consumption and
                CO2 emissions in ways that are direct and unambiguous, and
                impact air quality in ways that are indirect and ambiguous.
                 The following sections, included in prior rules setting fuel
                economy and CO2 standards and updated based on EPA's latest
                scientific assessments, describe the criteria and air toxics considered
                in this rule, and their health and environmental effects. Additionally,
                the section that follows describes how the estimated effects of each
                pollutant were modeled in this rulemaking. Section VII discusses the
                interactions between upstream, tailpipe, and highway travel that result
                in the net emissions of criteria and air toxic pollutants estimated as
                a result of this rule.
                (1) Particulate Matter
                (a) Background
                 Particulate matter (PM) is a complex mixture of solid particles and
                liquid droplets distributed among numerous atmospheric gases which
                interact with solid and liquid phases. Particles range in size from
                those smaller than 1 nanometer (10-\9\ meter) to over 100
                micrometers ([micro]m, or 10-\6\ meter) in diameter (for
                reference, a typical strand of human hair is 50-70 [micro]m in diameter
                and a grain of fine beach sand is about typically 90 [micro]m in
                diameter). Atmospheric particles can be grouped into several classes
                according to their aerodynamic and physical sizes. Generally, the three
                broad classes of particles include ultrafine particles (UFPs, generally
                considered as particulates with a diameter less than or equal to 0.1
                [micro]m [typically based on physical size, thermal diffusivity or
                electrical mobility]), ``fine'' particles (PM2.5; particles
                with a nominal mean aerodynamic diameter less than or equal to 2.5
                [micro]m), and ``thoracic'' particles (PM10; particles with
                a nominal mean aerodynamic diameter less than or equal to 10 [micro]m).
                Particles that fall within the size range between PM2.5 and
                PM10 are referred to as ``thoracic coarse particles''
                (PM10-2.5 particles with a nominal mean aerodynamic diameter
                greater than 2.5 [micro]m and less than or equal to 10 [micro]m). EPA
                currently has standards that regulate PM2.5 and
                PM10.\2234\
                ---------------------------------------------------------------------------
                 \2234\ Regulatory definitions of PM size fractions and
                information on reference and equivalent methods for measuring PM in
                ambient air are provided in 40 CFR parts 50, 53, and 58. With regard
                to national ambient air quality standards (NAAQS) which provide
                protection against health and welfare effects, the 24-hour
                PM10 standard provides protection against effects
                associated with short-term exposure to thoracic coarse particles
                (i.e. PM10--2.5).
                ---------------------------------------------------------------------------
                 Most particles are found in the lower troposphere, where they can
                have residence times ranging from a few hours to weeks. Particles are
                removed from the atmosphere by wet deposition, such as when they are
                carried by rain or snow, or by dry deposition, when particles settle
                out of suspension due to gravity. Atmospheric lifetimes are generally
                longest for PM2.5, which often remains in the atmosphere for
                days to weeks before being removed by wet or dry deposition. \2235\In
                contrast, atmospheric lifetimes for UFP and PM10-2.5 are
                shorter. Within hours, UFP can undergo coagulation and condensation
                that lead to formation of larger particles in the accumulation mode, or
                can be removed from the atmosphere by evaporation, deposition, or
                reactions with other atmospheric components. PM10-2.5 are
                also generally removed from the atmosphere within hours, through wet or
                dry deposition.\2236\
                ---------------------------------------------------------------------------
                 \2235\ U.S. EPA. Integrated Science Assessment (ISA) for
                Particulate Matter (Final Report. 2019), U.S. Environmental
                Protection Agency, Washington DC, EPA/600/R-19/188, 2019. Table 2-1.
                 \2236\ U.S. EPA. Integrated Science Assessment (ISA) for
                Particulate Matter (Final Report, 2019). U.S Environmental
                Protection Agency, Washington, DC, EPA/600/R-19/188, 2019. Table 2-
                1.
                ---------------------------------------------------------------------------
                 Particulate matter consists of both primary and secondary
                particles. Primary particles are emitted directly from sources, such as
                combustion-related activities (e.g., industrial activities, motor
                vehicles, biomass burning), while secondary particles are formed
                through atmospheric chemical reactions of gaseous precursors (e.g.,
                sulfur oxides (SOx), nitrogen oxides (NOx) and volatile organic
                compounds (VOCs) and ammonia). From 2000 to 2017, national annual
                average PM2.5 concentrations have declined by over
                40%,\2237\ largely reflecting reductions in emissions of precursor
                gases.
                ---------------------------------------------------------------------------
                 \2237\ See https://www.epa.gov/air-trends/particulate-matter-pm25-trends and https://www.epa.gov/air-trends/particulate-matter-pm25-trends#pmnat for more information.
                ---------------------------------------------------------------------------
                [[Page 24860]]
                (b) Health Effects of PM
                 Scientific evidence spanning animal toxicological, controlled human
                exposure, and epidemiologic studies shows that exposure to ambient PM
                is associated with a broad range of health effects. The Integrated
                Science Assessment for Particulate Matter (PM ISA) (U.S. EPA 2009)
                synthesizes the toxicological, clinical and epidemiological evidence to
                determine whether each pollutant is causally related to an array of
                adverse human health outcomes associated with either acute (i.e., hours
                or days-long) or chronic (i.e. years-long) exposure; for each outcome,
                the ISA reports this relationship to be causal, likely to be causal,
                suggestive of a causal relationship, inadequate to infer a causal
                relationship or not likely to be a causal relationship.
                 In brief, the ISA for PM2.5 found acute exposure to
                PM2.5 to be causally related to cardiovascular effects and
                mortality (i.e., premature death), and respiratory effects as likely-
                to-be-causally related. The ISA identified cardiovascular effects and
                total mortality as being causally related to long-term exposure to
                PM2.5 and respiratory effects as likely-to-be-causal; and
                the evidence was suggestive of a causal relationship for reproductive
                and developmental effects as well as cancer, mutagenicity and
                genotoxicity. The ISA for ozone found acute exposure to ozone to be
                causally related to respiratory effects, a likely-to-be-causal
                relationship with cardiovascular effects and total mortality and a
                suggestive relationship for central nervous system effects. Among
                chronic effects, the ISA reported a likely-to-be-causal relationship
                for respiratory outcomes and respiratory mortality, and suggestive
                relationship for cardiovascular effects, reproductive and developmental
                effects, central nervous system effects, and total mortality. DOT
                follows EPA's approach of estimating the incidence of air pollution
                effects for those health effects above where the ISA classified as
                either causal or likely-to-be-causal.
                 EPA's more recent Integrated Science Assessment for Particulate
                Matter (PM ISA), which was finalized in December 2019,\2238\ summarizes
                the most recent health effects evidence for short- and long-term
                exposures to PM2.5, PM10-2.5, and ultrafine
                particles, characterizing the strength of the evidence and whether the
                relationship is likely to be causal nature in nature. The 2019 P.M. ISA
                reinforces the findings of the 2009 ISA, and supports the decision to
                continue monetizing the respiratory and cardiovascular health endpoints
                monetized in the current analysis. EPA is currently in the process of
                considering how the 2019 ISA and eventual decision by the Administrator
                regarding the National Ambient Air Quality Standards for particulate
                matter will be used to update forthcoming regulatory impact analysis.
                ---------------------------------------------------------------------------
                 \2238\ U.S. EPA. Integrated Science Assessment (ISA) for
                Particulate Matter (Final Report, 2019). U.S. Environmental
                Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
                ---------------------------------------------------------------------------
                (c) Current Concentrations
                 There are two primary NAAQS for PM2.5: an annual
                standard (12.0 micrograms per cubic meter ([mu]g/m\3\)) set in 2012 and
                a 24-hour standard (35 [mu]g/m\3\) set in 2006, and two secondary NAAQS
                for PM2.5: an annual standard (15.0 [mu]g/m\3\) set in 1997
                and a 24-hour standard (35 [mu]g/m\3\) set in 2006.\2239\
                ---------------------------------------------------------------------------
                 \2239\ The EPA is currently reviewing the PM NAAQS and
                anticipates completing this review in late 2020 Available at https://www.epa.gov/naaqs/particulate-matter-pm-air-quality-standards).
                ---------------------------------------------------------------------------
                 There are many areas of the country that are currently in
                nonattainment for the annual and 24-hour primary PM2.5
                NAAQS. As of January 31, 2020, more than 19 million people lived in the
                4 areas that are designated as nonattainment for the 1997 annual
                PM2.5 NAAQS. These PM2.5 nonattainment areas are
                comprised of 14 full or partial counties. As of January 31, 2020, 6
                areas are designated as nonattainment for the 2012 annual
                PM2.5 NAAQS; these areas are composed of 16 full or partial
                counties with a population of more than 20 million. As of January 31,
                2020, 14 areas are designated as nonattainment for the 2006 24-hour
                PM2.5 NAAQS; these areas are composed of 41 full or partial
                counties with a population of more than 31 million. In total, there are
                currently 17 PM2.5 nonattainment areas with a population of
                more than 32 million people.
                 The EPA has already adopted many mobile source emission control
                programs that are expected to reduce ambient PM concentrations. As a
                result of these and other federal, state and local programs, the number
                of areas that fail to meet the PM2.5 NAAQS in the future is
                expected to decrease. However, even with the implementation of all
                current state and federal regulations, there are projected to be
                counties violating the PM2.5 NAAQS well into the future.
                (2) Ozone
                (a) Background
                 Ground-level ozone pollution is typically formed through reactions
                involving VOC and NOX in the lower atmosphere in the
                presence of sunlight. These pollutants, often referred to as ozone
                precursors, are emitted by many types of sources, such as highway and
                nonroad motor vehicles and engines, power plants, chemical plants,
                refineries, makers of consumer and commercial products, industrial
                facilities, and smaller area sources.
                 The science of ozone formation, transport, and accumulation is
                complex. Ground-level ozone is produced and destroyed in a cyclical set
                of chemical reactions, many of which are sensitive to temperature and
                sunlight. When ambient temperatures and sunlight levels remain high for
                several days and the air is relatively stagnant, ozone and its
                precursors can build up and result in more ozone than typically occurs
                on a single high-temperature day. Ozone and its precursors can be
                transported hundreds of miles downwind from precursor emissions,
                resulting in elevated ozone levels even in areas with low local VOC or
                NOX emissions.
                (b) Health Effects of Ozone
                 This section provides a summary of the health effects associated
                with exposure to ambient concentrations of ozone.\2240\ The information
                in this section is based on the information and conclusions in the
                February 2013 Integrated Science Assessment for Ozone (Ozone ISA),
                which formed the basis for EPA's revision to the primary and secondary
                standards in 2015.\2241\ The Ozone ISA concludes that human exposures
                to ambient concentrations of ozone are associated with a number of
                adverse health effects and characterizes the weight of evidence for
                these health effects.\2242\ The discussion below
                [[Page 24861]]
                highlights the Ozone ISA's conclusions pertaining to health effects
                associated with both short-term and long-term periods of exposure to
                ozone.
                ---------------------------------------------------------------------------
                 \2240\ Human exposure to ozone varies over time due to changes
                in ambient ozone concentration and because people move between
                locations which have notable different ozone concentrations. Also,
                the amount of ozone delivered to the lung is not only influenced by
                the ambient concentrations but also by the individuals breathing
                route and rate.
                 \2241\ U.S. EPA. Integrated Science Assessment of Ozone and
                Related Photochemical Oxidants (Final Report). U.S. Environmental
                Protection Agency, Washington, DC, EPA/600/R-10/076F, 2013. The ISA
                is available at http://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=247492#Download.
                 \2242\ The ISA evaluates evidence and draws conclusions on the
                causal nature of relationship between relevant pollutant exposures
                and health effects, assigning one of five ``weight of evidence''
                determinations: causal relationship, likely to be a causal
                relationship, suggestive of, but not sufficient to infer, a causal
                relationship, inadequate to infer a causal relationship, and not
                likely to be a causal relationship. For more information on these
                levels of evidence, please refer to Table II in the Preamble of the
                ISA.
                ---------------------------------------------------------------------------
                 For short-term exposure to ozone, the Ozone ISA concludes that
                respiratory effects, including lung function decrements, pulmonary
                inflammation, exacerbation of asthma, respiratory-related hospital
                admissions, and mortality, are causally associated with ozone exposure.
                It also concludes that cardiovascular effects, including decreased
                cardiac function and increased vascular disease, and total mortality
                are likely to be causally associated with short-term exposure to ozone
                and that evidence is suggestive of a causal relationship between
                central nervous system effects and short-term exposure to ozone.
                 For long-term exposure to ozone, the Ozone ISA concludes that
                respiratory effects, including new onset asthma, pulmonary inflammation
                and injury, are likely to be causally related with ozone exposure. The
                Ozone ISA characterizes the evidence as suggestive of a causal
                relationship for associations between long-term ozone exposure and
                cardiovascular effects, reproductive and developmental effects, central
                nervous system effects and total mortality. The evidence is inadequate
                to infer a causal relationship between chronic ozone exposure and
                increased risk of lung cancer.
                 Finally, inter-individual variation in human responses to ozone
                exposure can result in some groups being at increased risk for
                detrimental effects in response to exposure. In addition, some groups
                are at increased risk of exposure due to their activities, such as
                outdoor workers or children. The Ozone ISA identified several groups
                that are at increased risk for ozone-related health effects. These
                groups are people with asthma, children and older adults, individuals
                with reduced intake of certain nutrients (i.e., Vitamins C and E),
                outdoor workers, and individuals having certain genetic variants
                related to oxidative metabolism or inflammation. Ozone exposure during
                childhood can have lasting effects through adulthood. Such effects
                include altered function of the respiratory and immune systems.
                Children absorb higher doses (normalized to lung surface area) of
                ambient ozone, compared to adults, due to their increased time spent
                outdoors, higher ventilation rates relative to body size, and a
                tendency to breathe a greater fraction of air through the mouth.
                Children also have a higher asthma prevalence compared to adults.
                (c) Current Concentrations
                 The primary and secondary NAAQS for ozone are 8-hour standards with
                a level of 0.07 ppm. The most recent revision to the ozone standards
                was in 2015; the previous 8-hour ozone primary standard, set in 2008,
                had a level of 0.075 ppm.\2243\ As of January 31, 2020, there were 36
                ozone nonattainment areas for the 2008 ozone NAAQS, composed of 153
                full or partial counties, with a population of more than 99 million. As
                of January 31, 2020, there were 51 ozone nonattainment areas for the
                2015 ozone NAAQS, composed of 206 full or partial countries, with a
                population of more than 122 million. In total, there are currently 59
                ozone nonattainment areas with a population of more than 127 million
                people.
                ---------------------------------------------------------------------------
                 \2243\ The EPA is currently reviewing the PM NAAQS and
                anticipates completing this review in late 2020 Available at
                (https://www.epa.gov/naaqs/ozone-o3-air-quality-standards).
                ---------------------------------------------------------------------------
                 States with ozone nonattainment areas are required to take action
                to bring those areas into attainment. The attainment date assigned to
                an ozone nonattainment area is based on the area's classification. The
                attainment dates for areas designated nonattainment for the 2008 8-hour
                ozone NAAQS are in the 2015 to 2032 timeframe, depending on the
                severity of the problem in each area. Nonattainment area attainment
                dates associated with areas designated for the 2015 NAAQS will be in
                the 2021-2038 timeframe, depending on the severity of the problem in
                each area.
                 EPA has already adopted many emission control programs that are
                expected to reduce ambient ozone levels. As a result of these and other
                federal, state and local programs, 8-hour ozone levels are expected to
                improve in the future. However, even with the implementation of all
                current state and federal regulations, there are projected to be
                counties violating the ozone NAAQS well into the future.
                (3) Nitrogen Oxides
                (a) Background
                 Oxides of nitrogen (NOX) refers to nitric oxide and
                nitrogen dioxide (NO2). For the NOX NAAQS,
                NO2 is the indicator. Most NO2 is formed in the
                air through the oxidation of nitric oxide (NO) emitted when fuel is
                burned at a high temperature. NOX is also a major
                contributor to secondary PM2.5 formation. NOX and
                VOC are the two major precursors of ozone.
                (b) Health Effects of Nitrogen Oxides
                 The most recent review of the health effects of oxides of nitrogen
                completed by EPA can be found in the 2016 Integrated Science Assessment
                for Oxides of Nitrogen--Health Criteria (Oxides of Nitrogen ISA).\2244\
                The primary source of NO2 is motor vehicle emissions, and
                ambient NO2 concentrations tend to be highly correlated with
                other traffic-related pollutants. Thus, a key issue in characterizing
                the causality of NO2-health effect relationships was
                evaluating the extent to which studies supported an effect of
                NO2 that is independent of other traffic-related pollutants.
                EPA concluded that the findings for asthma exacerbation integrated from
                epidemiologic and controlled human exposure studies provided evidence
                that is sufficient to infer a causal relationship between respiratory
                effects and short-term NO2 exposure. The strongest evidence
                supporting an independent effect of NO2 exposure comes from
                controlled human exposure studies demonstrating increased airway
                responsiveness in individuals with asthma following ambient-relevant
                NO2 exposures. The coherence of this evidence with
                epidemiologic findings for asthma hospital admissions and ED visits as
                well as lung function decrements and increased pulmonary inflammation
                in children with asthma describe a plausible pathway by which
                NO2 exposure can cause an asthma exacerbation. The 2016 ISA
                for Oxides of Nitrogen also concluded that there is likely to be a
                causal relationship between long-term NO2 exposure and
                respiratory effects. This conclusion is based on new epidemiologic
                evidence for associations of NO2 with asthma development in
                children combined with biological plausibility from experimental
                studies.
                ---------------------------------------------------------------------------
                 \2244\ U.S. EPA. Integrated Science Assessment for Oxides of
                Nitrogen--Health Criteria (2016 Final Report). U.S. Environmental
                Protection Agency, Washington, DC, EPA/600/R-15/068, 2016.
                ---------------------------------------------------------------------------
                 In evaluating a broader range of health effects, the 2016 ISA for
                Oxides of Nitrogen concluded evidence is ``suggestive of, but not
                sufficient to infer, a causal relationship'' between short-term
                NO2 exposure and cardiovascular effects and mortality and
                between long-term NO2 exposure and cardiovascular effects
                and diabetes, birth outcomes, and cancer. In addition, the scientific
                evidence is inadequate (insufficient consistency of epidemiologic and
                toxicological evidence) to infer a causal relationship for long-term
                NO2 exposure with
                [[Page 24862]]
                fertility, reproduction, and pregnancy, as well as with postnatal
                development. A key uncertainty in understanding the relationship
                between these non-respiratory health effects and short- or long-term
                exposure to NO2 is copollutant confounding, particularly by
                other roadway pollutants. The available evidence for non-respiratory
                health effects does not adequately address whether NO2 has
                an independent effect or whether it primarily represents effects
                related to other or a mixture of traffic-related pollutants.
                 The 2016 ISA for Oxides of Nitrogen concluded that people with
                asthma, children, and older adults are at increased risk for
                NO2-related health effects. In these groups and life stages,
                NO2 is consistently related to larger effects on outcomes
                related to asthma exacerbation, for which there is confidence in the
                relationship with NO2 exposure.
                (c) Current Concentrations
                 On April 6, 2018, based on a review of the full body of scientific
                evidence, EPA issued a decision to retain the current primary NAAQS for
                NO2. The EPA has concluded that the current NAAQS are
                requisite to protect the public health, including the at-risk
                populations of older adults, children and people with asthma, with an
                adequate margin of safety. The primary NAAQS for NO2 are a
                one-hour standard with a level of 100 ppb, based on the three-year
                average of 98th percentile of the annual distribution of daily maximum
                one-hour concentrations, and an annual standard at a level of 53 ppb.
                (4) Sulfur Oxides
                (a) Background
                 Sulfur dioxide (SO2), a member of the sulfur oxide
                (SOX) family of gases, is formed from burning fuels
                containing sulfur (e.g., coal or oil derived), extracting gasoline from
                oil, or extracting metals from ore. SO2 and its gas phase
                oxidation products can dissolve in water droplets and further oxidize
                to form sulfuric acid which reacts with ammonia to form sulfates, which
                are important components of ambient PM.
                (b) Health Effects of SO2
                 This section provides an overview of the health effects associated
                with SO2. Additional information on the health effects of
                SO2 can be found in the 2017 Integrated Science Assessment
                for Sulfur Oxides--Health Criteria (SOX ISA).\2245\
                Following an extensive evaluation of health evidence from animal
                toxicological, controlled human exposure, and epidemiologic studies,
                the EPA has concluded that there is a causal relationship between
                respiratory health effects and short -term exposure to SO2.
                The immediate effect or SO2 on the respiratory system in
                humans is bronchoconstriction. People with asthma are more sensitive to
                the effects of SO2, likely resulting from preexisting
                inflammation associated with this disease. In addition to those with
                asthma (both children and adults), there is suggestive evidence that
                all children and older adults may be at increased risk of
                SO2-related health effects. In free-breathing laboratory
                studies involving controlled human exposures to SO2,
                respiratory effects have consistently been observed following 5-10 min
                exposures at SO2 concentrations >= 400 ppb in people with
                asthma engaged in moderate to heavy levels of exercise, with
                respiratory effects occurring at concentrations as low as
                200 ppb in some individuals with asthma. A clear
                concentration-response relationship has been demonstrated in these
                studies following exposures to SO2 at concentrations between
                200 and 1000 ppb, both in terms of increasing severity of
                respiratory symptoms and decrements in lung function, as well as the
                percentage of individuals with asthma adversely affected. Epidemiologic
                studies have reported positive associations between short-term ambient
                SO2 concentrations and hospital admissions and emergency
                department visits for asthma and for all respiratory causes,
                particularly among children and older adults (>=65 years). The studies
                provide supportive evidence for the causal relationship.
                ---------------------------------------------------------------------------
                 \2245\ U.S. EPA (2017). Integrated Science Assessment (ISA) for
                Sulfur Oxides. Health Criteria (Final Report). EPA 600/R-17/451.
                Washington, DC, U.S. EPA.
                ---------------------------------------------------------------------------
                 For long-term SO2 exposure and respiratory effects, the
                EPA has concluded that the evidence is suggestive or a causal
                relationship. This conclusion is based on new epidemiologic evidence
                for positive associations between long-term SO2 exposure and
                increases in asthma incidence among children, together with animal
                toxicological evidence that provides a pathophysiologic basis for the
                development of asthma. However, uncertainty remains regarding the
                influence of other pollutants on the observed associations with
                SO2 because these epidemiologic studies have not examined
                the potential for copollutant confounding.
                 Consistent associations between short-term exposure to
                SO2 and mortality have been observed in epidemiologic
                studies, with larger effect estimates reported for respiratory
                mortality than for cardiovascular mortality. While this finding is
                consistent with the demonstrated effects of SO2 on
                respiratory morbidity, uncertainty remains with respect to the
                interpretation of these observed mortality associations due to
                potential confounding by various copollutants. Therefore, the EPA has
                concluded that the overall evidence is suggestive of a causal
                relationship between short-term exposure to SO2 and
                mortality.
                (c) Current Concentrations
                 On February 25, 2019, the EPA announced its decision to retain,
                without revision, the existing NAAQS for SOX of 75 ppb, as
                the annual 99th percentile of daily maximum SO2
                concentrations, averaged over three years (84 FR 9866, March 18, 2019).
                The existing primary (health-based) standard provides health protection
                for the at-risk group (people with asthma) against respiratory effects
                following short-term (e.g., 5-minute) exposures to SO2 in
                ambient air. The EPA has been finalizing the initial area designations
                for the 2010 SO2 NAAQS in phases and completed designations
                for most of the country in December 2017. The EPA is under a court
                order to finalize initial designations by December 31, 2020, for a
                remaining set of about 50 areas where states have deployed new
                SO2 monitoring networks. As of January 31, 2020 there are 34
                nonattainment areas for the 2010 SO2 NAAQS. As of January
                31, 2020 there also remain eight nonattainment areas for the primary
                annual SO2 NAAQS set in 1971.
                (5) Carbon Monoxide
                (a) Background
                 Carbon monoxide is a colorless, odorless gas emitted from
                combustion processes. Nationally, particularly in urban areas, the
                majority of CO emissions to ambient air come from mobile sources.\2246\
                ---------------------------------------------------------------------------
                 \2246\ U.S. EPA (2010). Integrated Science Assessment for Carbon
                Monoxide (Final Report). U.S. Environmental Protection Agency,
                Washington, DC, EPA/600/R-09/019F, 2010. Available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=218686. See Section
                2.1.
                ---------------------------------------------------------------------------
                (b) Health Effects of Carbon Monoxide
                 Information on the health effects of CO can be found in the January
                2010 Integrated Science Assessment for Carbon Monoxide (CO ISA)
                associated with the 2010 evaluation of the
                [[Page 24863]]
                NAAQS.\2247\ The CO ISA presents conclusions regarding the presence of
                causal relationships between CO exposure and categories of adverse
                health effects. This section provides a summary of the health effects
                associated with exposure to ambient concentrations of CO, along with
                the ISA conclusions.\2248\
                ---------------------------------------------------------------------------
                 \2247\ U.S. EPA (2010). Integrated Science Assessment for Carbon
                Monoxide (Final Report). U.S. Environmental Protection Agency,
                Washington, DC, EPA/600/R-09/019F, 2010. Available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=218686.
                 \2248\ Personal exposure includes contributions from many
                sources, and in many different environments. Total personal exposure
                to CO includes both ambient and nonambient components; and both
                components may contribute to adverse health effects.
                ---------------------------------------------------------------------------
                 Controlled human exposure studies of subjects with coronary artery
                disease show a decrease in the time to onset of exercise-induced angina
                (chest pain) and electrocardiogram changes following CO exposure. In
                addition, epidemiologic studies observed associations between short-
                term CO exposure and cardiovascular morbidity, particularly increased
                emergency room visits and hospital admissions for coronary heart
                disease (including ischemic heart disease, myocardial infarction, and
                angina). Some epidemiologic evidence is also available for increased
                hospital admissions and emergency room visits for congestive heart
                failure and cardiovascular disease as a whole. The CO ISA concludes
                that a causal relationship is likely to exist between short-term
                exposures to CO and cardiovascular morbidity. It also concludes that
                available data are inadequate to conclude that a causal relationship
                exists between long-term exposures to CO and cardiovascular morbidity.
                 Animal studies show various neurological effects with in-utero CO
                exposure. Controlled human exposure studies report central nervous
                system and behavioral effects following low-level CO exposures,
                although the findings have not been consistent across all studies. The
                CO ISA concludes the evidence is suggestive of a causal relationship
                with both short-and long-term exposure to CO and central nervous system
                effects.
                 A number of studies cited in the CO ISA have evaluated the role of
                CO exposure in birth outcomes such as preterm birth or cardiac birth
                defects. There is limited epidemiologic evidence of a CO-induced effect
                on preterm births and birth defects, with weak evidence for a decrease
                in birth weight. Animal toxicological studies have found perinatal CO
                exposure to affect birth weight, as well as other developmental
                outcomes. The CO ISA concludes the evidence is suggestive of a causal
                relationship between long-term exposures to CO and developmental
                effects and birth outcomes.
                 Epidemiologic studies provide evidence of associations between
                short-term CO concentrations and respiratory morbidity such as changes
                in pulmonary function, respiratory symptoms, and hospital admissions. A
                limited number of epidemiologic studies considered copollutants such as
                ozone, SO2, and PM in two-pollutant models and found that CO
                risk estimates were generally robust, although this limited evidence
                makes it difficult to disentangle effects attributed to CO itself from
                those of the larger complex air pollution mixture. Controlled human
                exposure studies have not extensively evaluated the effect of CO on
                respiratory morbidity. Animal studies at levels of 50-100 ppm CO show
                preliminary evidence of altered pulmonary vascular remodeling and
                oxidative injury. The CO ISA concludes that the evidence is suggestive
                of a causal relationship between short-term CO exposure and respiratory
                morbidity, and inadequate to conclude that a causal relationship exists
                between long-term exposure and respiratory morbidity.
                 Finally, the CO ISA concludes that the epidemiologic evidence is
                suggestive of a causal relationship between short-term concentrations
                of CO and mortality. Epidemiologic evidence suggests an association
                exists between short-term exposure to CO and mortality, but limited
                evidence is available to evaluate cause-specific mortality outcomes
                associated with CO exposure. In addition, the attenuation of CO risk
                estimates which was often observed in copollutant models contributes to
                the uncertainty as to whether CO is acting alone or as an indicator for
                other combustion-related pollutants. The CO ISA also concludes that
                there is not likely to be a causal relationship between relevant long-
                term exposures to CO and mortality.
                (c) Current Concentrations
                 There are two primary NAAQS for CO: an 8-hour standard (9 ppm) and
                a 1-hour standard (35 ppm). The primary NAAQS for CO were retained in
                August 2011. There are currently no CO nonattainment areas; as of
                September 27, 2010, all CO nonattainment areas have been predesignated
                to attainment.
                 The past designations were based on the existing community-wide
                monitoring network. EPA made an addition to the ambient air monitoring
                requirements for CO during the 2011 NAAQS review. Those new
                requirements called for CO monitors to be operated near roads in Core
                Based Statistical Areas (CBSAs) of 1 million or more persons (76 FR
                54294, August 31, 2011).
                (6) Diesel Exhaust
                (a) Background
                 Diesel exhaust consists of a complex mixture composed of
                particulate matter, carbon dioxide, oxygen, nitrogen, water vapor,
                carbon monoxide, nitrogen compounds, sulfur compounds, and numerous
                low-molecular-weight hydrocarbons. A number of these gaseous
                hydrocarbon components are individually known to be toxic, including
                aldehydes, benzene and 1,3-butadiene. The diesel particulate matter
                present in diesel exhaust consists mostly of fine particles (< 2.5
                [micro]m), of which a significant fraction is ultrafine particles (<
                0.1 [micro]m). These particles have a large surface area which makes
                them an excellent medium for adsorbing organics, and their small size
                makes them highly respirable. Many of the organic compounds present in
                the gases and on the particles, such as polycyclic organic matter, are
                individually known to have mutagenic and carcinogenic properties.
                 Diesel exhaust varies significantly in chemical composition and
                particle sizes between different engine types (heavy-duty, light-duty),
                engine operating conditions (idle, acceleration, deceleration), and
                fuel formulations (high/low sulfur fuel). Also, there are emissions
                differences between on-road and nonroad engines because the nonroad
                engines are generally of older technology. After being emitted in the
                engine exhaust, diesel exhaust undergoes dilution as well as chemical
                and physical changes in the atmosphere. The lifetime for some of the
                compounds present in diesel exhaust ranges from hours to days.
                (b) Health Effects of Diesel Exhaust
                 In EPA's 2002 Diesel Health Assessment Document (Diesel HAD),
                exposure to diesel exhaust was classified as likely to be carcinogenic
                to humans by inhalation from environmental exposures, in accordance
                with the revised draft 1996/1999 EPA cancer
                guidelines.2249 2250 A number of
                [[Page 24864]]
                other agencies (National Institute for Occupational Safety and Health,
                the International Agency for Research on Cancer, the World Health
                Organization, California EPA, and the U.S. Department of Health and
                Human Services) had made similar hazard classifications prior to 2002.
                EPA also concluded in the 2002 Diesel HAD that it was not possible to
                calculate a cancer unit risk for diesel exhaust due to limitations in
                the exposure data for the occupational groups or the absence of a dose-
                response relationship.
                ---------------------------------------------------------------------------
                 \2249\ U.S. EPA. (1999). Guidelines for Carcinogen Risk
                Assessment. Review Draft. NCEA-F-0644, July. Washington, DC: U.S.
                EPA. Retrieved on March 19, 2009 from http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=54932.
                 \2250\ U.S. EPA (2002). Health Assessment Document for Diesel
                Engine Exhaust. EPA/600/8-90/057F Office of Research and
                Development, Washington DC. Retrieved on March 17, 2009 from http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=29060. pp. 1-1 & 1-2.
                ---------------------------------------------------------------------------
                 In the absence of a cancer unit risk, the Diesel HAD sought to
                provide additional insight into the significance of the diesel exhaust
                cancer hazard by estimating possible ranges of risk that might be
                present in the population. An exploratory analysis was used to
                characterize a range of possible lung cancer risk. The outcome was that
                environmental risks of cancer from long-term diesel exhaust exposures
                could plausibly range from as low as 10-5 to as high as
                10-3. Because of uncertainties, the analysis acknowledged
                that the risks could be lower than 10-5, and a zero risk
                from diesel exhaust exposure could not be ruled out.
                 Non-cancer health effects of acute and chronic exposure to diesel
                exhaust emissions are also of concern to EPA. EPA derived a diesel
                exhaust reference concentration (RfC) from consideration of four well-
                conducted chronic rat inhalation studies showing adverse pulmonary
                effects. The RfC is 5 [micro]g/m\3\ for diesel exhaust measured as
                diesel particulate matter. This RfC does not consider allergenic
                effects such as those associated with asthma or immunologic or the
                potential for cardiac effects. There was emerging evidence in 2002,
                discussed in the Diesel HAD, that exposure to diesel exhaust can
                exacerbate these effects, but the exposure-response data were lacking
                at that time to derive an RfC based on these then-emerging
                considerations. The EPA Diesel HAD stated, ``With [diesel particulate
                matter] being a ubiquitous component of ambient PM, there is an
                uncertainty about the adequacy of the existing [diesel exhaust]
                noncancer database to identify all of the pertinent [diesel exhaust]-
                caused noncancer health hazards.'' The Diesel HAD also noted ``that
                acute exposure to [diesel exhaust] has been associated with irritation
                of the eye, nose, and throat, respiratory symptoms (cough and phlegm),
                and neurophysiological symptoms such as headache, lightheadedness,
                nausea, vomiting, and numbness or tingling of the extremities.'' The
                Diesel HAD noted that the cancer and noncancer hazard conclusions
                applied to the general use of diesel engines then on the market and as
                cleaner engines replace a substantial number of existing ones, the
                applicability of the conclusions would need to be reevaluated.
                 It is important to note that the Diesel HAD also briefly summarized
                health effects associated with ambient PM and discusses EPA's then-
                annual PM2.5 NAAQS of 15 [micro]g/m\3\. In 2012, EPA revised
                the annual PM2.5 NAAQS to 12 [micro]g/m\3\. There is a large
                and extensive body of human data showing a wide spectrum of adverse
                health effects associated with exposure to ambient PM, of which diesel
                exhaust is an important component. The PM2.5 NAAQS is
                designed to provide protection from the noncancer health effects and
                premature mortality attributed to exposure to PM2.5. The
                contribution of diesel PM to total ambient PM varies in different
                regions of the country and also, within a region, from one area to
                another. The contribution can be high in near-roadway environments, for
                example, or in other locations where diesel engine use is concentrated.
                 Since 2002, several new studies have been published which continue
                to report increased lung cancer risk with occupational exposure to
                diesel exhaust from older engines. Of particular note since 2011 are
                three new epidemiology studies which have examined lung cancer in
                occupational populations, for example, truck drivers, underground
                nonmetal miners and other diesel motor-related occupations. These
                studies reported increased risk of lung cancer with exposure to diesel
                exhaust with evidence of positive exposure-response relationships to
                varying degrees.2251 2252 2253 These newer studies (along
                with others that have appeared in the scientific literature) add to the
                evidence EPA evaluated in the 2002 Diesel HAD and further reinforces
                the concern that diesel exhaust exposure likely poses a lung cancer
                hazard. The findings from these newer studies do not necessarily apply
                to newer technology diesel engines because the newer engines have large
                reductions in the emission constituents compared to older technology
                diesel engines.
                ---------------------------------------------------------------------------
                 \2251\ Garshick, Eric, Francine Laden, Jaime E. Hart, Mary E.
                Davis, Ellen A. Eisen, and Thomas J. Smith. 2012. Lung cancer and
                elemental carbon exposure in trucking industry workers.
                Environmental Health Perspectives 120(9), 1301-06.
                 \2252\ Silverman, D.T., Samanic, C.M., Lubin, J.H., Blair, A.E.,
                Stewart, P.A., Vermeulen, R., & Attfield, M.D. (2012). The diesel
                exhaust in miners study: a nested case-control study of lung cancer
                and diesel exhaust. Journal of the National Cancer Institute.
                 \2253\ Olsson, Ann C., et al. ``Exposure to diesel motor exhaust
                and lung cancer risk in a pooled analysis from case-control studies
                in Europe and Canada.'' American journal of respiratory and critical
                care medicine 183.7 (2011): 941-48.
                ---------------------------------------------------------------------------
                 In light of the growing body of scientific literature evaluating
                the health effects of exposure to diesel exhaust, in June 2012 the
                World Health Organization's International Agency for Research on Cancer
                (IARC), a recognized international authority on the carcinogenic
                potential of chemicals and other agents, evaluated the full range of
                cancer-related health effects data for diesel engine exhaust. IARC
                concluded that diesel exhaust should be regarded as ``carcinogenic to
                humans.'' \2254\ This designation was an update from its 1988
                evaluation that considered the evidence to be indicative of a
                ``probable human carcinogen.''
                ---------------------------------------------------------------------------
                 \2254\ IARC (International Agency for Research on Cancer)
                (2013). Diesel and gasoline engine exhausts and some nitroarenes.
                IARC Monographs Volume 105. Available at http://monographs.iarc.fr/ENG/Monographs/vol105/index.php.
                ---------------------------------------------------------------------------
                (c) Current Concentrations
                 Because DPM is part of overall ambient PM and cannot be easily
                distinguished from overall PM, the agencies do not have direct
                measurements of DPM in the ambient air. DPM concentrations are
                estimated using ambient air quality modeling based on DPM emission
                inventories. DPM emission inventories are computed as the exhaust PM
                emissions from mobile sources combusting diesel or residual oil fuel.
                DPM concentrations were recently estimated as part of the 2014 NATA.
                Areas with high concentrations are clustered in the Northeast, Great
                Lake States, California, and the Gulf Coast States and are also
                distributed throughout the rest of the U.S.
                (7) Air Toxics
                (a) Background
                 Light-duty vehicle emissions contribute to ambient levels of air
                toxics that are known or suspected human or animal carcinogens, or that
                have noncancer health effects. The population experiences an elevated
                risk of cancer and other noncancer health effects from exposure to the
                class of pollutants known collectively as ``air toxics.'' \2255\ These
                compounds include, but are not limited to, benzene, 1,3-
                [[Page 24865]]
                butadiene, formaldehyde, acetaldehyde, acrolein, polycyclic organic
                matter, and naphthalene. These compounds were identified as national or
                regional risk drivers or contributors in the 2014 or past National-
                scale Air Toxics Assessment and have significant inventory
                contributions from mobile sources.2256 2257
                ---------------------------------------------------------------------------
                 \2255\ U.S. EPA (2015). Summary of Results for the 2011
                National-Scale Assessment. http://www3.epa.gov/sites/production/files/2015-12/documents/2011-nata-summary-results.pdf.
                 \2256\ U.S EPA (2018) Technical Support Document EPA's 2014
                National Air Toxics Assessment. Available at https://www.epa.gov/national-air-toxics-assessment/2014-nata-assessment-results.
                 \2257\ U.S. EPA (2015). 2011 National Air Toxics Assessment.
                http://www3.epa.gov/national-air-toxics-assessment/2011-national-air-toxics-assessment.
                ---------------------------------------------------------------------------
                (b) Benzene
                 EPA's Integrated Risk Information System (IRIS) database lists
                benzene as a known human carcinogen (causing leukemia) by all routes of
                exposure, and concludes that exposure is associated with additional
                health effects, including genetic changes in both humans and animals
                and increased proliferation of bone marrow cells in
                mice.2258 2259 2260 EPA states in its IRIS database that
                data indicate a causal relationship between benzene exposure and acute
                lymphocytic leukemia and suggest a relationship between benzene
                exposure and chronic non-lymphocytic leukemia and chronic lymphocytic
                leukemia. EPA's IRIS documentation for benzene also lists a range of
                2.2 x 10-6 to 7.8 x10-6 per [micro]g/m\3\ as the unit risk estimate
                (URE) for benzene.2261 2262 The International Agency for
                Research on Cancer (IARC) has determined that benzene is a human
                carcinogen and the U.S. Department of Health and Human Services (DHHS)
                has characterized benzene as a known human
                carcinogen.2263 2264
                ---------------------------------------------------------------------------
                 \2258\ U.S. EPA. (2000). Integrated Risk Information System File
                for Benzene. This material is available electronically at: http://www3.epa.gov/iris/subst/0276.htm.
                 \2259\ International Agency for Research on Cancer, IARC
                monographs on the evaluation of carcinogenic risk of chemicals to
                humans, Volume 29, some industrial chemicals and dyestuffs,
                International Agency for Research on Cancer, World Health
                Organization, Lyon, France 1982.
                 \2260\ Irons, R.D.; Stillman, W.S.; Colagiovanni, D.B.; Henry,
                V.A. (1992). Synergistic action of the benzene metabolite
                hydroquinone on myelopoietic stimulating activity of granulocyte/
                macrophage colony-stimulating factor in vitro, Proc. Natl. Acad.
                Sci. 89:3691-3695.
                 \2261\ A unit risk estimate is defined as the increase in the
                lifetime risk of an individual who is exposed for a lifetime to 1
                [micro]g/m\3\ benzene in air.
                 \2262\ U.S. EPA (2000). Integrated Risk Information System File
                for Benzene. This material is available electronically at: http://www3.epa.gov/iris/subst/0276.htm.
                 \2263\ International Agency for Research on Cancer (IARC, 2018.
                Monographs on the evaluation of carcinogenic risks to humans, volume
                120. World Health Organization--Lyon France. Available at http://publications.iarc.fr/Book-And-ReportSeries/Iarc-Monographs-On-The-ldentification-Of-Carcinogenic-Hazards-To-Humans/Benzene-2018.
                 \2264\ NTP (National Toxicology Program). 2016. Report on
                Carcinogens, Fourteenth Edition.; Research Triangle Park, NC: U.S.
                Department of Health and Human Services Public Health Service.
                Available at https://ntp.niehs.nih.gov/go/roc.
                ---------------------------------------------------------------------------
                 A number of adverse noncancer health effects including blood
                disorders, such as pre- leukemia and aplastic anemia, have also been
                associated with long-term exposure to benzene. The most sensitive
                noncancer effect observed in humans, based on current data, is the
                depression of the absolute lymphocyte count in blood. EPA's inhalation
                reference concentration (RfC) for benzene is 30 [micro]g/m\3\. The RfC
                is based on suppressed absolute lymphocyte counts seen in humans under
                occupational exposure conditions. In addition, recent work, including
                studies sponsored by the Health Effects Institute, provides evidence
                that biochemical responses are occurring at lower levels of benzene
                exposure than previously known.2265 2266 2267 2268 EPA's
                IRIS program has not yet evaluated these new data. EPA does not
                currently have an acute reference concentration for benzene. The Agency
                for Toxic Substances and Disease Registry (ATSDR) Minimal Risk Level
                (MRL) for acute exposure to benzene is 29 [micro]g/m\3\ for 1-14 days
                exposure.
                ---------------------------------------------------------------------------
                 \2265\ Qu, O.; Shore, R.; Li, G.; Jin, X.; Chen, C.L.; Cohen,
                B.; Melikian, A.; Eastmond, D.; Rappaport, S.; Li, H.; Rupa, D.;
                Suramaya, R.; Songnian, W.; Huifant, Y.; Meng, M.; Winnik, M.; Kwok,
                E.; Li, Y.; Mu, R.; Xu, B.; Zhang, X.; Li, K. (2003). HEI Report
                115, Validation & Evaluation of Biomarkers in Workers Exposed to
                Benzene in China.
                 \2266\ Qu, Q., R. Shore, G. Li, X. Jin, L.C. Chen, B. Cohen, et
                al. (2002). Hematological changes among Chinese workers with a broad
                range of benzene exposures. Am. J. Industr. Med. 42: 275-285.
                 \2267\ Lan, Qing, Zhang, L., Li, G., Vermeulen, R., et al.
                (2004). Hematotoxically in Workers Exposed to Low Levels of Benzene.
                Science 306: 1774-1776.
                 \2268\ Turtletaub, K.W. and Mani, C. (2003). Benzene metabolism
                in rodents at doses relevant to human exposure from Urban Air.
                Research Reports Health Effect Inst. Report No.113.
                ---------------------------------------------------------------------------
                (c) 1,3-Butadiene
                 EPA has characterized 1,3-butadiene as carcinogenic to humans by
                inhalation.2269 2270 The IARC has determined that 1,3-
                butadiene is a human carcinogen and the U.S. DHHS has characterized
                1,3-butadiene as a known human
                carcinogen.2271 2272 2273 2274 There are numerous studies
                consistently demonstrating that 1,3-butadiene is metabolized into
                genotoxic metabolites by experimental animals and humans. The specific
                mechanisms of 1,3-butadiene-induced carcinogenesis are unknown;
                however, the scientific evidence strongly suggests that the
                carcinogenic effects are mediated by genotoxic metabolites. Animal data
                suggest that females may be more sensitive than males for cancer
                effects associated with 1,3-butadiene exposure; there are insufficient
                data in humans from which to draw conclusions about sensitive
                subpopulations. The URE for 1,3-butadiene is 3 x 10-5 per
                [micro]g/m\3\.\2275\ 1,3-butadiene also causes a variety of
                reproductive and developmental effects in mice; no human data on these
                effects are available. The most sensitive effect was ovarian atrophy
                observed in a lifetime bioassay of female mice.\2276\ Based on this
                critical effect and the benchmark concentration methodology, an RfC for
                chronic health effects was calculated at 0.9 ppb (approximately 2
                [micro]g/m\3\).
                ---------------------------------------------------------------------------
                 \2269\ U.S. EPA (2002). Health Assessment of 1,3-Butadiene.
                Office of Research and Development, National Center for
                Environmental Assessment, Washington Office, Washington, DC. Report
                No. EPA600-P-98-001F. This document is available electronically at
                http://www3.epa.gov/iris/supdocs/buta-sup.pdf.
                 \2270\ U.S. EPA (2002). ``Full IRIS Summary for 1,3-butadiene
                (CASRN 106-99-0)'' Environmental Protection Agency, Integrated Risk
                Information System (IRIS), Research and Development, National Center
                for Environmental Assessment, Washington, DC. Available at http://www3.epa.gov/iris/subst/0139.htm.
                 \2271\ International Agency for Research on Cancer (IARC)
                (1999). Monographs on the evaluation of carcinogenic risk of
                chemicals to humans, Volume 71, Re-evaluation of some organic
                chemicals, hydrazine and hydrogen peroxide World Health
                Organization, Lyon, France.
                 \2272\ International Agency for Research on Cancer (IARC).
                (2012). Monographs on the evaluation of carcinogenic risk of
                chemicals to humans, Volume 100F chemical agents and related
                occupations, World Health Organization, Lyon, France.
                 \2273\ International Agency for Research on Cancer (IARC).
                (2008). Monographs on the evaluation of carcinogenic risk of
                chemicals to humans, 1,3-Butadiene, Ethylene Oxide and Vinyl Halides
                (Vinyl Fluoride, Vinyl Chloride and Vinyl Bromide) Volume 97, World
                Health Organization, Lyon, France.
                 \2274\ NTP (National Toxicology Program). 201 6. Report on
                Carcinogens, Fourteenth Edition.; Research Triangle Park NC: U.S.
                Department of Health and Human Services Public Health Service.
                Available at https://ntp.niehs.nih.gov/go/rocl4.
                 \2275\ U.S. EPA (2002). ``Full IRIS Summary for 1,3-butadiene
                (CASRN 106-99-0)'' Environmental Protection Agency, Integrated Risk
                Information System (IRIS), Research and Development, National Center
                for Environmental Assessment, Washington, DC http://www3.epa.gov/iris/subst/0139.htm.
                 \2276\ Bevan, C.; Stadler, J.C.; Elliot, G.S.; et al. (1996).
                Subchronic toxicity of 4-vinylcyclohexene in rats and mice by
                inhalation. Fundam. Appl. Toxicol. 32:1-10.
                ---------------------------------------------------------------------------
                (d) Formaldehyde
                 In 1991, EPA concluded that formaldehyde is a carcinogen based on
                nasal tumors in animal bioassays.\2277\ An Inhalation URE for cancer
                and a Reference Dose for oral noncancer
                [[Page 24866]]
                effects were developed by the agency and posted on the IRIS database.
                Since that time, the National Toxicology Program (NTP) and
                International Agency for Research on Cancer (IARC) have concluded that
                formaldehyde is a known human carcinogen.2278 2279 2280
                ---------------------------------------------------------------------------
                 \2277\ EPA Integrated Risk Information System. Formaldehyde
                (CASRN 50-00-0) http://www3.epa.gov/iris/subst/0419/htm.
                 \2278\ NTP (National Toxicology Program). 2016. Report on
                Carcinogens. Fourteenth Edition.; Research Triangle Park, NC: U.S.
                Department of Health and Human Services. Public Health Service.
                Available at https://ntp.niehs.nih.gov/go/roc 14.
                 \2279\ IARC Monographs on the Evaluation of Carcinogenic Risks
                to Humans Volume 100F (2012): Formaldehyde.
                 \2280\ IARC Monographs on the Evaluation of Carcinogenic Risks
                to Humans Volume 88 (2006): Formaldehyde, 2- Butoxyethanol and 1 -
                tert-Butoxypropan-2-ol.
                ---------------------------------------------------------------------------
                 The conclusions by IARC and NTP reflect the results of
                epidemiologic research published since 1991 in combination with
                previous animal, human and mechanistic evidence. Research conducted by
                the National Cancer Institute reported an increased risk of
                nasopharyngeal cancer and specific lymph hematopoietic malignancies
                among workers exposed to formaldehyde.2281 2282 2283 A
                National Institute of Occupational Safety and Health study of garment
                workers also reported increased risk of death due to leukemia among
                workers exposed to formaldehyde.\2284\ Extended follow-up of a cohort
                of British chemical workers did not report evidence of an increase in
                nasopharyngeal or lymph hematopoietic cancers, but a continuing
                statistically significant excess in lung cancers was reported.\2285\
                Finally, a study of embalmers reported formaldehyde exposures to be
                associated with an increased risk of myeloid leukemia but not brain
                cancer.\2286\
                ---------------------------------------------------------------------------
                 \2281\ Hauptmann, M.; Lubin, J.H.; Stewart, P.A.; Hayes, R.B.;
                Blair, A. 2003. Mortality from lymphohematopoietic malignancies
                among workers in formaldehyde industries. Journal of the National
                Cancer Institute 95, pp. 1615-23.
                 \2282\ Hauptmann, M.; Lubin, J.H.; Stewart, P.A.; Hayes, R.B.;
                Blair, A. 2004. Mortality from solid cancers among workers in
                formaldehyde industries. American Journal of Epidemiology 159: 1117-
                30.
                 \2283\ Beane Freeman, L.E.; Blair, A.; Lubin, J.H.; Stewart,
                P.A.; Hayes, R.B.; Hoover, R.N.; Hauptmann, M. 2009. Mortality from
                lymph hematopoietic malignancies among workers in formaldehyde
                industries: The National Cancer Institute cohort. J. National Cancer
                Inst. 101: 751-61.
                 \2284\ Pinkerton, L.E. 2004. Mortality among a cohort of garment
                workers exposed to formaldehyde: an update. Occup. Environ. Med. 61:
                193-200.
                 \2285\ Coggon, D, EC Harris, J Poole, KT Palmer. 2003. Extended
                follow-up of a cohort of British chemical workers exposed to
                formaldehyde. J National Cancer Inst. 95:1608-15.
                 \2286\ Hauptmann, M.; Stewart P.A.; Lubin J.H.; Beane Freeman,
                L.E.; Hornung, R.W.; Herrick, R.F.; Hoover, R.N.; Fraumeni, J.F.;
                Hayes, R.B. 2009. Mortality from lymph hematopoietic malignancies
                and brain cancer among embalmers exposed to formaldehyde. Journal of
                the National Cancer Institute 101:1696-1708.
                ---------------------------------------------------------------------------
                 Health effects of formaldehyde in addition to cancer were reviewed
                by the Agency for Toxics Substances and Disease Registry in 1999,\2287\
                supplemented in 2010,\2288\ and by the World Health Organization.\2289\
                These organizations reviewed the scientific literature concerning
                health effects linked to formaldehyde exposure to evaluate hazards and
                dose response relationships and defined exposure concentrations for
                minimal risk levels (MRLs). The health endpoints reviewed included
                sensory irritation of eyes and respiratory tract, reduced pulmonary
                function, nasal histopathology, and immune system effects. In addition,
                research on reproductive and developmental effects and neurological
                effects were discussed along with several studies that suggest that
                formaldehyde may increase the risk of asthma--particularly in the
                young.
                ---------------------------------------------------------------------------
                 \2287\ ATSDR (1999). Toxicological Profile for Formaldehyde,
                U.S. Department of Health and Human Services (HHS), July 1999.
                 \2288\ ATSDR (2010). Addendum to the Toxicological Profile for
                Formaldehyde. U.S. Department of Health and Human Services (HHS),
                October 2010.
                 \2289\ IPCS (2002). Concise International Chemical Assessment
                Document 40. Formaldehyde. World Health Organization.
                ---------------------------------------------------------------------------
                 EPA released a draft Toxicological Review of Formaldehyde--
                Inhalation Assessment through the IRIS program for peer review by the
                National Research Council (NRC) and public comment in June 2010.\2290\
                The draft assessment reviewed more recent research from animal and
                human studies on cancer and other health effects. The NRC released
                their review report in April 2011.\2291\ EPA is currently developing a
                revised draft assessment in response to this review.
                ---------------------------------------------------------------------------
                 \2290\ EPA (2010). Toxicological Review of Formaldehyde (CAS No.
                50-00-0)-Inhalation Assessment: In Support of Summary Information on
                the Integrated Risk Information System (IRIS). External Review
                Draft. EPA/635/R-10/002A. U.S. Environmental Protection Agency,
                Washington DC. Available at http://cfpub.epa.gov/ncea/irs_drats/recordisplay.cfm?deid=223614.
                 \2291\ NRC (National Research Council) (2011). Review of the
                Environmental Protection Agency's Draft IRIS Assessment of
                Formaldehyde. Washington DC: National Academies Press. http://books.nap.edu/openbook.php?record_id=13142.
                ---------------------------------------------------------------------------
                (e) Acetaldehyde
                 Acetaldehyde is classified in EPA's IRIS database as a probable
                human carcinogen, based on nasal tumors in rats, and is considered
                toxic by the inhalation, oral, and intravenous routes.\2292\ The URE in
                IRIS for acetaldehyde is 2.2 x 10-6 per [micro]g/m\3\.\2293\
                Acetaldehyde is reasonably anticipated to be a human carcinogen by the
                U.S. DHHS in the 13th Report on Carcinogens and is classified as
                possibly carcinogenic to humans (Group 2B) by the
                IARC.2294 2295 Acetaldehyde is currently listed on the IRIS
                Program Multi-Year Agenda for reassessment within the next few years.
                ---------------------------------------------------------------------------
                 \2292\ U.S. EPA (1991). Integrated Risk Information System File
                of Acetaldehyde. Research and Development, National Center for
                Environmental Assessment, Washington, DC. This material is available
                electronically at http://www3.epa.gov/iris/subst/0290.htm.
                 \2293\ U.S. EPA (1991). Integrated Risk Information System File
                of Acetaldehyde. This material is available electronically at http://www3.epa.gov/iris/subst/0290.htm.
                 \2294\ NTP (National Toxicology Program) 2016. Report on
                Carcinogens Fourteenth Edition, Research Triangle Park, NC: U.S.
                Department of Health and Human Services. Public Health Service.
                Available at https://ntp.niehs.nih.gov/go/roc14.
                 \2295\ International Agency for Research on Cancer (IARC)
                (1999). Re-evaluation of some organic chemicals, hydrazine, and
                hydrogen peroxide. IARC Monographs on the Evaluation of Carcinogenic
                Risk of Chemical to Humans, Vol 71. Lyon, France.
                ---------------------------------------------------------------------------
                 The primary noncancer effects of exposure to acetaldehyde vapors
                include irritation of the eyes, skin, and respiratory tract.\2296\ In
                short-term (4 week) rat studies, degeneration of olfactory epithelium
                was observed at various concentration levels of acetaldehyde
                exposure.2297 2298 Data from these studies were used by EPA
                to develop an inhalation reference concentration of 9 [micro]g/m\3\.
                Some asthmatics have been shown to be a sensitive subpopulation to
                decrements in functional expiratory volume (FEV1 test) and
                bronchoconstriction upon acetaldehyde inhalation.\2299\
                ---------------------------------------------------------------------------
                 \2296\ U.S. EPA (1991). Integrated Risk Information System File
                of Acetaldehyde. This material is available electronically at http://www3.epa.gov/iris/subst/0290.htm.
                 \2297\ U.S. EPA. (2003). Integrated Risk Information System File
                of Acrolein. Research and Development, National Center for
                Environmental Assessment, Washington, DC. This material is available
                electronically at http://www3.epa.gov/iris/subst/0364.htm.
                 \2298\ Appleman, L.M., R.A. Woutersen, and V.J. Feron. (1982).
                Inhalation toxicity of acetaldehyde in rats. I. Acute and subacute
                studies. Toxicology. 23: 293-297.
                 \2299\ Myou, S.; Fujimura, M.; Nishi K.; Ohka, T.; and Matsuda,
                T. (1993) Aerosolized acetaldehyde induces histamine-mediated
                bronchoconstriction in asthmatics. Am. Rev. Respir. Dis. 148(4 Pt
                1): 940-943.
                ---------------------------------------------------------------------------
                (f) Acrolein
                 EPA most recently evaluated the toxicological and health effects
                literature related to acrolein in 2003 and concluded that the human
                carcinogenic potential of acrolein could not be determined because the
                available data were inadequate. No information was available on the
                carcinogenic effects of acrolein in humans and the animal data provided
                inadequate evidence of
                [[Page 24867]]
                carcinogenicity.\2300\ The IARC determined in 1995 that acrolein was
                not classifiable as to its carcinogenicity in humans.\2301\
                ---------------------------------------------------------------------------
                 \2300\ U.S. EPA (2003). Integrated Risk Information System File
                of Acrolein. Research and Development, National Center for
                Environmental Assessment, Washington, DC. This material is available
                at http://www3.epa.gov/iris/subst/0364.htm.
                 \2301\ International Agency for Research on Cancer (1995).
                Monographs on the evaluation of carcinogenic risk of chemicals to
                humans, Volume 63. Dry cleaning, some chlorinated solvents and other
                industrial chemicals, World Health Organization, Lyon, France.
                ---------------------------------------------------------------------------
                 Lesions to the lungs and upper respiratory tract of rats, rabbits,
                and hamsters have been observed after sub-chronic exposure to
                acrolein.\2302\ The agency has developed an RfC for acrolein of 0.02
                [micro]g/m\3\ and an RfD of 0.5 [micro]g/kg-day.\2303\
                ---------------------------------------------------------------------------
                 \2302\ U.S. EPA (2003). Integrated Risk Information System File
                of Acrolein. Office of Research and Development, National Center for
                Environmental Assessment, Washington, DC. This material is available
                at http://www3.epa.gov/iris/subst/0364.htm.
                 \2303\ U.S. EPA (2003). Integrated Risk Information System File
                of Acrolein. Office of Research and Development, National Center for
                Environmental Assessment, Washington, DC. This material is available
                at http://www3.epa.gov/iris/subst/0364.htm.
                ---------------------------------------------------------------------------
                 Acrolein is extremely acrid and irritating to humans when inhaled,
                with acute exposure resulting in upper respiratory tract irritation,
                mucus hypersecretion and congestion. The intense irritancy of this
                carbonyl has been demonstrated during controlled tests in human
                subjects, who suffer intolerable eye and nasal mucosal sensory
                reactions within minutes of exposure.\2304\ These data and additional
                studies regarding acute effects of human exposure to acrolein are
                summarized in EPA's 2003 Toxicological Review of Acrolein.\2305\
                Studies in humans indicate that levels as low as 0.09 ppm (0.21 mg/
                m\3\) for five minutes may elicit subjective complaints of eye
                irritation with increasing concentrations leading to more extensive
                eye, nose and respiratory symptoms. Acute exposures in animal studies
                report bronchial hyper-responsiveness. Based on animal data (more
                pronounced respiratory irritancy in mice with allergic airway disease
                in comparison to non-diseased mice) \2306\ and demonstration of similar
                effects in humans (e.g., reduction in respiratory rate), individuals
                with compromised respiratory function (e.g., emphysema, asthma) are
                expected to be at increased risk of developing adverse responses to
                strong respiratory irritants such as acrolein. EPA does not currently
                have an acute reference concentration for acrolein. The available
                health effect reference values for acrolein have been summarized by EPA
                and include an ATSDR MRL for acute exposure to acrolein of 7 [micro]g/
                m\3\ for 1-14 days' exposure; and Reference Exposure Level (REL) values
                from the California Office of Environmental Health Hazard Assessment
                (OEHHA) for one-hour and 8-hour exposures of 2.5 [micro]g/m\3\ and 0.7
                [micro]g/m\3\, respectively.\2307\
                ---------------------------------------------------------------------------
                 \2304\ U.S. EPA (2003). Toxicological review of acrolein in
                support of summary information on Integrated Risk Information System
                (IRIS) National Center for Environmental Assessment, Washington, DC.
                EPA/635/R-03/003. p. 10. Available online at: http://www3.epa.gov/ncea/iris/toxreviews/0364tr.pdf.
                 \2305\ U.S. EPA (2003). Toxicological review of acrolein in
                support of summary information on Integrated Risk Information System
                (IRIS) National Center for Environmental Assessment, Washington, DC.
                EPA/635/R-03/003. Available online at: http://www3.epa.gov/ncea/iris/toxreviews/0364tr.pdf.
                 \2306\ Morris JB, Symanowicz PT, Olsen JE, et al. (2003).
                Immediate sensory nerve-mediated respiratory responses to irritants
                in healthy and allergic airway-diseased mice. J Appl Physiol
                94(4):1563-71.
                 \2307\ U.S. EPA (2009). Graphical Arrays of Chemical-Specific
                Health Effect Reference Values for Inhalation Exposures (Final
                Report). U.S. Environmental Protection Agency, Washington, DC, EPA/
                600/R-09/061, 2009. Available at http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=211003.
                ---------------------------------------------------------------------------
                (g) Polycyclic Organic Matter
                 The term polycyclic organic matter (POM) defines a broad class of
                compounds that includes the polycyclic aromatic hydrocarbon compounds
                (PAHs). One of these compounds, naphthalene, is discussed separately
                below. POM compounds are formed primarily from combustion and are
                present in the atmosphere in gas and particulate form. Cancer is the
                major concern from exposure to POM. Epidemiologic studies have reported
                an increase in lung cancer in humans exposed to diesel exhaust, coke
                oven emissions, roofing tar emissions, and cigarette smoke; all of
                these mixtures contain POM compounds.2308 2309 Animal
                studies have reported respiratory tract tumors from inhalation exposure
                to benzo[a]pyrene and alimentary tract and liver tumors from oral
                exposure to benzo[a]pyrene.\2310\ In 1997 EPA classified seven PAHs
                (benzo[a]pyrene, benz[a]anthracene, chrysene, benzo[b]fluoranthene,
                benzo[k]fluoranthene, dibenz[a,h]anthracene, and indeno[1,2,3-
                cd]pyrene) as Group B2, probable human carcinogens.\2311\ Since that
                time, studies have found that maternal exposures to PAHs in a
                population of pregnant women were associated with several adverse birth
                outcomes, including low birth weight and reduced length at birth, as
                well as impaired cognitive development in preschool children (3 years
                of age).2312 2313 These and similar studies are being
                evaluated as a part of the ongoing IRIS reassessment of health effects
                associated with exposure to benzo[a]pyrene.
                ---------------------------------------------------------------------------
                 \2308\ Agency for Toxic Substances and Disease Registry (ATSDR).
                (1995). Toxicological profile for Polycyclic Aromatic Hydrocarbons
                (PAHs). Atlanta, GA: U.S. Department of Health and Human Services,
                Public Health Service. Available electronically at http://www.atsdr.cdc.gov/ToxProfiles/TP.asp?id=122&tid=25.
                 \2309\ U.S. EPA (2002). Health Assessment Document for Diesel
                Engine Exhaust. EPA/600/8-90/057F Office of Research and
                Development, Washington DC. http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=29060.
                 \2310\ International Agency for Research on Cancer (IARC).
                (2012). Monographs on the Evaluation of the Carcinogenic Risk of
                Chemicals for Humans, Chemical Agents and Related Occupations. Vol.
                100F. Lyon, France.
                 \2311\ U.S. EPA (1997). Integrated Risk Information System File
                of indeno (1,2,3-cd) pyrene. Research and Development, National
                Center for Environmental Assessment, Washington, DC. This material
                is available electronically at http://www3.epa.gov/ncea/iris/subst/0457.htm.
                 \2312\ Perera, F.P.; Rauh, V.; Tsai, W-Y.; et al. (2002). Effect
                of transplacental exposure to environmental pollutants on birth
                outcomes in a multiethnic population. Environ Health Perspect. 111:
                201-05.
                 \2313\ Perera, F.P.; Rauh, V.; Whyatt, R.M.; Tsai, W.Y.; Tang,
                D.; Diaz, D.; Hoepner, L.; Barr, D.; Tu, Y.H.; Camann, D.; Kinney,
                P. (2006). Effect of prenatal exposure to airborne polycyclic
                aromatic hydrocarbons on neurodevelopment in the first 3 years of
                life among inner-city children. Environ Health Perspect 114: 1287-
                92.
                ---------------------------------------------------------------------------
                (h) Naphthalene
                 Naphthalene is found in small quantities in gasoline and diesel
                fuels. Naphthalene emissions have been measured in larger quantities in
                both gasoline and diesel exhaust compared with evaporative emissions
                from mobile sources, indicating it is primarily a product of
                combustion. Acute (short-term) exposure of humans to naphthalene by
                inhalation, ingestion, or dermal contact is associated with hemolytic
                anemia and damage to the liver and the nervous system.\2314\ Chronic
                (long term) exposure of workers and rodents to naphthalene has been
                reported to cause cataracts and retinal damage.\2315\ The National
                Toxicology
                [[Page 24868]]
                Program listed naphthalene as ``reasonably anticipated to be a human
                carcinogen'' in 2004 on the basis of bioassays reporting clear evidence
                of carcinogenicity in rats and some evidence of carcinogenicity in
                mice.\2316\ California EPA has released a new risk assessment for
                naphthalene, and the IARC has reevaluated naphthalene and re-classified
                it as Group 2B: Possibly carcinogenic to humans.\2317\
                ---------------------------------------------------------------------------
                 \2314\ U.S. EPA (1998). Toxicological Review of Naphthalene
                (Reassessment of the Inhalation Cancer Risk), Environmental
                Protection Agency, Integrated Risk Information System, Research and
                Development, National Center for Environmental Assessment,
                Washington, DC. This material is available electronically at http://www3.epa.gov/iris/subst/0436.htm.
                 \2315\ U.S. EPA (1998). Toxicological Review of Naphthalene
                (Reassessment of the Inhalation Cancer Risk), Environmental
                Protection Agency, Integrated Risk Information System, Research and
                Development, National Center for Environmental Assessment,
                Washington, DC. This material is available electronically at http://www3.epa.gov/iris/subst/0436.htm.
                 \2316\ NTP (National Toxicology Program), 2016. Report on
                Carcinogens Fourteenth Edition, Research Triangle Park NC: U.S.
                Department of Health and Human Services, Public Health Service.
                Available at https://ntp.niehs.nih.gov/go/roc14.
                 \2317\ International Agency for Research on Cancer (IARC).
                (2002). Monographs on the Evaluation of the Carcinogenic Risk of
                Chemicals for Humans. Vol. 82. Lyon, France.
                ---------------------------------------------------------------------------
                 Naphthalene also causes a number of chronic non-cancer effects in
                animals, including abnormal cell changes and growth in respiratory and
                nasal tissues.\2318\ The current EPA IRIS assessment includes noncancer
                data on hyperplasia and metaplasia in nasal tissue that form the basis
                of the inhalation RfC of 3 [micro]g/m\3\.\2319\ The ATSDR MRL for acute
                exposure to naphthalene is 0.6 mg/kg/day.
                ---------------------------------------------------------------------------
                 \2318\ U.S. EPA (1998). Toxicological Review of Naphthalene,
                Environmental Protection Agency, Integrated Risk Information System,
                Research and Development, National Center for Environmental
                Assessment, Washington, DC. This material is available
                electronically at http://www3.epa.gov/iris/subst/0436.htm.
                 \2319\ U.S. EPA (1998). Toxicological Review of Naphthalene.
                Environmental Protection Agency, Integrated Risk Information System
                (IRIS), Research and Development, National Center for Environmental
                Assessment, Washington, DC. Available at http://www3.epa.gov/iris/subst/0436.htm.
                ---------------------------------------------------------------------------
                (i) Other Air Toxics
                 In addition to the compounds described above, other compounds in
                gaseous hydrocarbon and PM emissions from motor vehicles will be
                affected by this action. Mobile source air toxic compounds that will
                potentially be impacted include ethylbenzene, propionaldehyde, toluene,
                and xylene. Information regarding the health effects of these compounds
                can be found in EPA's IRIS database.\2320\
                ---------------------------------------------------------------------------
                 \2320\ U.S. EPA Integrated Risk Information System (IRIS)
                database is available at: www3.epa.gov/iris.
                ---------------------------------------------------------------------------
                (j) Current Concentrations
                 The most recent available data indicate that the majority of
                Americans continue to be exposed to ambient concentrations of air
                toxics at levels which have the potential to cause adverse health
                effects. The levels of air toxics to which people are exposed vary
                depending on where people live and work and the kinds of activities in
                which they engage, as discussed in detail in EPA's most recent Mobile
                Source Air Toxics Rule. According to the National Air Toxic Assessment
                (NATA) for 2014, mobile sources were responsible for 51 percent of
                outdoor anthropogenic toxic emissions and were the largest contributor
                to cancer and noncancer risk from directly emitted pollutants. Mobile
                sources are also significant contributors to precursor emissions which
                react to form air toxics. Formaldehyde is the largest contributor to
                cancer risk of all 71 pollutants quantitatively assessed in the 2014
                NATA. Mobile sources were responsible for more than 30 percent of
                primary anthropogenic emissions of this pollutant in 2014 and also
                contribute to formaldehyde precursor emissions. Benzene is also a large
                contributor to cancer risk, and mobile sources account for
                approximately 54 percent of ambient exposure. Over the years, EPA has
                implemented a number of mobile source and fuel controls which have
                resulted in VOC reductions, which also reduced formaldehyde, benzene
                and other air toxic emissions.
                (k) Exposure and Health Effects Associated With Traffic
                 Locations in close proximity to major roadways generally have
                elevated concentrations of many air pollutants emitted from motor
                vehicles. Hundreds of such studies have been published in peer-reviewed
                journals, concluding that concentrations of CO, NO, NO2,
                benzene, aldehydes, particulate matter, black carbon, and many other
                compounds are elevated in ambient air within approximately 300-600
                meters (approximately 1,000-2,000 feet) of major roadways. Highest
                concentrations of most pollutants emitted directly by motor vehicles
                are found at locations within 50 meters (approximately 165 feet) of the
                edge of a roadway's traffic lanes.
                 A large-scale review of air quality measurements in the vicinity of
                major roadways between 1978 and 2008 concluded that the pollutants with
                the steepest concentration gradients in vicinities of roadways were CO,
                ultrafine particles, metals, elemental carbon (EC), NO, NOX,
                and several VOCs.\2321\ These pollutants showed a large reduction in
                concentrations within 100 meters downwind of the roadway. Pollutants
                that showed more gradual reductions with distance from roadways
                included benzene, NO2, PM2.5, and
                PM10. In the review article, results varied based on the
                method of statistical analysis used to determine the trend.
                ---------------------------------------------------------------------------
                 \2321\ Karner, A.A.; Eisinger, D.S.; Niemeier, D.A. (2010).
                Near-roadway air quality: synthesizing the findings from real-world
                data. Environ Sci. Technol. 44: pp. 5334-44.
                ---------------------------------------------------------------------------
                 For pollutants with relatively high background concentrations
                relative to near-road concentrations, detecting concentration gradients
                can be difficult. For example, many aldehydes have high background
                concentrations as a result of photochemical breakdown of precursors
                from many different organic compounds. This can make detection of
                gradients around roadways and other primary emission sources difficult.
                However, several studies have measured aldehydes in multiple weather
                conditions and found higher concentrations of many carbonyls downwind
                of roadways.2322 2323 These findings suggest a substantial
                roadway source of these carbonyls.
                ---------------------------------------------------------------------------
                 \2322\ Liu, W.; Zhang, J.; Kwon, J.l; et l. (2006).
                Concentrations and source characteristics of airborne carbonyl
                comlbs measured outside urban residences. J Air Waste Manage Assoc.
                56: 1196-1204.
                 \2323\ Cahill, T.M.; Charles, M.J.; Seaman, V.Y. (2010).
                Development and application of a sensitive method to determine
                concentrations of acrolein and other carbonyls in ambient air.
                Health Effects Institute Research Report 149. Available at http://dx.doi.org.
                ---------------------------------------------------------------------------
                 In the past 15 years, many studies have been published with results
                reporting that populations who live, work, or go to school near high-
                traffic roadways experience higher rates of numerous adverse health
                effects, compared to populations far away from major roads.\2324\ In
                addition, numerous studies have found adverse health effects associated
                with spending time in traffic, such as commuting or walking along high-
                traffic roadways.2325 2326 2327 2328 The health outcomes
                with the strongest evidence linking them with traffic-associated air
                pollutants are respiratory effects, particularly in asthmatic children,
                and cardiovascular effects.
                ---------------------------------------------------------------------------
                 \2324\ In the widely-used PubMed database of health
                publications, between January 1, 1990 and August 18, 2011, 605
                publications contained the keywords ``traffic, pollution,
                epidemiology,'' with approximately half the studies published after
                2007.
                 \2325\ Laden, F.; Hart, J.E.; Smith, T.J.; Davis, M.E.;
                Garshick, E. (2007) Cause-specific mortality in the unionized U.S.
                trucking industry. Environmental Health Perspect 115:1192-96.
                 \2326\ Peters, A.; von Klot, S.; Heier, M.; Trentinaglia, I.;
                H[ouml]rmann, A.; Wichmann, H.E.; L[ouml]wel, H. (2004) Exposure to
                traffic and the onset of myocardial infarction. New England J Med
                351: 1721-30.
                 \2327\ Zanobetti, A.; Stone, P.H.; Spelzer, F.E.; Schwartz,
                J.D.; Coull, B.A.; Suh, H.H.; Nearling, B.D.; Mittleman, M.A.;
                Verrier, R.L.; Gold, D.R. (2009) T-wave alternans, air pollution and
                traffic in high-risk subjects. Am J Cardiol 104: 665-670.
                 \2328\ Dubowsky Adar, S.; Adamkiewicz, G.; Gold, D.R.; Schwartz,
                J.; Coull, B.A.; Suh, H. (2007) Ambient and microenvironmental
                particles and exhaled nitric oxide before and after a group bus
                trip. Environ Health Perspect 115: 507-512.
                ---------------------------------------------------------------------------
                [[Page 24869]]
                 Numerous reviews of this body of health literature have been
                published as well. In 2010, an expert panel of the Health Effects
                Institute (HEI) published a review of hundreds of exposure,
                epidemiology, and toxicology studies.\2329\ The panel rated how the
                evidence for each type of health outcome supported a conclusion of a
                causal association with traffic-associated air pollution as either
                ``sufficient,'' ``suggestive but not sufficient,'' or ``inadequate and
                insufficient.'' The panel categorized evidence of a causal association
                for exacerbation of childhood asthma as ``sufficient.'' The panel
                categorized evidence of a causal association for new onset asthma as
                between ``sufficient'' and ``suggestive but not sufficient.''
                ``Suggestive of a causal association'' was how the panel categorized
                evidence linking traffic-associated air pollutants with exacerbation of
                adult respiratory symptoms and lung function decrement. It categorized
                as ``inadequate and insufficient'' evidence of a causal relationship
                between traffic-related air pollution and health care utilization for
                respiratory problems, new onset adult asthma, chronic obstructive
                pulmonary disease (COPD), nonasthmatic respiratory allergy, and cancer
                in adults and children. Other literature reviews have been published
                with conclusions generally similar to the HEI
                panel's.2330 2331 2332 2333 However, in 2014, researchers
                from the U.S. Centers for Disease Control and Prevention (CDC)
                published a systematic review and meta-analysis of studies evaluating
                the risk of childhood leukemia associated with traffic exposure and
                reported positive associations between ``postnatal'' proximity to
                traffic and leukemia risks, but no such association for ``prenatal''
                exposures.\2334\
                ---------------------------------------------------------------------------
                 \2329\ Health Effects Institute Panel on the Health Effects of
                Traffic-Related Air Pollution (2010). Traffic-related air pollution:
                a critical review of the literature on emissions, exposure, and
                health effects. HEI Special Report 17. Available at http://www.healtheffects.org.
                 \2330\ Boothe, V.L.; Shendell, D.G. (2008). Potential health
                effects associated with residential proximity to freeways and
                primary roads: review of scientific literature, 1999-2006. J Environ
                Health 70: 33-41.
                 \2331\ Salam, M.T.; Islam, T.; Gilliland, F.D. (2008). Recent
                evidence for adverse effects of residential proximity to traffic
                sources on asthma. Curr Opin Pulm Med 14: 3-8.
                 \2332\ Sun, X.; Zhang, S.; Ma, X. (2014) No association between
                traffic density and risk of childhood leukemia: a meta-analysis.
                Asia Pac J Cancer Prev 15: 5229-32.
                 \2333\ Raaschou-Nielsen, O.; Reynolds, P. (2006). Air pollution
                and childhood cancer: a review of the epidemiological literature.
                Int J Cancer 118: 2920-9.
                 \2334\ Boothe, VL.; Boehmer, T.K.; Wendel, A.M.; Yip, F.Y.
                (2014) Residential traffic exposure and childhood leukemia: a
                systematic review and meta-analysis. Am J Prev Med 46: 413-422.
                ---------------------------------------------------------------------------
                 Health outcomes with few publications suggest the possibility of
                other effects still lacking sufficient evidence to draw definitive
                conclusions. Among these outcomes with a small number of positive
                studies are neurological impacts (e.g., autism and reduced cognitive
                function) and reproductive outcomes (e.g., preterm birth, low birth
                weight).2335 2336 2337 2338.
                ---------------------------------------------------------------------------
                 \2335\ Volk, H.E.; Hertz-Picciotto, I.; Delwiche, L.; et al.
                (2011). Residential proximity to freeways and autism in the CHARGE
                study. Environ Health Perspect 119: 873-77.
                 \2336\ Franco-Suglia, S.; Gryparis, A.; Wright, R.O.; et al.
                (2007). Association of black carbon with cognition among children in
                a prospective birth cohort study. Am J Epidemiol. doi: 10.1093/aje/
                kwm308. Available at http://dx.doi.org.
                 \2337\ Power, M.C.; Weisskopf, M.G.; Alexeef, SE; et al. (2011).
                Traffic-related air pollution and cognitive function in a cohort of
                older men. Environ Health Perspect 2011: 682-687.
                 \2338\ Wu, J.; Wilhelm, M.; Chung, J.; et al. (2011). Comparing
                exposure assessment methods for traffic-related air pollution in and
                adverse pregnancy outcome study. Environ Res 111: 685-6692.
                ---------------------------------------------------------------------------
                 In addition to health outcomes, particularly cardiopulmonary
                effects, conclusions of numerous studies suggest mechanisms by which
                traffic-related air pollution affects health. Numerous studies indicate
                that near-roadway exposures may increase systemic inflammation,
                affecting organ systems, including blood vessels and
                lungs.2339 2340 2341 2342 Long-term exposures in near-road
                environments have been associated with inflammation-associated
                conditions, such as atherosclerosis and
                asthma.2343 2344 2345
                ---------------------------------------------------------------------------
                 \2339\ Riediker, M. (2007). Cardiovascular effects of fine
                particulate matter components in highway patrol officers. Inhal
                Toxicol 19: 99-105. doi: 10.1080/08958370701495238 Available at
                http://dx.doi.org.
                 \2340\ Alexeef, SE; Coull, B.A.; Gryparis, A.; et al. (2011).
                Medium-term exposure to traffic-related air pollution and markers of
                inflammation and endothelial function. Environ Health Perspect 119:
                481-486. doi:10.1289/ehp.1002560 Available at http://dx.doi.org.
                 \2341\ Eckel. S.P.; Berhane, K.; Salam, M.T.; et al. (2011).
                Traffic-related pollution exposure and exhaled nitric oxide in the
                Children's Health Study. Environ Health Perspect (IN PRESS).
                doi:10.1289/ehp.1103516. Available at http://dx.doi.org.
                 \2342\ Zhang, J.; McCreanor, J.E.; Cullinan, P.; et al. (2009).
                Health effects of real-world exposure diesel exhaust in persons with
                asthma. Res Rep Health Effects Inst 138. Available at http://www.healtheffects.org.
                 \2343\ Adar, S.D.; Klein, R.; Klein, E.K.; et al. (2010). Air
                pollution and the microvasculatory: a cross-sectional assessment of
                in vivo retinal images in the population-based Multi-Ethnic Study of
                Atherosclerosis. PLoS Med 7(11): E1000372. doi:10.1371/
                journal.pmed.1000372. Available at http://dx.doi.org.
                 \2344\ Kan, H.; Heiss, G.; Rose, K.M.; et al. (2008).
                Prospective analysis of traffic exposure as a risk factor for
                incident coronary heart disease: the Atherosclerosis Risk in
                Communities (ARIC) study. Environ Health Perspect 116: 1463-1468.
                doi:10.1289/ehp.11290. Available at http://dx.doi.org.
                 \2345\ McConnell, R.; Islam, T.; Shankardass, K.; et al. (2010).
                Childhood incident asthma and traffic-related air pollution at home
                and school. Environ Health Perspect 1021-26.
                ---------------------------------------------------------------------------
                 Several studies suggest that some factors may increase
                susceptibility to the effects of traffic-associated air pollution.
                Several studies have found stronger respiratory associations in
                children experiencing chronic social stress, such as in violent
                neighborhoods or in homes with high family
                stress.2346 2347 2348
                ---------------------------------------------------------------------------
                 \2346\ Islam, T.; Urban, R.; Gauderman, W.J.; et al. (2011).
                Parental stress increases the detrimental effect of traffic exposure
                on children's lung function. Am J Respir Crit Care Med (In press).
                 \2347\ Clougherty, J.E.; Levy, J.I.; Kubzansky, L.D.; et al.
                (2007). Synergistic effects of traffic-related air pollution and
                exposure to violence on urban asthma etiology. Environ Health
                Perspect 115: 1140-46.
                 \2348\ Chen, E.; Schrier, H.M.; Strunk, R.C.; et al. (2008).
                Chronic traffic-related air pollution and stress interact to predict
                biologic and clinical outcomes in asthma. Environ Health Perspect
                116: 970-5.
                ---------------------------------------------------------------------------
                 The risks associated with residence, workplace, or schools near
                major roads are of potentially high public health significance due to
                the large population in such locations. According to the 2009 American
                Housing Survey, over 22 million homes (17.0 percent of all U.S. housing
                units) were located within 300 feet of an airport, railroad, or highway
                with four or more lanes. This corresponds to a population of more than
                50 million U.S. residents in close proximity to high-traffic roadways
                or other transportation sources. Based on 2010 Census data, a 2013
                publication estimated that 19 percent of the U.S. population (over 59
                million people) lived within 500 meters of roads with at least 25,000
                annual average daily traffic (AADT), while about 3.2 percent of the
                population lived within 100 meters (about 300 feet) of such
                roads.\2349\ Another 2013 study estimated that 3.7 percent of the U.S.
                population (about 11.3 million people) lived within 150 meters (about
                500 feet) of interstate highways or other freeways and
                expressways.\2350\ On average, populations near major roads have higher
                fractions of minority residents and lower socioeconomic status.
                Furthermore, on average, Americans spend more than an hour traveling
                each day, bringing nearly all residents into a
                [[Page 24870]]
                high-exposure microenvironment for part of the day.
                ---------------------------------------------------------------------------
                 \2349\ Rowangould, G.M. (2013). A census of the U.S. near-
                roadway population: public health and environmental justice
                considerations. Transportation Research Part D 25: 59-67.
                 \2350\ Boehmer, T.K.; Foster, S.L.; Henry, J.R.; Woghiren-
                Akinnifesi, E.L.; Yip, F.Y. (2013) Residential proximity to major
                highways--United States, 2010. Morbidity and Mortality Weekly Report
                62(3); 46-50.
                ---------------------------------------------------------------------------
                 In light of these concerns, EPA has required through the NAAQS
                process that air quality monitors be placed near high-traffic roadways
                for determining concentrations of CO, NO2, and
                PM2.5 (in addition to those existing monitors located in
                neighborhoods and other locations farther away from pollution sources).
                Near-roadway monitors for NO2 began operation between 2014
                and 2017 in Core Based Statistical Areas (CBSAs) with population of at
                least 500,000. Monitors for CO and PM2.5 began operation
                between 2015 and 2017. These monitors will further the understanding of
                exposure in these locations.
                 EPA and DOT continue to research near-road air quality, including
                the types of pollutants found in high concentrations near major roads
                and health problems associated with the mixture of pollutants near
                roads.
                (8) Environmental Effects of Non-GHG Pollutants
                (a) Visibility
                 Visibility can be defined as the degree to which the atmosphere is
                transparent to visible light.\2351\ Visibility impairment is caused by
                light scattering and absorption by suspended particles and gases.
                Visibility is important because it has direct significance to people's
                enjoyment of daily activities in all parts of the country. Individuals
                value good visibility for the well-being it provides them directly,
                where they live and work, and in places where they enjoy recreational
                opportunities. Visibility is also highly valued in significant natural
                areas, such as national parks and wilderness areas, and special
                emphasis is given to protecting visibility in these areas. For more
                information on visibility see the final 2019 p.m.
                ISA.2352 2353
                ---------------------------------------------------------------------------
                 \2351\ National Research Council, (1993). Protecting Visibility
                in National Parks and Wilderness Areas. National Academy of Sciences
                Committee on Haze in National Parks and Wilderness Areas. National
                Academy Press, Washington, DC. Available at http://www.nap.edu/books/0309048443/html/.
                 \2352\ U.S. EPA. Integrated Science Assessment (ISA) for
                Particulate Matter (Final Report 2019). U.S. Environmental
                Protection Agency, Washington, DC, EPA/600/R-19/188, 2019.
                 \2353\ There is an ongoing review of the ISA for Oxides of
                Nitrogen Oxides of Sulfur, and Particulate Matter (Ecological
                Criteria), Available at https://wwwepa.gov/isa/integrated-science-assessment-isa-oxides-nitrogen-oxides-sulfur-andparticulate-matter.
                ---------------------------------------------------------------------------
                 EPA is working to address visibility impairment. Reductions in air
                pollution from implementation of various programs associated with the
                Clean Air Act Amendments of 1990 (CAAA) provisions have resulted in
                substantial improvements in visibility and will continue to do so in
                the future. Because trends in haze are closely associated with trends
                in particulate sulfate and nitrate due to the relationship between
                their concentration and light extinction, visibility trends have
                improved as emissions of SO2 and NOX have
                decreased over time due to air pollution regulations such as the Acid
                Rain Program.\2354\
                ---------------------------------------------------------------------------
                 \2354\ U.S. EPA (2009). Final Report: Integrated Science
                Assessment for Particulate Matter. U.S. Environmental Protection
                Agency, Washington, DC, EPA/600/R-08/139F, 2009.
                ---------------------------------------------------------------------------
                 In the Clean Air Act Amendments of 1977, Congress recognized
                visibility's value to society by establishing a national goal to
                protect national parks and wilderness areas from visibility impairment
                caused by manmade pollution.\2355\ In 1999, EPA finalized the regional
                haze program to protect the visibility in Mandatory Class I Federal
                areas.\2356\ There are 156 national parks, forests and wilderness areas
                categorized as Mandatory Class I Federal areas.\2357\ These areas are
                defined in CAA Section 162 as those national parks exceeding 6,000
                acres, wilderness areas and memorial parks exceeding 5,000 acres, and
                all international parks which were in existence on August 7, 1977.
                ---------------------------------------------------------------------------
                 \2355\ See Section 169(a) of the Clean Air Act.
                 \2356\ 64 FR 35714 (July 1, 1999).
                 \2357\ 62 FR 38680-81 (July 18, 1997).
                ---------------------------------------------------------------------------
                 EPA has also concluded that PM2.5 causes adverse effects
                on visibility in other areas that are not targeted by the Regional Haze
                Rule, such as urban areas, depending on PM2.5 concentrations
                and other factors such as dry chemical composition and relative
                humidity (i.e., an indicator of the water composition of the
                particles). EPA revised the PM2.5 standards in December 2012
                and established a target level of protection that is expected to be met
                through attainment of the existing secondary standards for
                PM2.5.
                (b) Plant and Ecosystem Effects of Ozone
                 The welfare effects of ozone include effects on ecosystems, which
                can be observed across a variety of scales, i.e. subcellular, cellular,
                leaf, whole plant, population and ecosystem. Ozone can produce both
                acute and chronic injury in sensitive species depending on the
                concentration level and the duration of the exposure.\2358\ In those
                sensitive species,\2359\ effects from repeated exposure to ozone
                throughout the growing season of the plant can tend to accumulate, so
                that even relatively low concentrations experienced for a longer
                duration have the potential to create chronic stress on
                vegetation.\2360\ Ozone damage to sensitive species includes impaired
                photosynthesis and visible injury to leaves. The impairment of
                photosynthesis, the process by which the plant makes carbohydrates (its
                source of energy and food), can lead to reduced crop yields, timber
                production, and plant productivity and growth. Impaired photosynthesis
                can also lead to a reduction in root growth and carbohydrate storage
                below ground, resulting in other, more subtle plant and ecosystems
                impacts.\2361\ These latter impacts include increased susceptibility of
                plants to insect attack, disease, harsh weather, interspecies
                competition and overall decreased plant vigor. The adverse effects of
                ozone on areas with sensitive species could potentially lead to species
                shifts and loss from the affected ecosystems,\2362\ resulting in a loss
                or reduction in associated ecosystem goods and services. Additionally,
                visible ozone injury to leaves can result in a loss of aesthetic value
                in areas of special scenic significance like national parks and
                wilderness areas and reduced use of sensitive ornamentals in
                landscaping.\2363\
                ---------------------------------------------------------------------------
                 \2358\ 73 FR 16486 (March 27, 2008).
                 \2359\ 73 FR 16491 (March 27, 2008). Only a small percentage of
                all the plant species growing within the U.S. (over 43,000 species
                have been catalogued in the USDA PLANTS database) have been studied
                with respect to ozone sensitivity.
                 \2360\ The concentration at which ozone levels overwhelm a
                plant's ability to detoxify or compensate for oxidant exposure
                varies. Thus, whether a plant is classified as sensitive or tolerant
                depends in part on the exposure levels being considered. Chapter 9,
                Section 9.3.4 of U.S. EPA, 2013 Integrated Science Assessment for
                Ozone and Related Photochemical Oxidants. Office of Research and
                Development/National Center for Environmental Assessment. U.S.
                Environmental Protection Agency. EPA 600/R-10/076F.
                 \2361\ 73 FR 16492 (March 27, 2008).
                 \2362\ 73 FR 16493-94 (March 27, 2008). Ozone impacts could be
                occurring in areas where plant species sensitive to ozone have not
                yet been studied or identified.
                 \2363\ 73 FR 16490-97 (March 27, 2008).
                ---------------------------------------------------------------------------
                 The most recent Integrated Science Assessment (ISA) for Ozone
                presents more detailed information on how ozone affects vegetation and
                ecosystems.2364 2365 The ISA concludes that ambient
                concentrations of ozone are associated with a number of adverse welfare
                effects and characterizes the
                [[Page 24871]]
                weight of evidence for different effects associated with ozone.\2366\
                The ISA concludes that visible foliar injury effects on some
                vegetation, reduced vegetation growth, reduced productivity in
                terrestrial ecosystems, reduced yield and quality of some agricultural
                crops, and alteration of below-ground biogeochemical cycles are
                causally associated with exposure to ozone. It also concludes that
                reduced carbon sequestration in terrestrial ecosystems, alteration of
                terrestrial ecosystem water cycling, and alteration of terrestrial
                community composition are likely to be causally associated with
                exposure to ozone.
                ---------------------------------------------------------------------------
                 \2364\ U.S. EPA. Integrated Science Assessment of Ozone and
                Related Photochemical Oxidants (Final Report). U.S. Environmental
                Protection Agency, Washington, DC, EPA/600/R-10/076F, 2013. The ISA
                is available at http://cfpub.epa.gov/ncea/isa/recordisplay.cfm?deid=247492#Download.
                 \2365\ There is an ongoing review of the ozone NAAQS, EPA
                intends to finalize an updated Integrated Science Assessment in
                early 2020 Available at (https://www.epa.gov naaqs/ozone-o3-
                standards-integrated-science-assessments-currentreview).
                 \2366\ The Ozone ISA evaluates the evidence associated with
                different ozone related health and welfare effects, assigning one of
                five ``weight of evidence'' determinations: causal relationship,
                likely to be a causal relationship, suggestive of a causal
                relationship, inadequate to infer a causal relationship, and not
                likely to be a causal relationship. For more information on these
                levels of evidence, please refer to Table II of the ISA.
                ---------------------------------------------------------------------------
                (c) Atmospheric Deposition
                 Wet and dry deposition of ambient particulate matter delivers a
                complex mixture of metals (e.g., mercury, zinc, lead, nickel, aluminum,
                and cadmium), organic compounds (e.g., polycyclic organic matter,
                dioxins, and furans) and inorganic compounds (e.g., nitrate, sulfate)
                to terrestrial and aquatic ecosystems. The chemical form of the
                compounds deposited depends on a variety of factors including ambient
                conditions (e.g., temperature, humidity, oxidant levels) and the
                sources of the material. Chemical and physical transformations of the
                compounds occur in the atmosphere as well as the media onto which they
                deposit. These transformations in turn influence the fate,
                bioavailability and potential toxicity of these compounds.
                 Adverse impacts to human health and the environment can occur when
                particulate matter is deposited to soils, water, and biota.\2367\
                Deposition of heavy metals or other toxics may lead to the human
                ingestion of contaminated fish, impairment of drinking water, damage to
                terrestrial, freshwater and marine ecosystem components, and limits to
                recreational uses. Atmospheric deposition has been identified as a key
                component of the environmental and human health hazard posed by several
                pollutants including mercury, dioxin and PCBs.\2368\
                ---------------------------------------------------------------------------
                 \2367\ U.S. EPA. Integrated Science Assessment for Particulate
                Matter (Final Report). U.S. Environmental Protection Agency,
                Washington, DC, EPA/600/R-08/139F, 2009.
                 \2368\ U.S. EPA (2000). Deposition of Air Pollutants to the
                Great Waters: Third Report to Congress. Office of Air Quality
                Planning and Standards. EPA-453/R-00-0005.
                ---------------------------------------------------------------------------
                 The ecological effects of acidifying deposition and nutrient
                enrichment are detailed in the Integrated Science Assessment for Oxides
                of Nitrogen and Sulfur-Ecological Criteria.2369 2370
                Atmospheric deposition of nitrogen and sulfur contributes to
                acidification, altering biogeochemistry and affecting animal and plant
                life in terrestrial and aquatic ecosystems across the United States.
                The sensitivity of terrestrial and aquatic ecosystems to acidification
                from nitrogen and sulfur deposition is predominantly governed by
                geology. Prolonged exposure to excess nitrogen and sulfur deposition in
                sensitive areas acidifies lakes, rivers and soils. Increased acidity in
                surface waters creates inhospitable conditions for biota and affects
                the abundance and biodiversity of fishes, zooplankton and
                macroinvertebrates and ecosystem function. Over time, acidifying
                deposition also removes essential nutrients from forest soils,
                depleting the capacity of soils to neutralize future acid loadings and
                negatively affecting forest sustainability. Major effects in forests
                include a decline in sensitive tree species, such as red spruce (Picea
                rubens) and sugar maple (Acer saccharum). In addition to the role
                nitrogen deposition plays in acidification, nitrogen deposition also
                leads to nutrient enrichment and altered biogeochemical cycling. In
                aquatic systems increased nitrogen can alter species assemblages and
                cause eutrophication. In terrestrial systems nitrogen loading can lead
                to loss of nitrogen-sensitive lichen species, decreased biodiversity of
                grasslands, meadows and other sensitive habitats, and increased
                potential for invasive species.
                ---------------------------------------------------------------------------
                 \2369\ NOX and SOX secondary ISA2369 U.S.
                EPA. Integrated Science Assessment (ISA) for Oxides of Nitrogen and
                Sulfur Ecological Criteria (Final Report). U.S. Environmental
                Protection Agency, Washington, DC, EPA/600/R-08/082F, 2008.
                 \2370\ There is an ongoing review of the ISA for Oxides and
                Nitrogen, Oxides of Sulfur, and Particulate Matter (Ecological
                Criteria), Available at https://www.epa.gov/isa/integrated-science-assessment-isa-oxides-nitrogen-oxides-sulfur-and-particulate-matter.
                ---------------------------------------------------------------------------
                 Building materials including metals, stones, cements, and paints
                undergo natural weathering processes from exposure to environmental
                elements (e.g., wind, moisture, temperature fluctuations, sunlight,
                etc.). Pollution can worsen and accelerate these effects. Deposition of
                PM is associated with both physical damage (materials damage effects)
                and impaired aesthetic qualities (soiling effects). Wet and dry
                deposition of PM can physically affect materials, adding to the effects
                of natural weathering processes, by potentially promoting or
                accelerating the corrosion of metals, by degrading paints and by
                deteriorating building materials such as stone, concrete and
                marble.\2371\ The effects of PM are exacerbated by the presence of
                acidic gases and can be additive or synergistic due to the complex
                mixture of pollutants in the air and surface characteristics of the
                material. Acidic deposition has been shown to have an effect on
                materials including zinc/galvanized steel and other metal, carbonate
                stone (as monuments and building facings), and surface coatings
                (paints).\2372\ The effects on historic buildings and outdoor works of
                art are of particular concern because of the uniqueness and
                irreplaceability of many of these objects. In addition to aesthetic and
                functional effects on metals, stone and glass, altered energy
                efficiency of photovoltaic panels by PM deposition is also becoming an
                important consideration for impacts of air pollutants on materials.
                ---------------------------------------------------------------------------
                 \2371\ U.S. EPA. Integrated Science Assessment (ISA) for
                Particulate Matter (Final Report, 2019). U.S Environmental
                Protection Agency, Washington, DC, EPA/600/R-l9/188, 2019.
                 \2372\ Irving, P.M., e.d. 1991. Acid Deposition: State of
                Science and Technology, Volume III, Terrestrial, Materials, Health,
                and Visibility Effects, The U.S. National Acid Precipitation
                Assessment Program, Chapter 24, pp. 24-76.
                ---------------------------------------------------------------------------
                (d) Environmental Effects of Air Toxics
                 Emissions from producing, transporting and combusting fuel
                contribute to ambient levels of pollutants that contribute to adverse
                effects on vegetation. Volatile organic compounds, some of which are
                considered air toxics, have long been suspected to play a role in
                vegetation damage.\2373\ In laboratory experiments, a wide range of
                tolerance to VOCs has been observed.\2374\ Decreases in harvested seed
                pod weight have been reported for the more sensitive plants, and some
                studies have reported effects on seed germination, flowering and fruit
                ripening. Effects of individual VOCs or their role in conjunction with
                other stressors (e.g., acidification, drought, temperature extremes)
                have not been well studied. In a recent study of a mixture of VOCs
                including ethanol and toluene on herbaceous plants, significant effects
                on seed production, leaf water content and photosynthetic
                [[Page 24872]]
                efficiency were reported for some plant species.\2375\
                ---------------------------------------------------------------------------
                 \2373\ U.S. EPA (1991). Effects of organic chemicals in the
                atmosphere on terrestrial plants. EPA/600/3-91/001.
                 \2374\ Cape JN, ID Leith, J Binnie, J Content, M Donkin, M
                Skewes, DN Price AR Brown, AD Sharpe. (2003). Effects of VOCs on
                herbaceous plants in an open-top chamber experiment. Environ.
                Pollut. 124:341-343.
                 \2375\ Cape JN, ID Leith, J Binnie, J Content, M Donkin, M
                Skewes, DN Price AR Brown, AD Sharpe. (2003). Effects of VOCs on
                herbaceous plants in an open-top chamber experiment. Environ.
                Pollut. 124:341-343.
                ---------------------------------------------------------------------------
                 Research suggests an adverse impact of vehicle exhaust on plants,
                which has in some cases been attributed to aromatic compounds and in
                other cases to nitrogen oxides.2376 2377 2378 The impacts of
                VOCs on plant reproduction may have long-term implications for
                biodiversity and survival of native species near major roadways. Most
                of the studies of the impacts of VOCs on vegetation have focused on
                short-term exposure and few studies have focused on long-term effects
                of VOCs on vegetation and the potential for metabolites of these
                compounds to affect herbivores or insects.
                ---------------------------------------------------------------------------
                 \2376\ Viskari E.-L. (2000). Epicuticular wax of Norway spruce
                needles as indicator of traffic pollutant deposition. Water, Air,
                and Soil Pollut. 121:327-337.
                 \2377\ Ugrekhelidze D, F Korte, G Kvesitadze (1997). Uptake and
                transformation of benzene and toluene by plant leaves. Ecotox.
                Environ. Safety 37:24-29.
                 \2378\ Kammerbauer H, H Selinger, R Rommelt, A Ziegler-Jons, D
                Knoppik, B Hock. (1987). Toxic components of motor vehicle emissions
                for the spruce Picea abies. Environ. Pollut. 48: 235-43.
                ---------------------------------------------------------------------------
                (c) How the Agencies Estimated Impacts on Emissions
                 The rule implements an emissions inventory methodology for
                estimating impacts. Vehicle emissions inventories are often described
                as three-legged stools, comprised of activity (i.e., miles traveled,
                hours operated, or gallons of gasoline burned), population (or number
                of vehicles), and emission factors. An emissions factor is a
                representative value that attempts to relate the quantity of a
                pollutant released to the atmosphere with an activity associated with
                the release of that pollutant.\2379\ Depending on the vehicle activity
                available, emission factors may be on a distance-, time-, or fuel-
                basis. For example, an emissions inventory for a light-duty fleet could
                simply be the vehicle miles traveled multiplied by the appropriate per-
                mile emission factor for a chosen pollutant.
                ---------------------------------------------------------------------------
                 \2379\ USEPA, Basics Information of Air Emissions Factors and
                Quantification, https://www.epa.gov/air-emissions-factors-and-quantification/basic-information-air-emissions-factors-and-quantification.
                ---------------------------------------------------------------------------
                 As described in Section VI.A, Overview of Methods, the agencies
                used specific models to develop inputs to the CAFE model, such as fuel
                prices and emission factors. The CAFE model estimates how manufacturers
                might respond to a given regulatory scenario (CAFE/CO2
                standards) and fuel prices, and what impact that response will have on
                emissions. As mentioned above, the agencies have used DOT's CAFE model
                to estimate impacts of the CAFE and CO2 standards
                promulgated today. Details of the analysis are presented below and in
                the accompanying RIA, EIS, and model documentation. To estimate the
                response on emissions, several steps are involved. The estimation of
                emissions involves accounting for vehicular fuel type (e.g., gasoline,
                diesel, electric) and fuel economy (accounting for the estimated gap,
                discussed below, between ``laboratory'' and actual on-road fuel
                economy), vehicular turnover and travel demand, fuel properties (carbon
                content), and upstream process emissions. Like other models, the CAFE
                model includes procedures to estimate annual rates at which new
                vehicles are used and subsequently scrapped. Together, these procedures
                result in, for each vehicle model in each model year, estimates of the
                number remaining in service in each calendar year, as well as the
                annual mileage accumulation (i.e. VMT) in each calendar year.
                Quantities of emissions derive from this vehicle operation.
                 For every vehicle model in the market file, the model estimates the
                VMT per vehicle (using the assumed VMT schedule, the vehicle fuel
                economy, fuel price, and the rebound assumption). Those miles are
                multiplied by the number off each vehicle model/configuration remaining
                in service in any given calendar year. Fuel consumption is the product
                of miles driven and fuel economy, which can be tracked by model year
                cohort in the model. Carbon dioxide emissions from vehicle tailpipes
                are the simple product of gallons consumed and the carbon content of
                each gallon. As discussed in the CAFE model overview, the simulated
                application of technology results in estimates of the cost, fuel type,
                fuel economy, and fuel share applicable to each vehicle model in each
                model year. Together with quantities of travel, and with estimates of
                the ``gap'' between ``laboratory'' and ``on-road'' fuel economy, these
                enable calculation of quantities of fuel consumed in each year during
                the useful life of each vehicle model produced in each model year. The
                model calculates emissions of CO2, CH4, and
                N2O, criteria pollutants, and air toxics, reporting
                emissions both from vehicle tailpipes and from upstream processes
                (e.g., petroleum refining) involving in producing and supplying fuels.
                 In order to calculate calendar year fuel consumption, the model
                needs to account for the inherited on-road fleet in addition to the
                model year cohorts affected by this rule. Using the VMT of the average
                passenger car and light truck from each cohort, the model computes the
                fuel consumption of each model year class of vehicles for its age in a
                given CY. The sum across all ages (and thus, model year cohorts) in a
                given CY provides estimated CY fuel consumption.
                 For this rule, vehicle tailpipe (downstream) and upstream emission
                inventories were developed separately. In addition to the tailpipe
                emissions of carbon dioxide, each gallon of gasoline produced for
                consumption by the on-road fleet has associated ``upstream'' emissions
                that occur in the extraction, transportation, refining, and
                distribution of the fuel. The tailpipe inventories apply per-mile
                emission factors from the Motor Vehicle Emission Simulator (MOVES) and
                the upstream inventories apply per-gallon of fuel consumed emission
                factors from the Argonne National Laboratory's Greenhouse gases,
                Regulated Emissions, and Energy use in Transportation (GREET) Model.
                The model accounts for upstream emissions and reports them accordingly.
                More detailed descriptions of emission data sources and calculations
                are provided in the following section.
                 The agencies received several comments on estimation of criteria
                pollutant impacts in the NPRM. As discussed elsewhere in this preamble,
                EDF modified aspects of the CAFE model as part of their comments to the
                agencies. Specifically in regards to criteria pollutant emissions, EDF
                made several alternative assumptions, including assertions that
                criteria pollutant impacts were not as negligible as the agencies
                claimed, and that fatalities due to criteria pollutant emissions would
                be higher than the agencies showed in the NPRM. The agencies declined
                to adopt EDF's suggested changes to the model and inputs, but did make
                the changes discussed in this section that refined the agencies'
                accounting of criteria pollutant emissions and explicitly modeled
                criteria pollutant fatalities, as discussed below.
                 Also discussed elsewhere in this preamble, some commenters
                expressed that the agencies' analysis (by implication, their modeling)
                should account for some States' mandates that manufacturers sell
                minimum quantities of ``Zero Emission Vehicles'' (ZEVs).\2380\ These
                commenters stressed the
                [[Page 24873]]
                importance of the ZEV mandate in relation to maintaining air quality
                requirements and reducing effects of climate change.
                ---------------------------------------------------------------------------
                 \2380\ CBD et al., NHTSA-2018-0067-12123; States and Cities,
                NHTSA-2018-0067-11735; SCAQMD, NHTSA-2018-0067-11813.
                ---------------------------------------------------------------------------
                 The reference case analysis for today's rule, like that for the
                proposal, does not simulate compliance with ZEV mandates,\2381\ because
                such mandates are subject to preemption under EPCA and are therefore
                not enforceable. As discussed in the One National Program Action,
                California and other states remain free to revise their overall average
                emissions standards to further reduce ozone forming emissions and seek
                a waiver of Clean Air Act preemption from EPA, as described above,
                while not violating NHTSA's preemption authority. These States and
                local governments would continue to be allowed to take other actions so
                long as those are not related to fuel economy and are consistent with
                any other relevant Federal law.
                ---------------------------------------------------------------------------
                 \2381\ The NPRM version of the model included experimental
                capabilities to account for mandates and credits for the sale of
                ZEVs, but the agencies did not utilize those capabilities for the
                NPRM for the same reasons discussed above.
                ---------------------------------------------------------------------------
                (1) Activity Levels
                 As discussed in Section VI.A, for each vehicle model/configuration
                in each model year during 2017-2050, the CAFE model estimates and
                records the fuel type (e.g., gasoline, electricity), fuel economy, and
                number of units sold in the U.S. The model also makes use of an
                aggregated representation of vehicles sold in the U.S. during 1978-
                2016. The model estimates the numbers of each cohort of vehicles
                remaining in service in each calendar year, and the amount of driving
                accumulated by each such cohort in each calendar year. The CAFE model
                estimates annual vehicle-miles of travel (VMT) for each individual car
                and light truck model produced in each model year at each age of their
                lifetimes, which extend for a maximum of 40 years. Since a vehicle's
                age is equal to the current calendar year minus the model year in which
                it was originally produced, the age span of each vehicle model's
                lifetime corresponds to a sequence of 40 calendar years beginning in
                the calendar year corresponding to the model year it was
                produced.\2382\ These estimates reflect the gradual decline in the
                fraction of each car and light truck model's original model year
                production volume that is expected to remain in service during each
                year of its lifetime, as well as the well-documented decline in their
                typical use as they age. Using this relationship, the CAFE model
                calculates total VMT for the entire fleet of cars and light trucks in
                service during each calendar year spanned by the agencies' analysis.
                ---------------------------------------------------------------------------
                 \2382\ In practice, many vehicle models bearing a given model
                year designation become available for sale in the preceding calendar
                year, and their sales can extend through the following calendar year
                as well. However, the CAFE model does not attempt to distinguish
                between model years and calendar years; vehicles bearing a model
                year designation are assumed to be produced and sold in that same
                calendar year.
                ---------------------------------------------------------------------------
                 Based on these estimates, the model also calculates quantities of
                each type of fuel or energy, including gasoline, diesel, and
                electricity, consumed in each calendar year. By combining these with
                estimates of each model's fuel or energy efficiency, the model also
                estimates the quantity and energy content of each type of fuel consumed
                by cars and light trucks at each age, or viewed another way, during
                each calendar year of their lifetimes. As with the accounting of VMT,
                these estimates of annual fuel or energy consumption for each vehicle
                model and model year combination are combined to calculate the total
                volume of each type of fuel or energy consumed during each calendar
                year, as well as its aggregate energy content.
                 The procedures the CAFE model uses to estimate annual VMT for
                individual car and light truck models produced during each model year
                over their lifetimes and to combine these into estimates of annual
                fleet-wide travel during each future calendar year, together with the
                sources of its estimates of their survival rates and average use at
                each age, are described in detail in Section VI.D.1 of this final rule.
                The data and procedures it employs to convert these estimates of VMT to
                fuel and energy consumption by individual model, and to aggregate the
                results to calculate total consumption and energy content of each fuel
                type during future calendar years, are also described in detail in that
                same section.
                 The model documentation accompanying today's notice describes these
                procedures in detail.\2383\ The quantities of travel and fuel
                consumption estimated for the cross section of model years and calendar
                years constitutes a set of ``activity levels'' based on which the model
                calculates emissions. The model does so by multiplying activity levels
                by emission factors. As indicated in the previous section, the
                resulting estimates of vehicle use (VMT), fuel consumption, and fuel
                energy content are combined with emission factors drawn from various
                sources to estimate emissions of GHGs, criteria air pollutant, and
                airborne toxic compound that occur throughout the fuel supply and
                distribution process, as well as during vehicle operation, storage, and
                refueling. Emission factors measure the mass of each GHG or criteria
                pollutant emitted per vehicle-mile of travel, gallon of fuel consumed,
                or unit of fuel energy content. The following section identifies the
                sources of these emission factors and explains in detail how the CAFE
                model applies them to its estimates of vehicle travel, fuel use, and
                fuel energy consumption to estimate total annual emissions of each GHG,
                criteria pollutant, and airborne toxic.
                ---------------------------------------------------------------------------
                 \2383\ CAFE model documentation is available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
                ---------------------------------------------------------------------------
                (2) What emission factors did the agencies apply?
                (a) Tailpipe (Downstream) Emission Factors
                 In a full fuel cycle analysis, emissions that occur from the
                fueling pump to vehicle wheels are usually referred to as tailpipe or
                simply downstream emissions. Today's rule primarily impacts
                CO2 emissions. The agencies have calculated tailpipe
                CO2 emissions based on fuel consumption and fuel properties
                (i.e., fuel density and carbon content) that result in gram per gallon
                emission factors. For all other exhaust constituents (except sulfur
                dioxide, discussed below), the agencies have calculated emissions by
                applying per-mile emission factors to quantities of travel (i.e., VMT).
                This rulemaking's tailpipe emission factors are from EPA's Motor
                Vehicle Emission Simulator (MOVES), which serves as the federal
                regulatory model for mobile-source emission inventories, with a few
                notable exceptions. In particular, light-duty gasoline and diesel
                tailpipe emission factors for the following criteria pollutants,
                greenhouse gases (other than CO2), and air toxics are drawn
                from MOVES2014a: \2384\
                ---------------------------------------------------------------------------
                 \2384\ For the emission factors informing the Final EIS,
                updating to MOVES 2014b would have produced values identical to
                those based on MOVES 2014a.
                 Criteria pollutants
                 [cir] Carbon monoxide (CO),
                 [cir] Volatile organic compounds (VOC),
                 [cir] Nitrogen oxides (NOX), and
                 [cir] Fine particulate matter (PM2.5)
                 Greenhouse gases
                 [cir] Methane (CH4), and
                 [cir] Nitrous oxide (N2O)
                 Air toxics
                 [cir] Acetaldehyde,
                 [cir] Acrolein,
                 [cir] Benzene,
                 [cir] Butadiene,
                 [cir] Formaldehyde,
                 [cir] Diesel particulate matter (DPM10), and
                [[Page 24874]]
                 [cir] Methyl tert-butyl ether (MTBE)
                 These MOVES-based emission factors are specified separately for
                gasoline and diesel vehicles, by model year (ranging from MY 1975 to
                2050), and by vehicle age (ranging from zero to 39 years old). The
                structure of criteria pollutant emission standards is such that these
                factors do not vary with fuel economy unless a change in fuel type
                (e.g., from gasoline to electricity) is involved.
                 Since tailpipe sulfur dioxide (SO2) emissions are
                dependent on the sulfur content of the fuel, a single SO2
                emission factor in grams per million British thermal units (MMBTU) of
                fuel consumed is applied respectively for gasoline, diesel, and ethanol
                (E85) across all model years after MY 2017 based on a longitudinal
                analysis in MOVES.
                 As previously mentioned, EDF submitted supplemental comments on
                SO2 emissions, stating that ``SO2 emissions
                should be proportional to fuel consumption'' and ``that the tailpipe
                SO2 emissions by calendar year from the Volpe Model do not
                change proportionally to the changes in fuel consumption across various
                CO2 control scenarios.'' \2385\ The version of the model
                supporting the 2012 final rule calculated tailpipe SO2
                emissions on a gram per gallon basis. Supporting the ensuing rulemaking
                regarding heavy-duty pickups and vans, and the 2016 draft TAR, EPA
                staff provided SO2 emission factors specified on a gram per
                mile basis. DOT modified the model in order to apply these
                SO2 emission factors as provided by EPA. The CAFE Model
                documentation released with the NPRM clearly describes how the agencies
                calculated emissions in the model. Although the version of model
                applied for the NPRM did not change this approach to calculating
                tailpipe SO2 emissions, the agencies agree that
                SO2 emissions should be proportional to fuel consumption,
                and DOT has revised the model accordingly. For SO2
                emissions, the inputs to the model include the number of grams of
                SO2 emitted by a vehicle per gallon of fuel consumed by the
                vehicle.
                ---------------------------------------------------------------------------
                 \2385\ EDF, NHTSA-2018-0067-12363.
                ---------------------------------------------------------------------------
                 The agencies also received comments on the use of MOVES. Most
                notably, the National Farmers Union stated ``Concerns have been raised
                regarding the models used by EPA to determine emissions from fuels.
                Third-party reviews have shown that MOVES2014 may be inadequate as a
                tool for estimating the exhaust emissions of gasoline blends containing
                more than 10 percent ethanol. The model's results for mid-level ethanol
                blends have been shown to be inconsistent with other results from the
                scientific literature for both exhaust emissions and evaporative
                emissions, including results from real-world emissions testing.''
                \2386\ The agencies considered comments on the use of MOVES and ethanol
                blends and notes that MOVES may be unreliable for fuel blends over E10;
                however, MOVES is not designed to model mid-level ethanol blends.
                MOVES2014 is designed to model ethanol volumes up to 15 percent (E0 to
                E15), and it can also model E85 (ethanol volumes of 70 to 85 percent),
                but MOVES2014 is not designed to model intermediate fuel blends.
                Moreover, the agencies did not explicitly consider blends above E10 as
                part of the analysis, but rather ethanol blending is considered in
                relation to how to achieve a higher octane level and a higher anti-
                known index.
                ---------------------------------------------------------------------------
                 \2386\ National Farmers Union, NHTSA-2018-0067-11972.
                ---------------------------------------------------------------------------
                 The Pennsylvania Department of Environmental Protection stated that
                there may be a significant State-specific rebound effect in
                Pennsylvania given Pennsylvania's regional role in natural gas and
                petroleum processing and refining. According to this commenter, the
                proposed rule does not adequately take into account significant local,
                State, and regional air quality impacts because it dilutes the
                emissions impact of the rule across the entire Nation. The Center for
                Biological Diversity, the Consumer Federation of America, and other
                commenters expressed concern that the proposed rule would increase
                criteria pollutants in areas with large minority populations,
                especially those in areas near oil refineries.
                 Results of these tailpipe emissions calculations are summarized
                below in Section VII and in the FRIA accompanying today's notice, and
                presented in greater detail in the accompanying Final EIS.
                (b) Upstream Emission Factors
                 Fuel cycle emissions occurring between the extraction well and the
                fueling pump are often called upstream emissions. This rule has drawn
                upstream emission factors exclusively from the Greenhouse gases,
                Regulated Emissions, and Energy use in Transportation (GREET) model,
                developed by the U.S. Department of Energy's Argonne National
                Laboratory. The upstream gasoline, diesel, and electricity emission
                factors for criteria pollutants--namely, CO, VOC, NOX,
                PM2.5, and SO2--and greenhouse gases--namely,
                CO2, CH4, and N2O--have been updated
                with GREET 2018 data. The upstream emission factors for the air toxics
                mentioned above were unchanged from the proposal. For the final rule,
                upstream emission factors cover the following analysis years, 2017,
                2020, 2025, 2030, 2035, 2040, 2045, and 2050, and four distinct
                upstream processes:
                 Petroleum Extraction,
                 Petroleum Transportation,
                 Petroleum Refining, and
                 Fuel Transportation, Storage, and Distribution (TS&D).
                 These upstream emission factors for each fuel type and analysis
                year were generated by a process using emission factor values found in
                the GREET 2018 spreadsheet tool and adjustment factors where
                appropriate. Emission factors for the petroleum extraction process are
                the aggregation of different crude feedstock--such as crude oil, oil
                sands, and shale oil--emission factors multiplied by their associated
                adjustments for transportation to refineries losses, storage losses,
                and energy share by crude feedstock. Emission factors for the petroleum
                transportation process are emissions by crude feedstock sources--such
                as crude oil fields, surface and in-situ mining, and shale reserves--
                and multiplied the associated energy shares. Emission factors for the
                petroleum refining are the sum of the crude input, combustion, and non-
                combustion products multiplied by the transportation of blended fuel
                loss factors. The refining emission factors applies a non-ethanol
                energy content adjustment for gasoline, blended at E10. Diesel does not
                have any such ethanol content adjustment. Emission factors for the Fuel
                TS&D process are based on the blended fuel transportation and
                distribution emissions as well as an energy content factor for both the
                petroleum and ethanol portions of the fuels. Again, diesel does not
                have an ethanol adjustment.
                 The aggregated upstream emission factors used in the rule are
                aggregated across the four processes for each fuel type and analysis
                year. The aggregated upstream emission factor in the sum of the fuel
                TS&D emission factor, the petroleum refining emission factor multiplied
                by the share of fuel savings leading to reduced domestic refining, the
                pair of petroleum extraction and transportation emission factors
                multiplied by both the share of fuel savings and the share of reduced
                domestic refining from domestic crude. The upstream adjustments are
                replicated from the proposal.
                 Finally, the upstream emission factors for electricity are also
                updated with GREET 2018 data. Upstream electricity emissions factors
                are derived from
                [[Page 24875]]
                electricity for transportation use feedstock and fuel emissions by
                analysis year. As the analysis supporting the proposal noted, there are
                three possible supply ``pathways'' for fuel consumed by the U.S. light-
                duty vehicle fleet:
                 1. Importing fuel that has been refined overseas into the U.S.
                 2. Refining fuel in the U.S. from crude petroleum produced overseas
                and imported into the U.S.
                 3. Refining fuel in the U.S. from crude petroleum produced in the
                U.S.\2387\
                ---------------------------------------------------------------------------
                 \2387\ The proposal assumed that all fuel refined outside the
                U.S. and then imported into the U.S. would be refined from petroleum
                that was also produced outside the U.S. Although some of it could be
                refined from crude petroleum produced in the U.S. and exported, the
                analysis assumed that the fraction supplied via this pathway is
                negligible.
                ---------------------------------------------------------------------------
                 The distribution of fuel consumed within the U.S. that is supplied
                via each of these pathways has important implications for domestic
                ``upstream'' emissions, because each pathway produces domestic
                emissions arising from a different combination of activities that occur
                within the U.S. For example, pathway 1 involves domestic emissions that
                occur during crude petroleum extraction, transportation of crude oil
                from production or nearby temporary storage facilities to domestic
                refineries, refining of crude petroleum to produce transportation
                fuels, and storage and distribution of refined fuels.\2388\ In
                contrast, pathway 2 generates domestic emissions during transportation
                of crude petroleum from U.S. coastal ports to domestic refineries, as
                well as from fuel refining, storage, and distribution, while pathway 3
                produces domestic emissions only from storage and distribution of
                refined fuel.
                ---------------------------------------------------------------------------
                 \2388\ By longstanding EPA convention, emissions that occur when
                vehicles are being refueled at retail stations or vehicle storage
                depots (such as buses) are ascribed to vehicle use, rather than to
                fuel supply.
                ---------------------------------------------------------------------------
                 The analysis supporting the proposal made two central assumptions
                in estimating upstream emissions from fuel supply. First, 50 percent of
                any change in domestic fuel consumption by cars and light trucks
                operating on petroleum-based liquid fuels (gasoline and diesel) would
                be reflected in changes in imports of refined fuel, while the remaining
                50 percent would be reflected in changes in the volume of those fuels
                refined domestically. Second, 90 percent of any change in the volume of
                fuel refined domestically was assumed to be reflected in changes in the
                volume of crude petroleum imported into the U.S, with the remaining 10
                percent reflected in changes in the volume of crude petroleum produced
                within the U.S. The agencies developed these assumptions to analyze the
                environmental impacts of alternative CAFE and CO2 standards
                for model years 2012-2016, and have continued to rely in their analyses
                supporting subsequent rules.
                 To illustrate the effect of these assumptions, for each increase in
                domestic fuel consumption of 100 gallons, 50 additional gallons would
                be supplied via pathway 1 (refined outside the U.S. and imported in
                already-refined form). Additional fuel supplied via pathway 2 (U.S.
                domestic refining of imported crude oil) would account for 90 percent
                of the remaining 50 gallons of increased consumption, or 45 gallons.
                Finally, the remaining 5 gallons of increased fuel consumed within the
                U.S. would be supplied via pathway 3 (domestic refining of crude oil
                produced within the U.S.). This same breakdown was applied to changes
                in fuel consumption estimated to occur throughout the analysis period
                used for the proposal, which extended from 2017 through 2050.
                 The agencies estimated the resulting changes in upstream emissions
                of criteria air pollutants and airborne toxics occurring within the
                U.S. by applying emission factors for the appropriate stages of the
                fuel supply chain (petroleum extraction, petroleum transportation to
                refineries, fuel refining, and fuel storage and distribution) to the
                changes in the total energy content of fuel supplied by each pathway,
                and summed the results.\2389\ The energy content of fuel rather than
                its volume was used as the basis for estimating emissions, because
                emission factors are typically expressed in mass per unit of fuel
                energy supplied--for example, grams per million Btu--rather than per
                unit volume of fuel supplied.
                ---------------------------------------------------------------------------
                 \2389\ Increases in upstream GHG emissions were calculated from
                the increase in U.S. domestic fuel consumption, without regard to
                whether they occurred within the U.S.
                ---------------------------------------------------------------------------
                 In the proposal, the agencies made no explicit assumptions about
                the future mix of electric generating capacity that would be used to
                supply increased electricity consumed by BEVs and PHEVs. Instead, the
                agencies implicitly relied on the assumptions about future evolution of
                the nationwide mix of generation sources that were reflected in the
                U.S. average emission factors for electricity produced to power
                transportation vehicles, including cars and light trucks, which as
                described previously were drawn from the most recent version of Argonne
                National Laboratory's GREET model that was available at the time of the
                proposal. These assumptions were consistent with those made by EIA in
                its AEO 2017 Reference case analysis and publications.\2390\
                ---------------------------------------------------------------------------
                 \2390\ https://greet.es.anl.gov/publication-greet-2017-summary.
                ---------------------------------------------------------------------------
                 While the agencies' use of these assumptions to estimate upstream
                emissions did not prompt widespread comments on their analyses in
                support of previous CAFE rulemakings, the more recent proposal did draw
                a large number of comments focusing on those same assumptions. Most
                commenters asserted that the entirety of any increase in consumption of
                petroleum-based fuels by cars and light trucks resulting from the
                proposal would be met via increased domestic refining, primarily from
                crude petroleum produced in the U.S., and would thus generate
                additional upstream emissions within the U.S. throughout the fuel
                supply process. Even some commenters who argued elsewhere that the U.S.
                would continue to be a large-scale importer of petroleum asserted that
                the entire increase in fuel consumption resulting from the proposal
                would be refined from additional domestically-produced petroleum.\2391\
                ---------------------------------------------------------------------------
                 \2391\ For example, IPI notes that AEO 2019 shows the U.S. will
                continue to import crude petroleum through 2050, and will remain a
                net importer as measured by the energy content rather than the
                volume of U.S. petroleum exports and imports; see IPI, NHTSA-2018-
                0067-12213. Similarly, EDF argued that because U.S. petroleum
                imports have been declining and gasoline imports are currently low,
                the best assumption was that the entire increase in gasoline
                consumption resulting from the proposal would be supplied from
                increased domestic refining of U.S.-produced crude petroleum; see
                EDF, NHTSA-2018-0067-12108.
                ---------------------------------------------------------------------------
                 As a consequence, most commenters argued that the agencies'
                analysis of the proposal significantly underestimated the increases in
                upstream emissions that were likely to result, with some also asserting
                that the increases in emissions of criteria air pollutants would cause
                potentially serious degradation of air quality in the areas surrounding
                U.S. refineries. For example, EDF stated, ``NHTSA assumed that 50% of
                all the gasoline saved by more stringent CAFE and CO2
                standards would have been imported (i.e., refined overseas). . . . It
                is difficult to see how this could be the case when the nation is
                producing enough crude oil to be a net exporter. It is also difficult
                to see how this could be the case when gasoline consumption is
                decreasing and sufficient domestic refining capacity exists to fulfill
                today's demand, let alone decreased demand in the future. . . .
                Assuming that 100% of the differences in gasoline consumption between
                control scenarios will be refined in the U.S. appears to be much more
                consistent with the available data. Likewise, it seems reasonable to
                assume
                [[Page 24876]]
                that differences in the crude oil requirements of the various scenarios
                will also affect domestic production more so than imports.'' \2392\
                ---------------------------------------------------------------------------
                 \2392\ EDF, NHTSA-2018-0067-12108, p. 53. Others making similar
                assertions include IPI, NHTSA-2018-0067-12213, p. 5.
                ---------------------------------------------------------------------------
                 However, one commenter did agree with the agencies' assessment of
                the proposal's likely impact on U.S. petroleum imports, noting that
                ``Through 2050, there will only be a small increase in domestic oil
                production due to increased demand, well under 1%. . . . The vast
                majority (88% through 2050) of the additional petroleum that will be
                required to fuel light-duty vehicles in the proposed case will be
                imported. This assessment is not too far off of a single comment in the
                NPRM, `Using NEMS, it was estimated that 50% of increased gasoline
                consumption would be supplied by increased domestic refining and that
                90% of this additional refining would use imported crude petroleum.' ''
                \2393\
                ---------------------------------------------------------------------------
                 \2393\ David Gohlke, EPA-HQ-OAR-2018-0283-5082, p. 1.
                ---------------------------------------------------------------------------
                 The agencies note that there seems to be considerable confusion
                among commenters about the agencies' assumptions regarding import
                shares, and what they are attempting to measure. The agencies'
                assumptions are intended to measure the effects of changes in
                consumption of petroleum-derived transportation fuels by cars and light
                trucks that are attributable to this final rule on changes in U.S.
                production and imports of crude petroleum, in domestic refining of
                crude petroleum to produce transportation fuels, and in the volume of
                refined fuel distributed for domestic consumption. While recent data on
                U.S. fuel consumption, domestic production and imports of crude
                petroleum, and imports of refined petroleum products may be useful in
                estimating these desired measures, they are not themselves measures of
                the marginal impacts of changes in fuel consumption on the volumes of
                fuel supplied via each of the supply pathways described previously.
                 Instead, the agencies rely on two types of information to estimate
                the current and likely future values of the desired measures. First,
                they examine recent changes in domestic consumption of petroleum-based
                motor fuels--particularly gasoline, since it is the primary fuel used
                by vehicles that are subject to CAFE and CO2 standards--and
                compare them to the accompanying changes in the three gasoline supply
                pathways, namely domestic petroleum production, U.S. imports of crude
                petroleum, and U.S. imports of refined gasoline (or components that are
                blended domestically to produce gasoline). Second, the agencies examine
                differences in forecasts of U.S. petroleum production, fuel refining,
                and imports of refined fuel under alternative future scenarios that
                were included in AEO 2018 whose projections of domestic fuel
                consumption differ in ways that include alternative CAFE standards.
                While this latter approach would ideally compare scenarios that differ
                only in their assumptions about the stringency of CAFE and
                CO2 standards but are otherwise strictly comparable, such
                idealized comparisons are rarely possible because other factors almost
                always differ as well between the alternative scenarios being compared.
                (i) Assumptions Used To Analyze Impacts of the Final Rule on Petroleum
                Imports and Emissions
                 In response to comments, the agencies conducted a detailed
                examination of recent changes in U.S. fuel consumption, domestic fuel
                refining, and U.S. imports and exports of crude petroleum as well as
                refined fuel (primarily gasoline). This included comparing changes in
                these variables at both the national aggregate level and for three
                separate regions of the U.S. In addition, they examined differences in
                the forecast values of these variables under alternative assumptions
                about fuel economy standards, although as indicated above these
                comparisons are complicated by the fact that factors other than CAFE
                and CO2 standards also differ between these alternative
                scenarios.
                 The agencies also identified a fourth ``pathway'' to supply the
                increase in U.S. gasoline consumption anticipated to result from this
                final rule. The U.S. is now a net exporter of refined gasoline (and
                products that are blended to produce gasoline), and the volume of U.S
                gasoline exports is likely to increase for at least the next two
                decades. This introduces the possibility that some--and perhaps all--of
                the anticipated increase in domestic gasoline consumption will be met
                simply by redirecting U.S. gasoline exports to serve domestic
                consumption. This additional source of supply would result in no
                increase in domestic refining activity, and thus no increase in
                emissions from refining of petroleum-based transportation fuels.\2394\
                ---------------------------------------------------------------------------
                 \2394\ Increased domestic emissions would only occur in this
                case to the extent that domestic distribution of gasoline entailed
                higher emissions than transporting it to U.S. coastal ports for
                export.
                ---------------------------------------------------------------------------
                 Throughout most of the past half-century, the nation has been a
                large net importer of crude petroleum, taking its price as determined
                in world markets and importing the volumes necessary to meet the
                difference between U.S. demand for refined petroleum products and
                domestic supplies. Throughout this period, the U.S. has also been
                largely self-sufficient in refining, meaning that the gap between
                domestic demand for refined products and the volumes refined from crude
                petroleum extracted within the U.S. was primarily met by domestic
                refining of imported crude petroleum, with only marginal volumes of
                gasoline and other products imported or exported. U.S. refinery
                capacity and output generally increased over this period in proportion
                to growth in domestic consumption of fuel and other products refined
                from petroleum.
                 In the past decade, however, this situation has changed
                dramatically. U.S. production of crude petroleum has more than doubled
                since 2008, making the nation one of the world's largest producers,
                while net imports of crude oil and refined products have declined by
                nearly 80 percent.\2395\ Domestic gasoline consumption declined by more
                than 6 percent between 2007 and 2012, and recovered to its 2007 levels
                only as recently as 2016, remaining near or slightly below its 2016
                level since then.\2396\ As a consequence, the U.S. shifted from being a
                net importer of refined petroleum products to a net exporter in 2011,
                and has become a net exporter of gasoline and ``blending stock'' since
                2016.\2397\
                ---------------------------------------------------------------------------
                 \2395\ These and other petroleum statistics cited here were
                calculated from data available at EIA, Petroleum and Other Liquids,
                2019, https://www.eia.gov/petroleum/data.php. U.S. production of
                crude petroleum rose from 1.83 billion barrels in 2008 to 4.01
                billion barrels in 2018, or by 119%, During that same period, net
                U.S. imports of crude petroleum and refined products declined from
                4.07 billion to 0.85 billion barrels, or by 79%. Net U.S. imports
                are the difference between the nation's total (or gross) imports
                from elsewhere in the world and the volumes it exports to other
                nations.
                 \2396\ U.S. gasoline consumption declined from 3.39 billion
                barrels in 2007 to 3.18 billion barrels in 2012, or by 6.2 percent,
                rose to 3.41 billion barrels in 2016, and remained near that level
                through 2018.
                 \2397\ In 2010, U.S. net imports of refined petroleum products
                were 98 million barrels, but by 2011 U.S. net exports were 160
                million barrels. U.S. net exports of refined products then increased
                steadily through 2018, reaching 1.23 billion barrels in that year.
                In 2015, U.S. net imports of gasoline and blending components
                totaled 19 million barrels, but by 2016, U.S. net exports were 20
                million barrels, and grew to 93 million barrels in 2018. Another
                recent change in petroleum markets has been the increasing
                production and trade in gasoline blendstock in domestic and
                international petroleum trade. While in earlier periods refineries
                normally produced finished gasoline and shipped it to local storage
                terminals for distribution and retailing, in recent years,
                refineries have increasingly shifted to producing standardized
                gasoline blendstocks, such as Reformulated Blendstock for Oxygenate
                Blending (or ``RBOB''), which are then shipped and blended with
                ethanol or other additives to make finished gasoline that meets
                local regulatory requirements or customer specifications. Although
                this process has clear cost and operational advantages, particularly
                with extensive geographic and seasonal variation in gasoline
                formulations, it complicates the tabulation and comparison of
                petroleum statistics. In both EIA and most international trade
                statistics, finished gasoline and blendstocks are treated as
                separate products, and as reported in EIA statistics, large volumes
                of finished gasoline are now produced from blendstocks by local
                ``blenders,'' rather than by more centralized ``refiners.'' In
                addition, the volume of refinery production of gasoline and
                blendstock is now systematically lower than consumption of finished
                gasoline, because up to 10 percent of the volume of gasoline sold at
                retail can be made up of ethanol that is blended into gasoline after
                it leaves the refinery.
                ---------------------------------------------------------------------------
                [[Page 24877]]
                 Over the past decade, increased availability of crude petroleum and
                other refinery feedstocks in combination with declining gasoline
                consumption has presented U.S. refiners with a choice between
                continuing to produce gasoline at or near their capacity while boosting
                exports, or cutting back on refinery output. U.S. refiners elected not
                to cut back on their production of gasoline; instead, they actually
                increased the volume they refined. U.S. production of finished gasoline
                increased by 9 percent between 2007 and 2018.
                 The excess of gasoline production resulting from increased refinery
                capacity and stable consumption has partly displaced previous gasoline
                and blendstock imports, with the remainder taking the form of increased
                U.S. exports. Thus, as Figure VI-92 below shows, the nation now has a
                capacity to produce gasoline that considerably exceeds its current
                domestic consumption. This surplus of gasoline appears likely to
                increase in coming few years, as EIA's Annual Energy Outlook 2019
                reference case (EIA, 2019) anticipates that domestic gasoline
                consumption will continue to decline until nearly 2040. Therefore, the
                U.S. seems likely to remain a net exporter of gasoline through the next
                three decades.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.513
                 Although EIA's Annual Energy Outlook does not include separate
                forecasts of gasoline exports and imports, that same agency's Short
                Term Energy Outlook projects that U.S. gasoline exports will continue
                to rise through 2020 (EIA, 2019).\2398\ Combined with EIA's reference
                case forecast in the AEO 2019, the forecasts of declining U.S. gasoline
                consumption and rising net exports of refined petroleum products
                suggest that the United States will remain a growing net exporter of
                refined petroleum products--including gasoline--through nearly 2040. In
                turn, this suggests that any increase in domestic gasoline consumption
                resulting from this final rule is likely to
                [[Page 24878]]
                low anticipated growth in U.S. exports, rather than prompting growth in
                domestic refining and associated upstream emissions.
                ---------------------------------------------------------------------------
                 \2398\ AEO does not forecast gasoline refining, imports, or
                exports separately, instead reporting them as part of total refined
                petroleum products.
                ---------------------------------------------------------------------------
                 Regional patterns of U.S. gasoline consumption, refining, and trade
                also suggests that redirecting U.S. gasoline exports to domestic
                markets is likely to be an important source of additional supply to
                meet any increase in U.S. consumption stemming from this final rule.
                The nation's East Coast (which comprises the Energy Information
                Administration's Production and Distribution District 1, or PADD 1)
                currently accounts for about 32 percent of U.S. gasoline consumption,
                but has historically produced significantly less than gasoline than it
                consumes. As Figure VI-93 below shows, the gap between consumption and
                local supply within PADD1 has recently narrowed, as gasoline production
                along the East Coast has increased rapidly in recent years, while
                shipments into the region from the remainder of the U.S. and foreign
                imports (which come mostly from Canada) declined. In June 2019,
                however, press reports suggested that that one of the largest East
                Coast refineries (Philadelphia Energy Solutions, which represents some
                28 percent of East Coast refining capacity) would be closed.\2399\ At
                the same time, construction of new refineries continues to be hindered
                by the density of population concentrations and commercial development
                along the nation's East Coast, casting doubt on the potential for
                continued increases in local gasoline refining and supply within PADD
                1.
                ---------------------------------------------------------------------------
                 \2399\ Seba, E. (2019, July 5). Philadelphia refinery closing
                reverses two years of U.S. capacity gains. Retrieved September 19,
                2019, from Reuters: https://www.reuters.com/article/us-usa-refinery-blast-capacity/philadelphia-refinery-closing-reverses-two-years-of-u-s-capacity-gains-idUSKCN1U0283.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.514
                BILLING CODE 4910-59-C
                 As a consequence, it seems likely that at least in the near term,
                any increase in gasoline consumption along the Nation's East Coast in
                response to this rule would be supplied primarily by Gulf Coast
                refineries or increased foreign imports, rather than from increased
                production in East Coast refineries. Pipelines available to transport
                refined petroleum products from Gulf Coast refineries to the East Coast
                may also face capacity limitations, in which case most of any increase
                in gasoline consumption there would need to be met by increased imports
                from abroad. Over the longer term, however, it is possible that
                increases in East Coast gasoline consumption could be met partly by
                expanded refining activity within the region.
                 The West Coast, which includes Nevada and Arizona (EIA's PADD 5),
                currently accounts for 168 percent of
                [[Page 24879]]
                U.S. gasoline consumption. Almost all of the gasoline consumed in that
                region is also refined within it, although small volumes are shipped
                into Arizona from neighboring PADDs by pipeline, and small volumes are
                also exported to Latin America by tanker. The West Coast is relatively
                isolated from other U.S. sources of refined gasoline by long
                transportation distances and limited pipeline capacity, while import
                terminals for crude petroleum are relatively numerous, and it therefore
                appears more likely that marginal increases in gasoline consumption
                from the rule will be met from increases in local (i.e., within-PADD)
                refining. Figure VI-94 shows that this has been the case in recent
                decades, as growth in gasoline production within PADD 5 throughout that
                period has closely paralleled growth in local consumption, while net
                exports have remained minimal.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.515
                 The central region of the United States (PADDs 2-4) accounts for
                the remaining 52 percent of current U.S. gasoline consumption, while
                producing about three-quarters of the nation's gasoline and blendstock.
                Although as Figure VI-95 shows the central region was a minor net
                exporter of gasoline as recently as 2007, it now exports some 800,000
                barrels per day of gasoline and blendstock, and has accounted for
                virtually all of the recent growth in U.S. exports of these two
                categories of refined products. Recent press reports indicate that
                firms are currently making significant new investments to add refining
                capacity on the Gulf Coast to process the growing supply of U.S. shale
                oil (Douglas, 2019), and with the projected future decline in U.S.
                consumption, any additional gasoline refined there is likely to
                increase U.S. exports. Thus, future increases in gasoline consumption
                in the central region of the U.S. of the magnitude likely to result
                from adopting these final standards is expected to be met by diverting
                gasoline exports to domestic consumption, even in the absence of
                additional refinery investments.
                [[Page 24880]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.516
                BILLING CODE 4910-59-C
                 Table VI-278 below compares recent changes in gasoline consumption
                and various sources of supply for these three U.S. regions during the
                recent period (2012-18) when gasoline consumption has generally
                increased. As it shows, recent increases in consumption along the U.S.
                East Coast have been supplied by increased production within the
                region. As noted previously, however, it appears likely that production
                capacity there will contract significantly in the near term, and that
                future increases in consumption will need to be met from foreign
                imports or shipments from other U.S. regions. As the table also shows,
                recent increases in gasoline production in the Midwest and Gulf Coast
                region have been adequate to supply increased consumption within the
                region as well as major increases in foreign exports and shipments to
                other U.S. regions. Finally, increased consumption on the Nation's West
                Coast appears to have been met via a combination of increased
                production within the region and drawdowns of previously accumulated
                inventories (not shown in the table).
                 At the national level, where net shipments among regions
                necessarily cancel one another (resulting in the zero entry for Net
                Receipts from Other PADDS shown in the table), recent increases in
                production have been sufficient to meet increased domestic consumption,
                while simultaneously enabling a major increase in exports. This
                suggests that from the nationwide aggregate perspective, incremental
                increases in domestic gasoline consumption resulting from this rule
                could be met by a reduction in U.S. exports of domestically-refined
                gasoline to other nations, accompanied by increases in shipments from
                the Midwest and Gulf Coast regions to the nation's East and West
                Coasts.
                [[Page 24881]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.517
                 To summarize, based on changes in the various sources of supply
                that have accompanied recent changes in consumption within different
                regions of the U.S., the agencies anticipate that:
                 Most of any marginal increases in U.S. gasoline
                consumption resulting from this rule that occur on the East Coast of
                the U.S. is likely to be met in the near term by increased transfers
                from other regions of the U.S. or higher foreign imports, and possibly
                by expanded refining activity in the longer term;
                 Most of any marginal increases in U.S. gasoline
                consumption resulting from this rule that occur on the West Coast is
                likely to be supplied by increased gasoline refining within that
                region; and
                 Most or all of any marginal increase in U.S. gasoline
                consumption resulting from this rule that occurs in the Central region
                is likely to be supplied by redirecting foreign exports to supply
                markets within that region.
                 With these expectations and acknowledging the uncertainty
                surrounding them, the agencies have concluded that assuming 50 percent
                of any increase in U.S. gasoline consumption will lead to increased
                domestic refining activity--and thus to increases in domestic refinery
                emissions--continues to be reasonable, and perhaps even overstates the
                expected increase in domestic refinery emissions. In particular, the
                agencies find that assuming 50 percent is more reasonable than assuming
                that either none or 100 percent of any change in gasoline consumption
                will be translated into changes in domestic gasoline refining. Thus,
                the agencies have elected to continue to employ the 50 percent
                assumption in their central analysis, and to examine the sensitivity of
                its results to varying this fraction over the entire possible range,
                from zero to 100 percent.
                (ii) Changes in Crude Oil Supply to Domestic Refineries
                 The agencies also re-evaluated their assumption that 90 percent of
                the increase in crude petroleum refined in the U.S. to produce
                additional gasoline consumed as a result of this rule would be imported
                from abroad (thus resulting in increased emissions for its storage at
                import terminals, and transportation to domestic refineries), while the
                remaining 10 percent would be produced domestically (thus resulting in
                emissions from its extraction, local storage, and transportation to
                U.S. refineries). As discussed in more detail below, the agencies
                conclude that domestic petroleum production responds primarily to
                technological innovations, investments in exploration and development
                of new domestic sources of oil, and variation in the world price of
                petroleum, rather than to U.S. demand for refined products such as
                gasoline. As a consequence, they conclude that any increase in gasoline
                consumption attributable to this final rule is unlikely by itself to
                have a significant effect on domestic petroleum production, and that
                their previous assumption continues to be reasonable.
                 U.S. oil production is primarily a function of development
                opportunities identified during prior exploration programs, innovations
                in the technological for drilling and extracting crude petroleum,
                producer's expectations regarding future world petroleum prices, and
                the U.S. tax and regulatory situations surrounding petroleum
                exploration and production. Crude oil is a fungible, non-perishable
                commodity, and can usually be transported among local oil markets
                around the globe at some cost. As a consequence, the price of oil in a
                U.S. domestic market such as Texas is highly correlated with its price
                in markets located in Northern Europe, the Far East, and the Middle
                East.
                 In contrast, U.S. gasoline consumption depends on a broad array of
                factors that overlap only partially with the determinants of U.S. crude
                petroleum production. These include domestic economic growth and its
                consequences for transportation demand, current and future vehicle fuel
                economy, gasoline prices, excise and sales taxes levied on gasoline,
                technological and cultural changes, vehicle prices, and the evolution
                of transportation systems and the built environment.
                 As a consequence, changes in U.S. consumption and supply of
                petroleum products will primarily be reflected in changes the
                destinations of domestically produced and imported crude petroleum,
                rather than in changes in their production volumes. To the extent that
                changes in U.S. gasoline demand for lead to changes in the volume
                refined domestically (the subject of the previous analysis), increased
                refining activity is thus likely to be reflected in a shift in U.S.
                imports or exports of crude oil, rather than in a change in U.S.
                production of crude oil. Instead, any effect of this rule on U.S. crude
                oil production would arise primarily from the impact of increased
                domestic gasoline demand on global oil prices, which will be limited by
                the fact that U.S. gasoline demand accounts for a relatively small
                share of total global demand for petroleum products, and by
                [[Page 24882]]
                the response of global supply to any upward pressure on prices. Thus,
                any effect of this rule on U.S. petroleum production is likely to be
                extremely modest.\2400\
                ---------------------------------------------------------------------------
                 \2400\ U.S. gasoline consumption currently accounts for about 9%
                of total global demand for refined petroleum products, and the AEO
                2019 reference case projects that this will decline to 6% by the
                year 2035, and remain at that level through 2050. These figures are
                calculated from AEO 2019 Reference Case, Tables 11 and 21, available
                at https://www.eia.gov/outlooks/aeo/tables_ref.php.
                ---------------------------------------------------------------------------
                 Localized and temporary changes in domestic production might arise
                in response to capacity limitations or transportation bottlenecks
                associated with particular regions or refineries, which could
                temporarily create markets for higher-priced crude oil. However, these
                situations would normally be localized and prevail for only a limited
                time. At the same time, the effects of any change in domestic petroleum
                consumption on world oil prices would be attenuated, because as
                indicated previously the impact of increased domestic consumption would
                be felt on prices and volumes supplied in the much larger global
                petroleum market, rather than confined to the smaller U.S. market. Any
                resulting changes in global oil prices and petroleum production would
                inevitably be small when viewed on a world scale, and likely to prompt
                only minimal responses in U.S. petroleum supply.
                 As one indication of the likely minimal impacts of higher U.S.
                gasoline consumption on U.S. production of crude petroleum, EIA's
                Annual Energy Outlook 2018 included a side case called ``No New
                Efficiency Requirements,'' which included a freeze on U.S. fuel economy
                standards beginning in 2020. Although this scenario does not correspond
                exactly to either the agencies' earlier proposal or this final rule,
                comparing its results to those from the AEO 2018 reference case
                illustrates the insensitivity of domestic crude oil production to
                increases in gasoline consumption, as represented in EIA's National
                Energy Modeling System (NEMS).
                 Figure VI-96 below presents such a comparison, showing historical
                trends is U.S. gasoline consumption and petroleum production, and
                comparing their projected future trends in the AEO 2018 Reference Case
                and No New Efficiency Requirements alternative. As the figure
                illustrates, the large increase in U.S. gasoline consumption under the
                latter scenario relative to the Reference Case is accompanied by an
                almost indiscernible change in U.S. crude petroleum production, for
                exactly the reasons described above.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.518
                [[Page 24883]]
                BILLING CODE 4910-59-C
                 The agencies conclude that in the context of the current global
                petroleum market, increases in U.S. gasoline demand on the scale likely
                to result from this final rule are unlikely to produce changes in the
                market that prompt a significant increase in domestic petroleum
                production. Instead, they are likely to affect mainly the destinations
                and uses of crude petroleum--including refining gasoline within the
                U.S.--that is already being supplied to the global market. As a
                consequence, the agencies have elected to retain our previous
                assumption that any increase in domestic gasoline refining that occurs
                as a consequence of adopting this final rule is unlikely by itself to
                lead to a significant increase in domestic crude oil production or in
                the associated upstream emissions. Specifically, the agencies continue
                to assume that 10 percent of any increase in domestic gasoline refining
                would utilize increased U.S. production of crude petroleum.
                 The agencies chose to model upstream emissions in order to generate
                full fuel cycle emissions--using GREET for the upstream component and
                MOVES for the downstream component--because each alternative has
                varying levels of fuel consumption, and the specific gallons of
                gasoline, diesel, E85, and other fuels evaluated in today's rule will
                lead to different tailpipe and upstream emission outcomes.
                 While it may be fair to characterize MOVES and GREET as partial
                equilibrium models rather than general equilibrium models, the agencies
                did not make any modifications to the MOVES or GREET emission factors
                themselves. Changes in emission results were initiated through changes
                in fleet composition or activity, especially changes in vehicle miles
                travelled as well as vehicle sales and population. Other changes were
                made to average vehicle mass and road load coefficients such as
                aerodynamic drag and rolling resistance corresponding to the various
                regulatory alternatives. Each alternative consists of a package of
                technology changes, so a particular technology change was not modeled
                alone and would need to be evaluated separately to quantify incremental
                changes. Please consult the FRIA for quantified impacts for the
                technology packages laid out by alternative.
                d) How Did the Agencies Estimate and Value Health Impacts From Changes
                in Air Quality
                 The agencies' analyses estimates changes in the population-wide
                incidence of selected health impacts, as well as changes in the
                aggregate monetary value of those health impacts that may occur from
                the changes in emissions of criteria air pollutants projected to result
                from this final rule and the alternative that were considered. As with
                other estimated impacts of the final rule and alternatives, these
                changes are measured from a baseline that is represented by the
                adoption of the augural CAFE standards and the extension of EPA's
                updated CO2 estimates, providing a more precise accounting
                of physical impacts and costs and benefits of the standards, and also
                directly responds to comments, as discussed below.\2401\
                ---------------------------------------------------------------------------
                 \2401\ See EPA, Office of Air and Radiation, Office of Air
                Quality Planning and Standards, Technical Support Document,
                Estimating the Benefit per Ton of Reducing PM2.5
                Precursors from 17 Sectors, February 2018, available at https://www.epa.gov/sites/production/files/2018-02/documents/sourceapportionmentbpttsd_2018.pdf.
                ---------------------------------------------------------------------------
                 Many commenters expressed concern over the health impacts from
                increased GHG emissions and criteria pollutants. The American Lung
                Association et al. stated ``Today, nearly 40 percent of Americans--more
                than 124 million--live in communities in nonattainment for ozone and
                particulate matter, with many residents impacted more severely by local
                pollution sources, including near-road pollution. . . . Near-road
                pollution has been found to increase asthma attacks in children,
                cardiovascular health impacts, impaired lung function and premature
                death. . . . Reducing VOC emissions will help reduce the burden of
                these carcinogens on many communities, especially those living or
                working near these roadways.'' \2402\ As discussed in this Section, the
                agencies agree with these statements and have considered health effects
                as part of the analysis for today's rule. The Institute for Policy
                Integrity stated ``the agencies fixate on alleged on-road fatality
                effects while arbitrarily ignoring the mortalities, morbidities, and
                other welfare effects associated with emissions.'' \2403\ As described
                in this Section, in the analysis for this rule, the agencies estimate
                both air quality-related fatalities and their costs, in addition to the
                agencies' analysis on vehicle-related fatalities. Many public
                commenters also expressed concern for health issues associated with
                increased pollutants and emissions over what was anticipated by the
                agencies' 2012 analysis. The agencies carefully considered these
                comments and provided additional analysis to consider health impacts,
                as described below.
                ---------------------------------------------------------------------------
                 \2402\ American Lung Association et al., NHTSA-2018-0067-11765.
                 \2403\ Institute for Policy Integrity, NHTSA-2018-0067-12213.
                ---------------------------------------------------------------------------
                 The estimated health impacts reflect the nationwide baseline level
                of emissions of each pollutant, an assumed geographic distribution of
                increased emissions, the resulting changes in concentrations of
                criteria pollutants at various locations nationwide (some of which
                reflect accumulations of emissions, while others are chemical by-
                products formed in atmospheric reactions), increased exposure of the
                U.S. population to unhealthful concentrations of each pollutant, and
                the consequences of increased exposure for the aggregate frequency of
                each health impact. The agencies' analysis assumes that the increases
                in upstream and vehicle emissions are distributed in proportion to
                current emissions associated with fuel supply and vehicle use. This is
                consistent with the way EPA estimates health impacts and health damage
                costs for the refining and on-road mobile sources sectors, since those
                are estimated by assuming an increase in emissions from those sectors
                that is distributed in proportion to current emissions from each one,
                and estimating the resulting changes in accumulations of air
                pollutants, population exposure, health impacts, and associated
                monetary value. The accompanying estimates of per-ton damage costs
                apply unit values to the increased frequency of each health effect,
                representing the dollar costs or estimated willingness-to-pay to avoid
                its occurrence, and combine the results to estimate total damage costs.
                 EPA analysts utilize a large volume of underlying data, a number of
                intermediate calculations, and many simplifying assumptions to develop
                these estimates of health impacts and health damage costs per ton of
                additional emissions, and discussing these in detail is well beyond the
                scope of this rule. These underlying data, assumptions, and
                calculations are described in detail in the document that reports the
                values used for the agencies' analysis.\2404\ EPA quantifies health
                impacts and damage costs for emissions from 17 separate sectors of U.S.
                economic activity, and reports values for increases in premature
                mortality and the combined costs of damages from premature mortality
                and various other health impacts per ton of PM2.5, nitrate,
                [[Page 24884]]
                and sulfate emissions.\2405\ These values include high and low
                estimates of both premature mortality and health damage costs, which
                primarily reflect alternative published estimates of the premature
                mortality impact of PM2.5 emissions.\2406\ Alternative
                values are also reported for 3 percent and 7 percent discount rates;
                discounting affects the values because of the delay (or ``latency
                period'') between exposure to air pollution and the development of some
                health impacts, most notably premature deaths.
                ---------------------------------------------------------------------------
                 \2404\ See EPA, Office of Air and Radiation, Office of Air
                Quality Planning and Standards, Technical Support Document,
                Estimating the Benefit per Ton of Reducing PM2.5
                Precursors from 17 Sectors, February 2018, available at https://www.epa.gov/sites/production/files/2018-02/documents/sourceapportionmentbpttsd_2018.pdf.
                 \2405\ Premature mortality includes deaths that are estimated to
                occur before the normally expected life span of persons with
                specified demographic characteristics.
                 \2406\ Estimated willingness to pay to avoid premature death
                accounts for 98% of the total health damage costs included in these
                estimates; see EPA, p. 10.
                ---------------------------------------------------------------------------
                 The agencies' analysis uses those values for the petroleum refining
                sector (sector 15) to represent impacts resulting from emissions that
                occur during the fuel production and distribution process (upstream
                emissions), and those for the on-road mobile source sector (sector 13)
                to represent the impacts of emissions resulting from car and light
                truck use. The agencies apply EPA's estimates of per-ton increases in
                premature mortality and health damage costs for these sectors to their
                estimates of changes in nationwide total emissions of PM2.5,
                nitrogen oxides (NOx), and sulfur dioxide (SO2) from the
                fuel supply process and from car and light truck use.
                 Table VI-279 and Table VI-280 below report values the agencies used
                in the estimates of premature mortality impacts and total health damage
                costs per ton of emissions to analyze the consequences of this final
                rule. The results for this analysis are provided in Section VII of this
                rule. The dollar values reported in the tables below differ slightly
                from those reported in the underlying source, because they have been
                adjusted from the 2015$ used in that source to the 2018 dollars used
                throughout this analysis. Values for intervening years were
                interpolated from those shown in the tables, and values for the year
                2030 shown in the tables were assumed to prevail for years beyond 2030.
                The agencies' central analysis of the rule uses averages of the low and
                high values shown in each table, while the low and high values
                themselves are used in the sensitivity analyses described in Section
                VII of this rule.
                BILLING CODE 4910-59-P
                [[Page 24885]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.519
                [[Page 24886]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.520
                BILLING CODE 4910-59-C
                 The valuation of premature mortality effects rely on the results of
                ``benefits per ton'' approach (BPT). This approach is a reduced form
                approach, which is
                [[Page 24887]]
                less complex than full-scale air quality modeling, requiring less
                agency resources and time. Based on EPA's work to examine reduced form
                approach, the BPT may yield estimates of PM2.5--benefits for
                the mobile sector that are as much as 10 percent greater than those
                estimated when using full air quality modeling.
                 The EPA is currently working on a systematic comparison of results
                from its BPT technique and other reduced-form techniques with results
                from full-form photochemical modelling. While this analysis employed
                photochemical modeling simulations, we acknowledge that the Agency has
                elsewhere applied reduced-form techniques. The summary report from the
                ``Reduced Form Tool Evaluation Project'', which has not yet been peer
                reviewed, is available on EPA's website at https://www.epa.gov/benmap/reduced-form-evaluation-project-report. Under the scenarios examined in
                that report, EPA's BPT approach in the 2012 rule (which was based off a
                2005 inventory) may yield estimates of PM2.5--benefits for
                the mobile sector that are as much as 10 percent greater than those
                estimated when using full air quality modeling. The estimate increases
                to 30 percent greater for the electricity sector. The EPA continues to
                work to develop refined reduced-form approaches for estimating
                PM2.5 benefits.
                 In addition, considerable uncertainty surrounds many of the
                assumptions and other inputs used in the agencies' analysis of economic
                and environmental impacts likely to result from adopting the final
                standards, rather than ratifying the augural standards. Perhaps most
                notably, because fuel prices are inherently volatile and forecasts of
                their future level depend critically on developments in the often
                unstable and politicized global oil market, those forecasts are
                inherently uncertain, as evidenced by the fact that actual gasoline
                prices are well below those the agencies relied on in their 2012
                analysis of CAFE and CO2 standards for model years 2017-25.
                While the agencies' current analysis updates those projections to
                reflect EIA's 2019 Annual Energy Outlook, which now anticipates that
                future prices will remain well below those the agencies projected in
                their 2012 analysis, it remains possible that EIA's current forecast
                will continue to overestimate actual future prices (of course, EIA's
                current forecast could also prove to be too low, although the recent
                record suggests a larger risk that the opposite will be the case).
                Further, gasoline prices are only one of a number of assumptions about
                which the agencies have reason to be uncertain; others include the fuel
                economy and other features of car and light truck models that
                manufacturers will offer during future model years, how buyers will
                respond to changes in the features of competing models in the face of
                future fuel prices and economic conditions, and how much they (and
                subsequent owners) will ultimately drive the models they purchase over
                their lifetimes. Uncertainty about all of these factors is reflected in
                similar risks that the agencies' projections of changes in vehicle use
                and fuel consumption under the final standards will prove to be in
                error. Finally, uncertainty about the agencies' companion projections
                of those standards' impacts on PM emissions and premature mortality is
                compounded by the currently unknown effects of future control
                technologies and regulations on actual refinery and vehicle emissions,
                as well as by the sources of potential error in estimating the effects
                of changes in emissions on premature mortality discussed above.
                Although it may seem that the agencies' estimates of increases in
                premature mortality resulting from the final standards are more likely
                to be too high than too low, it is extremely difficult to anticipate
                whether this is actually the case.
                 Separately, the DEIS and FEIS accompanying this rule describe that
                the BPT estimates are subject to several assumptions and uncertainties
                that make it difficult to draw conclusions about the estimated monetary
                values.\2407\ Non-exhaustively, these reasons include that estimates do
                not reflect local variability in population density, meteorology,
                exposure, baseline health incidence rates, or other local factors that
                might lead to an overestimate or underestimate of the actual benefits
                of controlling fine particulates, and that the health impact studies
                include several sources of uncertainties, including: Within-study
                variability (the precision with which a given study estimates the
                relationship between air quality changes and health impacts), across-
                study variation (different published studies of the same pollutant/
                health effect relationship typically do not report identical findings,
                and in some cases the differences are substantial), the application of
                concentration-response functions nationwide (does not account for any
                relationship between region and health impact to the extent that there
                is such a relationship), and extrapolation of impact functions across
                population (the agencies assumed that certain health impact functions
                applied to age ranges broader than those considered in the original
                epidemiological study).
                ---------------------------------------------------------------------------
                 \2407\ See DEIS and FEIS at Chapter 4, Air Quality--Health
                Impacts.
                ---------------------------------------------------------------------------
                 Full-scale photochemical modeling provides the needed spatial and
                temporal detail to more precisely estimate changes in ambient levels of
                these pollutants and their associated impacts on human health and
                welfare. This modeling provides insight into the uncertainties
                associated with the use of benefit-per-ton estimates. The agencies
                conducted a photochemical modeling analysis for the Final EIS using the
                same methods as in the previous CAFE Final EISs 2408 2409
                and the HD Fuel Efficiency Standards Phases 1 and 2 Final
                EISs.2410 2411 The air quality modeling and health effects
                analysis focused on ozone and fine particulate matter equal to or less
                than 2.5 microns in diameter (PM2.5). As indicated in the
                Draft EIS, the agencies performed photochemical air quality modeling
                based on the inputs and emissions forecasts used in the Draft EIS.
                Consistent with prior rulemakings and as described in the scoping
                notice, to accommodate the substantial time required to complete the
                air quality modeling analysis, NHTSA proposed to initiate air quality
                modeling before the inputs and emissions forecasts for the Final EIS
                were finalized.\2412\ NHTSA received no public comments in response to
                the scoping notice addressing this analytical approach, and the agency
                proceeded accordingly. Therefore, NHTSA used the inputs and emissions
                forecasts for the Proposed Action and alternatives as stated in the
                Draft EIS for the analysis in this final rulemaking. For additional
                [[Page 24888]]
                information on the scoping notice and comments received, see Section X.
                ---------------------------------------------------------------------------
                 \2408\ NHTSA (2010). Final Environmental Impact Statement,
                Corporate Average Fuel Economy Standards, Passenger Cars and Light
                Trucks, Model Years 2012-2016. Washington, DC, National Highway
                Traffic Safety Administration.
                 \2409\ NHTSA (2012). Final Environmental Impact Statement,
                Corporate Average Fuel Economy Standards Passenger Cars and Light
                Trucks, Model Years 2017-2025, Docket No. NHTSA-2011-0056. July
                2012. Available at: https://one.nhtsa.gov/Laws-&-Regulations/CAFE-%E2%80%93-Fuel-Economy/Environmental-Impact-Statement-for-CAFE-Standards,-2017%E2%80%93202.
                 \2410\ NHTSA (2011). Final Environmental Impact Statement,
                Medium and Heavy-Duty Fuel Efficiency Improvement Program.
                Washington, DC, National Highway Traffic Safety Administration.
                 \2411\ NHTSA (2016). Phase 2 Fuel Efficiency Standards for
                Medium- and Heavy-Duty Engines and Vehicles. Final Environmental
                Impact Statement. Available at: https://www.nhtsa.gov/sites/nhtsa.dot.gov/files/mdhd2-final-eis.pdf.
                 \2412\ NHTSA, ``Notice of Intent to Prepare an Environmental
                Impact Statement for Model Year 2022-2025 Corporate Average Fuel
                Economy Standards,'' 82 FR 34740, 34743 fn. 15 (Jul. 26, 2017).
                ---------------------------------------------------------------------------
                 Some stakeholders submitted comments about the agencies' use of
                underlying NPRM modeling to conduct the photochemical modeling; for
                example, NCDEQ recognized the agencies statement that there was not
                sufficient time to collect the modeling, but stated that they
                ``strongly believe that the inputs and results should be readily
                available for public comment before the EIS and rulemaking are
                finalized.'' \2413\ Those comments are addressed in Section X and in
                the FEIS accompanying this rule. As part of EDF's alternative
                examination of the CAFE model and inputs, EDF utilized the same EPA
                benefit-per-ton method the agencies utilized for the final rule
                (discussed further below) to estimate health effects due to criteria
                pollutant emissions, concluding that the proposal would increase
                premature mortality due to increases in particulate matter emissions.
                EDF stated that these results indicated that the potential impacts of
                the rule are large, and accordingly, ``NHTSA and EPA must conduct
                detailed and thorough emission, photochemical and health effects
                modeling to quantify the effect of this or any other proposal to relax
                the CAFE and CO2 standards and increase upstream
                emissions.'' \2414\
                ---------------------------------------------------------------------------
                 \2413\ North Carolina Department of Environmental Quality,
                NHTSA-2018-0067-12025.
                 \2414\ Environmental Defense Fund, NHTSA-2018-0067-12108.
                ---------------------------------------------------------------------------
                 The agencies estimated air quality changes and health-related
                benefits at the national scale based on a detailed analysis of air
                quality and health effects throughout the contiguous 48 states.
                Different regions of the country could experience either a net increase
                or a net decrease in emissions because of the rule, depending on the
                relative magnitude of the changes in emissions from decreased fuel
                economy, decreased vehicle use, and increased fuel production and
                distribution under each alternative. The EIS air quality analysis
                addresses regional differences using grid-based air quality modeling
                and analysis techniques, which account for local and regional
                differences in emissions and many of the other factors (such as
                meteorology and atmospheric processes) that affect air quality and the
                resulting health effects at any given location. This air quality
                modeling analysis is intended as a screening application of both the
                Community Multiscale Air Quality (CMAQ) model and the Environmental
                Benefits Mapping and Analysis Program (BenMAP) tool for the purposes of
                quantifying and comparing the air quality and health-related benefits.
                 To examine and quantify the air quality and health-related benefits
                associated with implementing the final CAFE standards for MY 2021-2026
                light-duty vehicles, the agencies performed a national-scale
                photochemical air quality modeling and health benefit assessment with
                the following key steps:
                 Preparing emission inventories.
                 Modeling air quality.
                 Assessing air quality-related health impacts.
                 The following widely used tools were used for the air quality and
                health effects assessment:
                 Sparse-Matrix Operator Kernel Emissions (SMOKE) processing
                tool (version 3.7) to prepare model-ready emissions.
                 Community Multiscale Air Quality (CMAQ) model (version
                5.2.1) to quantify air quality changes for the different fuel economy
                alternatives.
                 Environmental Benefits Mapping and Analysis Program--
                Community Edition (BenMAP-CE) tool (version 1.4) to assess the health-
                related impacts of the simulated changes in air quality.
                 The national-scale modeling analysis employed the standard CMAQ
                continental modeling domain. The horizontal resolution of the grid for
                this modeling domain is 36 kilometers (22.4 miles). Air quality and
                health-related impacts were calculated for each grid cell in the entire
                contiguous United States (48 states). Although the modeling domain does
                not include all 50 states, nearly all of the affected emissions and
                population are included in the domain; therefore, the results are
                expected to represent those for a national-scale analysis. The agencies
                applied the CMAQ model for an annual simulation period using
                meteorological inputs for a base year of 2011.
                 The agencies performed modeling for 2035 (although the emission
                inputs represented a variety of different projection years, including
                2030, 2035, and 2040, based on best available data). As in the Draft
                EIS, the agencies chose 2035 for analysis of the various fuel economy
                alternatives because a large proportion of vehicles in operation are
                expected to meet the level of the standards set forth by 2035. EPA
                provided up-to-date, projected, national-scale emissions data for 2040
                for motor vehicles and for 2030 for all other sources. The emissions
                were processed for the 36-kilometer (22.4-mile) resolution modeling
                domain using SMOKE. The resulting model-ready inventories contain
                emissions for all criteria pollutants (as required for photochemical
                modeling) for multiple source categories (sectors), including on-road
                mobile sources, non-road mobile sources (e.g., construction equipment,
                locomotives, ships, and aircraft), electric generating unit (EGU) point
                sources, non-EGU point sources, area sources, and biogenic sources.
                 Following preparation of baseline emissions inventories, the
                baseline emissions for the light-duty vehicle portion of the on-road
                mobile emissions and the relevant upstream categories were replaced
                with data reflecting the alternatives analyzed in the Draft EIS. As
                discussed above, NHTSA calculated national estimates of on-road
                emissions for these vehicle classes for 2035, including both downstream
                and upstream emissions.
                 The agencies then applied CMAQ, using the emissions specific to
                each alternative. The simulated difference in air quality between the
                Draft EIS No Action Alternative and each action alternative represents
                the change in air quality associated with that alternative. Following
                the application of CMAQ, the agencies processed the CMAQ outputs for
                input to the BenMAP-CE health effects analysis tool, and used BenMAP-CE
                to estimate the health impacts and monetized health-related benefits
                associated with the changes in air quality simulated by CMAQ for each
                of the action alternatives. The BenMAP-CE tool includes health impact
                functions, which relate a change in the concentration of a pollutant
                with a change in the incidence of a health endpoint. BenMAP-CE also
                calculates the economic value of health impacts. For this study, the
                health effects analysis considered the effects of ozone and
                PM2.5. The PM2.5 analysis includes sulfate and
                nitrate particulates (secondary PM2.5) formed from emissions
                of SO2 (sulfur dioxide) and NOX, respectively.
                BenMAP-CE does not estimate health impacts associated with changes in
                directly emitted sulfur dioxide (SO2), carbon monoxide (CO),
                and other emissions. Health effects were calculated at the 36-kilometer
                scale (grid cell size) and aggregated nationally to determine overall
                impact.
                 Figure VI-97 shows the components of the air quality modeling and
                health-related benefits analysis. Note that both the emissions and
                meteorological inputs are used by SMOKE.
                [[Page 24889]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.521
                 Discussion of the photochemical modeling results is presented in
                the FEIS accompanying this final rule.
                E. Compliance Example Walk-Through
                 To illustrate the CAFE model's simulation of a manufacturer's
                potential response to fuel prices and new standards, the NPRM provided
                an example of how the preliminary version of the model showed, on a
                year-by-year basis, how GM could potentially respond to a set of CAFE
                standards, starting from MY 2016 (the latest year for which the
                agencies were able to develop a full and detailed characterization of
                the fleet of vehicles produced for sale in the U.S. at the time of
                publishing the NPRM). Although no analysis that does not rely heavily
                on a manufacturer's confidential product planning information can, with
                high fidelity, predict what that manufacturer will do, the CAFE model,
                by realistically reflecting product planning considerations in a
                detailed year-by-year context, can describe a course that manufacturer
                could realistically take. Indeed, when manufacturers provide
                information to the agencies, they often emphasize year-by-year plans.
                Although such information is typically considered confidential business
                information (CBI), public comments by the Alliance illustrate the
                concept for a hypothetical manufacturer. Although the illustration
                includes credit carry-back (aka borrowing) that most manufacturers have
                a history of avoiding, the illustration clearly demonstrates that the
                Alliance views product planning as a year-by-year exercise:
                BILLING CODE 4910-59-P
                [[Page 24890]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.522
                BILLING CODE 4910-59-C
                 Like the peer reviewers who examined the model's simulation of
                technology application and compliance, automakers have been widely
                supportive of the CAFE model's approach of year-by-year analysis
                informed by product planning realities. For example, Toyota commented,
                ``The preamble correctly notes that manufacturers try to keep costs
                down by applying most major changes mainly during vehicle redesigns and
                more modest changes during product refresh, and that redesign cycles
                for vehicle models can range from six to ten years, and eight to ten-
                years for powertrains. . . This appreciation for standard business
                practice enables the modeling to capture more accurately the way
                vehicles share engines, transmissions, and platforms. There are now
                more realistic limits placed on the number of engines and transmissions
                in a powertrain portfolio which better recognizes manufacturers must
                manage limited engineering resources and control supplier, production,
                and service costs.'' \2416\
                ---------------------------------------------------------------------------
                 \2415\ NHTSA-2018-0067-12073, at 28.
                 \2416\ NHTSA-2018-0067-12098, at 6.
                ---------------------------------------------------------------------------
                 The CAFE model's year-by-year approach to estimating manufacturers'
                potential responses to standards and fuel prices is consistent with
                EPCA/EISA's requirement that CAFE standards be set at the maximum
                feasible levels for each fleet (passenger car and light truck) in each
                model year. Some commenters correctly observe that the CAA (which
                provides no direction regarding tailpipe CO2 emissions
                standards) does not require such a year-by-year determination, but
                suggest, further, that EPA should refrain from making use of year-by-
                year analysis. In particular, CBD et. al. commented as follows:
                 Furthermore, the Volpe model and association [sic] tools are not
                designed in accordance with EPA's independent statutory authority
                under Clean Air Act Section 202. The Volpe and OMEGA models have an
                overarching difference in their architecture--one where the Volpe
                modeling approach is designed to match NHTSA's statutory authority,
                but not EPA's. The EPCA requirements drive the design of the Volpe
                model, in that it performs a year-by-year analysis in order to
                demonstrate that NHTSA is meeting its EPCA obligations. As a result,
                the Volpe model attempts to simulate for each manufacturer, by year,
                their refresh and redesign cadence across their vehicle platforms
                and then predict a manufacturer's technology deployment decision-
                making process for each platform. But under the Clean Air Act, EPA
                is not required to demonstrate that standards are set at the maximum
                feasible level year-by-year, as EPCA explicitly requires for
                NHTSA.\2417\
                ---------------------------------------------------------------------------
                 \2417\ NHTSA-2018-0067-12000, Appendix A, at 24-25.
                 Although CBD is correct that the CAA does not require a year-by-
                year determination or year-by-year analysis, CBD wrongly claims that
                the CAFE model's modeling approach is not ``in accordance'' with the
                CAA. CBD's claim is analogous to saying ``just say you want to drive
                across the country; don't bother looking at a map.'' As the NPRM
                demonstrated, the CAFE model can be used to simulate compliance with
                CO2 standards. That the model follows a year-by-year
                approach to doing so simply means that it takes greater pains to
                describe realistic pathways forward from a known model year.
                Manufacturers are by no means the only stakeholders to recognize that
                product planning is actually a year-by-year process. Supporting its
                comments on the agencies' proposal, CARB provided a study by Roush
                Industries, focusing on a potential design pathway for the Toyota
                RAV4.\2418\ While this report, which was cited by CARB in its comments,
                asserted the agencies' modeling underestimated fuel consumption
                benefits and overestimated costs, Roush, like the Alliance, clearly
                interpreted the question of realism as a
                [[Page 24891]]
                year-by-year question, as illustrated by the following chart in Roush's
                report:
                ---------------------------------------------------------------------------
                 \2418\ Rogers, G., ``Technical Review of: The Safer Affordable
                Fuel-Efficient (SAFE) Vehicles Rule for Model Years 2021-2026
                Passenger Cars and Light Trucks, Final Report.'' Roush Industries.
                October 25, 2018. See CARB, NHTSA-2018-0067-11984.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.523
                BILLING CODE 4910-59-C
                 While a year-by-year representation is essential to the estimation
                of pathways that individual manufacturers could realistically take to
                apply technologies to specific vehicle models, the CAFE model also
                accounts for a range of other important engineering and product
                planning considerations. For example, among specific vehicle models,
                engines and transmissions are often shared, and a given vehicle design
                platform may encompass a range of different specific vehicle models.
                This means not every configuration of every vehicle model can be as
                optimized for fuel economy as if each could be considered in isolation.
                This isn't to say that such optimization is technologically impossible,
                but rather to say that the resources involved in such optimization
                would be financially impracticable. Moreover, CAFE and CO2
                standards apply to fleets, not specific products. This means, for
                example, that if a given engine is shared among both passenger cars and
                light trucks, changes made to that engine in response to one fleet's
                standard will impact products in the other fleet. Consistent with the
                fact that CAFE and CO2 compliance applies to fleets on a
                year-by-year basis, the CAFE model explicitly accounts for sharing
                among specific model/configurations when simulating year-by-year
                compliance. The Roush report's authors ``have not performed a complete
                fleet-compliance simulation.'' \2420\ Therefore, even notwithstanding
                differences in estimates of redesign schedules and technology efficacy
                and costs, Roush's analysis of the RAV4 is highly idealized. As
                discussed below, together with inputs based on Toyota's actual MY 2017
                production, the CAFE model represents the RAV4 as encompassing multiple
                configurations, spanning both the passenger car and light truck
                regulatory classes, all on a common vehicle platform that includes
                several other vehicle models, and some RAV4s sharing engines with some
                Camrys. Compared to estimating the potential to apply technology to a
                handful of specific model/configurations in isolation, analysis that
                accounts for manufacturers' actual production considerations produces
                more realistic results.
                ---------------------------------------------------------------------------
                 \2419\ Rogers, G., ``Technical Review of: The Safer Affordable
                Fuel-Efficient (SAFE) Vehicles Rule for Model Years 2021-2026
                Passenger Cars and Light Trucks, Final Report,'' at 26. Roush
                Industries. October 25, 2018. See CARB, NHTSA-2018-0067-11984.
                 \2420\ Ibid. at 6.
                ---------------------------------------------------------------------------
                 Nothing about the CAA discourages realism in regulatory analysis,
                and even if the CAA did so, the CAFE model can easily be run for
                isolated model years, or run in a manner that otherwise ignores
                practical limits on development and manufacturing complexity.\2421\ EPA
                elected to use the CAFE model as designed because doing so produces a
                more realistic basis to estimate regulatory impacts. EPA considers its
                use of the CAFE model entirely consistent with all CAA and other
                statutory and other requirements governing the agency's development of
                [[Page 24892]]
                motor vehicle CO2 emissions standards which, unlike criteria
                pollutant standards, are specified on a year-by-year basis, and
                inherently involve the entirety of manufacturers' vehicles and fleets.
                ---------------------------------------------------------------------------
                 \2421\ Idealized simulation of compliance with a hypothetically
                isolated model year could be accomplished by, when running the
                model, setting the various ``start'' and ``end'' years to the same
                value. Sharing of engines and transmission among different model/
                configurations could be ignored by, in the CAFE model's ``market''
                input file, assigning each engine, transmission, and vehicle
                platform to a single model/configuration (e.g., such that each of
                the six versions of the RAV4 is on its own vehicle platform, and
                uses a dedicated engine and transmission).
                ---------------------------------------------------------------------------
                 Of course, like any other model, the CAFE model used for the NPRM
                had room for improvement. As discussed above, the agencies have
                responded to public comments by making changes to some aspects of the
                CAFE model itself. Only a few such changes, all of which are discussed
                above in greater detail, impact the CAFE model's simulation of
                manufacturers' application of fuel-saving technologies. Among these,
                three are especially important: First, the model now uses a more
                ``open'' application of its technology ``decision trees.'' While the
                primary objective of this change is to make the model's cost accounting
                more transparent (by recasting costs as absolute rather than
                incremental), it also makes the model somewhat more likely to identify
                and apply any highly cost-effective yet comparatively ``advanced''
                combinations of technology. Second, the model introduces a ``cost per
                credit'' metric for comparing available opportunities to add specific
                technologies to specific vehicles.\2422\ As discussed above and in the
                summary of the sensitivity analysis conducted for today's notice,
                changing from the NPRM's ``effective cost'' metric to this new ``cost
                per credit'' metric leads the model to, at least for the combination of
                inputs in today's central analysis, more frequently select less costly
                technology pathways than more costly pathways, at least when simulating
                compliance with CO2 standards. Third, the CAFE model can now
                extend its explicit simulation of manufacturers' technology application
                well into the future. Today's analysis extends this explicit simulation
                through model year 2050. Because today's reference case input estimates
                include continued increases in fuel prices alongside continued
                (``learning''-related) reductions in technology costs, extending the
                explicit simulation shows manufacturers making significant voluntary
                improvement in the longer term (e.g., after MY 2035), even if CAFE and
                CO2 remain unchanged.
                ---------------------------------------------------------------------------
                 \2422\ Notable comments on this metric appear at NHTSA-2018-
                0067-12039, Appendix, pp. 28-34, and at NHTSA-2018-0067-12108,
                Appendix B, pp. 66-70.
                ---------------------------------------------------------------------------
                 The agencies have also revised most of the inputs to the CAFE
                model, both to respond to comments and to better reflect an ever-
                changing world. Sections appearing above discuss changes to model
                inputs, such as the analysis fleet, technology-related inputs, and fuel
                prices. Many of these changes are important to the model's simulated
                application of fuel-saving technology. Updating the analysis fleet from
                a MY 2016 to a MY 2017 basis ensures that fuel economy and
                CO2 improvements manufacturers actually realized by adding
                technologies between those model years is accounted for, and ensures
                that changes in product offerings and production volumes between those
                model years are also accounted for. With this update, the agencies also
                more fully accounted for compliance credits accumulated prior to the
                MYs represented explicitly in today's analysis. Some manufacturers have
                accumulated large volumes of such credits, and are able to apply those
                credits well past MY 2016, and to trade them to other manufacturers.
                Updated vehicle simulations correct errors and make use of additional
                engine performance estimates (i.e., engine efficiency ``maps''), and
                cost estimates for some technologies reflect additional data and
                consideration of comments. Also, fuel prices in the forecast used for
                today's analysis are somewhat higher than those used for the NPRM; by
                itself, this change makes the model tend to show larger and more
                widespread voluntary fuel economy increases and accompanying
                CO2 emissions reductions, although this increased tendency
                is countered by the impact of changing to the ``cost per credit''
                metric.
                 The following example will illustrate the model's behavior when
                simulating compliance with CO2 standards. While the example
                focuses on the baseline CO2 standards and on a specific
                manufacturer (Toyota), and highlights a specific vehicle model (the
                Toyota RAV4), results for other scenarios, manufacturers, and vehicle
                models reflect application of the same logic. Because this example
                begins with the MY 2017 fleet, and does not make use of manufacturers'
                product plans (which the agencies have historically treated as
                confidential business information, today's analysis cannot and does not
                fully reflect manufacturers' actual product design decisions, even in
                the short term. Nevertheless, the analysis yields a realistic and
                detailed characterization of a path each manufacturer could take in
                response to a given set of standards and other input estimates (e.g.,
                of technology costs and fuel prices).
                 As discussed above, the model considers all models and model/
                configurations produced for sale in the U.S. by a given manufacturer.
                The Toyota Camry and Tundra are examples of specific Toyota passenger
                car and light truck models, Toyota produces a range of configurations
                (e.g., with different engines) of each of these vehicle models, and
                inputs to the CAFE model ensure that each such configuration is
                accounted for. CAFE model output files show the progressive application
                of technology to each model/configuration over time under each
                regulatory alternative. Here, focusing on different versions of one
                model, the RAV4, illustrates the process and results.
                 The RAV4 is one of the vehicle models included in a vehicle
                platform that also includes the Camry, Corolla, Prius, Lexus CT 200h,
                Lexus NX 200t, and Lexus NX 300h. As mentioned above, the CAFE model
                reflects the agencies' assumption that significant changes to vehicle
                structures or materials will most practicably be applied throughout a
                vehicle platform as models within the platform are redesigned. Within
                this platform, the CAFE model identifies the Corolla LE, at more than
                180,000 units produced in MY 2017, as the most likely ``leader'' for
                such changes. Inputs to today's analysis also show that most of the
                RAV4s produced for the U.S. in MY 2017 shared a 2.5L naturally
                aspirated 4-cylinder gasoline engine with many Camrys. The CAFE model
                identifies the Camry as the leader for new versions of that engine. The
                same inputs show many RAV4s shared a 6-speed automatic transmission
                with a range of other vehicle models, including the Avalon, Camry,
                Lexus ES 350, Highlander, Lexus NX 200t, and the CAFE model identifies
                the Camry as the most likely leader for changes to this transmission.
                Model inputs also show other RAV4s shared a different 6-speed automatic
                transmission with the Lexus NX 200t, and the CAFE model identifies the
                RAV4 as the most likely leader for changes to this transmission.
                Finally, the MY2017 RAV4 also included two ``strong'' (power split)
                hybrid-electric versions (SE and XLE). Although these shared an engine
                with other Toyota hybrids (Avalon, Camry, Lexus ES 300h and NX 300h),
                the CAFE model reflects the agencies' assumption that it could be
                practicable to ``split off'' plug-in (or fuel cell) configurations
                rather than necessarily replace all strong hybrids sharing an engine
                with PHEVs, BEVs, or FCVs.
                 Inputs for today's analysis have Toyota redesigning the RAV4 every
                five years, starting with MY 2019, and freshening the model 2-3 years
                after each redesign. Given this design cycle, and all the other inputs
                to today's analysis, the CAFE model shows that under the baseline
                CO2 standards,
                [[Page 24893]]
                Toyota could potentially make changes to the RAV4 summarized in the
                table that follows. The first changes occur in 2019, with Toyota
                improving aerodynamics of the hybrid RAV4s, and with the conventional
                RAV4s inheriting a new high compression ratio (HCR) engine introduced
                with the MY 2018 redesign of the Camry, and also adding 8-speed
                automatic (A8) transmissions,\2423\ improved accessories (IACC), and
                tires with reduced rolling resistance (ROLL20). With the MY 2024
                redesign, all versions of the RAV4 receive further aerodynamic
                improvements (AERO20) and ``Level 1'' mass reduction, engine friction
                reduction (EFR) is applied to the HCR engine the non-hybrid versions
                share with the Camry, and secondary axle disconnect (SAX) is applied to
                the non-hybrid versions of the RAV4. With the MY 2027 freshening,
                Toyota applies low-drag brakes to all the RAV4s. The MY 2029 redesign
                does not make any powertrain changes, but applies more significant mass
                reduction (MR3) to all RAV4s. In MY 2039, Toyota replaces the hybrid
                RAV4 SE and XLE with 200-mile (BEV200) and 300-mile (BEV300) electric
                vehicle, respectively.
                ---------------------------------------------------------------------------
                 \2423\ While it is not necessary for the compliance simulation
                to produce real predictions of manufacturer product designs, only
                plausible ones, these changes to the RAV4 did in fact occur during
                the 2019 redesign.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.524
                [[Page 24894]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.525
                 This progressive application of technology to the RAV4 produces a
                series of emission reductions shown in the following table (and, though
                not shown, corresponding fuel economy improvements). The table also
                shows the progression of CO2 targets for these vehicles,
                reflecting the fact that targets are higher for the hybrid and
                conventional AWD versions of the RAV4, classified as light trucks, than
                for the FWD RAV4s classified as passenger cars. Also notably, the
                conventional RAV4s never achieve their respective CO2
                emissions targets. This merely reflects the fact that credits for
                reducing A/C refrigerant leakage apply at the fleet level rather than
                on a per-vehicle basis and, in any event, Toyota can respond by
                improving CO2 levels enough among enough other vehicle
                models that Toyota's overall average CO2 levels comply with
                Toyota's overall requirements, taking into account the potential
                application of compliance credits.
                [[Page 24895]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.526
                 These CO2 values could be converted to equivalent fuel
                economy levels by multiplying their reciprocals by 8887 grams per
                gallon (e.g., 8887 g/gal x 1/(144 g/mi) = 62 mpg), differences in
                compliance provisions are such that results would be offset from actual
                fuel economy levels under CAFE standards. When simulating compliance
                with CAFE or CO2 standards, the CAFE model reports both fuel
                economy and CO2 targets and achieved levels, even when the
                model is ``enforcing'' compliance with only one of these sets of
                standards. When simulating
                [[Page 24896]]
                compliance with baseline CO2 standards, results for the
                example discussed here show the following fuel economy targets and
                achieved levels for the RAV4.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.527
                [[Page 24897]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.528
                 The progressive application of technology also produces increases
                (and some eventual decreases) in costs. For each RAV4 configuration,
                the following table shows costs beyond MY 2017 technology, in 2018
                dollars. The conventional RAV4s incur a significant cost increase in MY
                2019, primarily for the new HCR engine inherited from the Camry. Costs
                continue to increase through MY 2029 as additional technology
                accumulates, with another significant increase for MR4 in MY 2029.
                After MY 2029, technology costs for conventional RAV4s gradually
                decline through MY 2050, in response to ongoing learning. In MY 2039,
                the BEV200 RAV4 is less expensive than the HEV RAV4 it replaces,
                leading this version's cost to drop by about $500 between MY 2033 and
                MY 2034, and with learning, to fall quickly well below this version's
                MY 2017 cost. Conversely, the BEV300 RAV4 introduced in MY 2039 is
                about $950 more expensive than the MY 2038 hybrid RAV4 it replaces, and
                even with learning, the BEV300 remains more expensive through MY 2050
                than the hybrid RAV4. These BEVs are not needed for compliance; the
                model shows Toyota could introduce them because, if battery costs
                continue to decline while gasoline prices continue to increase, BEVs
                could eventually become attractive on an economic basis.
                [[Page 24898]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.529
                [[Page 24899]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.530
                BILLING CODE 4910-59-C
                 As mentioned above, by making sufficient improvements to other
                vehicle models, Toyota could refrain from making the conventional RAV4s
                meet their CO2 emissions targets. More broadly, Toyota can
                also use compliance credits to cover compliance gaps. The CAFE model
                accounts for the potential to transfer compliance credits between the
                passenger car (PC) and light truck (LT) fleets. The model also accounts
                for the potential to apply credits from prior model years (i.e.,
                credits that have been ``banked'' or, equivalently, ``carried
                forward''), including compliance credits earned prior to MY 2017. These
                aspects of the model interact with the model's accounting for multiyear
                planning--that is, the potential that a manufacturer, depending on its
                product design cadence and on the progression of standards, might apply
                ``extra'' technology in some model years in order to facilitate
                compliance in later model years. For example, if a manufacturer is only
                redesigning 15% of its fleet volume in MY 2025, that manufacturer might
                be best off--even setting aside credit banking--applying some ``extra''
                technology (i.e., technology that leads to overcompliance) as part of
                vehicle redesigns planned for MYs 2018-2024, and carrying that
                technology forward into MY 2025 when there are fewer opportunities
                available to reduce CO2 emissions in new models. As shown in
                Figure VI-100, in Toyota's case, the model shows that Toyota could
                offset its light truck compliance gaps during MY 2017-2019 by applying
                compliance credits earned for light trucks prior to MY 2017. The graph
                also shows Toyota applying extra technology to its passenger car fleet
                during MYs 2018-2024 in order to comply with the MY 2025 passenger car
                standard, but also to carry forward compliance credits and use those
                credits to offset large compliance gaps for Toyota's light truck fleet
                during MYs 2023-2027. After MY 2025, the model shows the effects of
                some technology continuing to be inherited (especially during MYs 2026-
                2030) from prior MYs, of Toyota continuing to make voluntary
                improvements where economically attractive (like the MY 2039 RAV4 EV
                mentioned above), and of Toyota continuing to transfer compliance
                credits from the passenger car to the light truck fleet.\2424\
                ---------------------------------------------------------------------------
                 \2424\ While the fleets (PC and LT) are shown separately for
                compliance purposes in this example, the ability to utilize credits
                from either fleet toward total model year compliance (in the current
                year, without caps or limits) means that the fleets for a
                manufacturer comply jointly in each model year.
                ---------------------------------------------------------------------------
                [[Page 24900]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.531
                 As the above figure shows, credit banking and transfers play an
                important role in Toyota's simulated response to the standards. If
                exercised in a manner that sets aside credit banking, the CAFE model
                shows Toyota increasing its application of fuel-saving technologies
                through MY 2025, and carrying those improvements forward, such that
                Toyota's overall average CO2 emission rate is 16 g/mi lower
                in MY 2025 when credit banking is not accounted for, as illustrated by
                the next chart appearing below. Though not shown here, accounting for
                credit banking also impacts the simulation other OEMs' compliance
                pathways, because inputs to today's analysis assume that Toyota would
                likely not need to use all of its pre-2017 compliance credits before
                these credits expire in 2021, and that Toyota could therefore sell
                those older credits other manufacturers (e.g., FCA, VW). By accounting
                for credit banking, the CAFE model thereby avoids considerable
                potential understatement of future CO2 emissions from light-
                duty vehicles.
                [[Page 24901]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.532
                 As indicated by the following chart, a failure to account for
                credit banking would also increase Toyota's modeled per-vehicle costs
                by nearly $1,000 in MY 2025. By accounting for credit banking, the CAFE
                model thus avoids considerable potential overstatement of compliance
                costs. Though not shown here, accounting for credit banking while also
                applying inputs that reflect Toyota's ability to sell older credits to
                some other OEMs also enables the CAFE model to avoid overstatement of
                compliance costs for those OEMs.
                [[Page 24902]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.533
                 While the model's simulation of manufacturers' potential responses
                to CAFE standards applies the same inputs and analytical methods, it
                does so accounting for several important statutory and regulatory
                differences between CO2 standards and CAFE standards, and
                for specific statutory direction regarding how CAFE standards are to be
                considered for purposes of setting standards at the maximum feasible
                levels in each model year. EPCA places specific limits on the amount of
                credit that can be transferred between fleets, and requires that
                domestic passenger cars meet minimum standards without applying
                credits. EPCA also requires that the determination of maximum feasible
                stringency set aside the potential to apply compliance credits or
                introduce new alternative fuel vehicles (include BEVs and FCVs, but not
                including plug-in HEVs) during the model years under consideration.
                Especially with standards that continue to become more stringent,
                applying these statutory constraints to the analysis leads the model to
                tend to show greater overcompliance with standards in earlier model
                years, because even setting aside the potential to carry forward or
                transfer credits, Toyota is likely to find it more practicable to apply
                some ``extra'' technology when redesigning vehicles during MYs 2017-
                2024 than to attempt to address MY 2025 standards by working with only
                vehicles scheduled to be redesigned in MY 2025. The model also tends to
                show greater overcompliance in later model years, because some of that
                extra technology from years leading up to the last year of stringency
                increases takes time to carry forward to ensuing model years. These
                aspects of the CAFE ``standard setting'' analysis are evident in the
                model's solution for Toyota, shown in the following figure. With the
                use of credits set aside after MY 2020, Toyota overcomplies with light
                truck standards during MYs 2018-2023 in order to carry technology
                forward into MY 2025. Although Toyota only marginally overcomplies with
                MY 2025 standards, the inheritance of technology during MYs 2026-2029
                contributes to increased overcompliance (which is to be expected given
                the degree of platform and powertrain sharing between the fleets).
                Continued increases in overcompliance after 2030 arise due to cost
                learning effects (especially for batteries) and increased fuel prices.
                [[Page 24903]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.534
                VII. What Does the Analysis Show, and What Does It Mean?
                A. Impacts of the Standards--Final and Alternatives
                 New CAFE and CO2 standards will have a range of impacts.
                EPCA/EISA and NEPA require DOT to consider such impacts when making
                decisions about new CAFE standards, and the CAA requires EPA to do so
                when making decisions about new emissions standards. Like past
                rulemakings, today's announcement is supported by the analysis of many
                potential impacts of new standards. Today's rulemaking finalizes new
                standards through model year 2026. While the CAFE model explicitly
                estimates manufacturers responses to standards through model year 2050
                and the associated impacts through calendar year 2089, today's
                rulemaking presents estimates of impacts on model years through MY
                2029, including impacts through these vehicles' full useful lives
                (i.e., for MY 2029 vehicles, through 2068). Today's rulemaking also
                presents estimates of overall impacts in each calendar year through
                2050, accounting for all model years through 2050. The agencies of
                course do not know today what will actually come to pass decades from
                now under the new final standards or under any of alternatives under
                consideration. The analysis is intended less as a forecast, than as an
                assessment--reflecting the best judgments regarding many different
                factors--of impacts that could occur.\2425\ As discussed below, the
                analysis was conducted using several defined alternatives to explore
                the sensitivity of this assessment to a variety of potential changes in
                key analytical inputs (e.g., fuel prices).
                ---------------------------------------------------------------------------
                 \2425\ ``Prediction is very difficult, especially if it's about
                the future.'' Attributed to Niels Bohr, Nobel Laureate in Physics.
                ---------------------------------------------------------------------------
                 This section summarizes various impacts of the final standards and
                other regulatory alternatives defined above. The no-action alternative
                provides the baseline relative to which all impacts are shown. Because
                the final standards (and the other alternatives considered), being of a
                ``deregulatory'' nature, are less stringent than the no-action
                alternative, all impacts are directionally opposite to impacts reported
                in recent CAFE and CO2 rulemakings. For example, while past
                rulemakings reported positive values for fuel consumption avoided under
                new standards, today's rulemaking reports negative values, as fuel
                consumption is expected be somewhat greater under today's new final
                standards than under standards defining the baseline no-action
                alternative. Reported negative values for avoided fuel consumption
                could also be properly interpreted as simply ``additional fuel
                consumption.'' Similarly, reported negative values for costs could be
                properly interpreted as ``avoided costs'' or ``benefits,'' and reported
                negative values for benefits could be properly interpreted as ``forgone
                benefits'' or ``costs.'' However, today's rulemaking retains reporting
                conventions consistent with past rulemakings, anticipating that,
                compared to other options, doing so will facilitate review by most
                stakeholders.
                 Today's analysis presents results for individual model years in two
                different ways. The first way is similar to past rulemakings and shows
                how
                [[Page 24904]]
                manufacturers could respond in each model year under the new final
                standards and each alternative covering MYs 2021/2022-2026. The second,
                expanding on the information provided in past rulemakings, evaluates
                incremental impacts of new standards for each model year, in turn. In
                past rulemaking analyses, NHTSA modeled year-by-year impacts under the
                aggregation of standards applied in all model years, and EPA modeled
                manufacturers' hypothetical compliance with a single model years'
                standards in that model year. Especially considering multiyear planning
                effects, neither approach provides a clear basis to attribute impacts
                to specific standards first introduced in each of a series of model
                years. For example, of the technology manufacturers applied in MY 2017,
                some would have been applied even under the MY 2014 standards, and some
                were likely applied to position manufacturers toward compliance with
                (including credit banking to be used toward) MY 2018 standards.
                Therefore, of the impacts attributable to the model year 2017 fleet,
                only a portion can be properly attributed to the MY 2017 standards, and
                the impacts of the MY 2017 standards involve fleets leading up and
                extending well beyond MY 2016. Considering this, the final standards
                were examined on an incremental basis, modeling each new model year's
                standards over the entire span of included model years, using those
                results as a baseline relative to which to measure impacts attributable
                to the next model year's standards. For example, incremental costs
                attributable to the new standards for MY 2023 are calculated as
                follows:
                COSTNew final,MY 2023 = (COSTNew final\through\MY 2023-COSTNo-
                Action\through\MY 2023)-(COSTNew final\through\MY 2022-COSTNo-
                Action\through\MY 2022)
                where
                COSTNew final,MY 2023: Incremental technology cost during MYs 2018-
                2029 and attributable to the new final standards for MY 2023.
                COSTNew final\through\MY 2022: Technology cost for MYs 2018-2029
                under new final standards through MY 2022.
                COSTNew final\through\MY 2023: Technology cost for MYs 2018-2029
                under new final standards through MY 2023.
                COSTNo-Action\through\MY 2022: Technology cost for MYs 2018-2029
                under no-action alternative standards through MY 2022.
                COSTNo-Action\through\MY 2023: Technology cost for MYs 2018-2029
                under no-action alternative standards through MY 2023.
                 Furthermore, today's analysis includes impacts on new vehicle sales
                volumes and the use (i.e., survival) of vehicles of all model years,
                such that standards introduced in a model year produce impacts
                attributable to vehicles having been in operation for some time. For
                example, as modeled here, standards for MY 2021 will impact the prices
                of new vehicles starting in MY 2017, and those price impacts will
                affect the survival of all vehicles still in operation in calendar
                years 2018 and beyond (e.g., MY 2021 standards impact the operation of
                MY 2007 vehicles in calendar year 2027). Therefore, while past
                rulemaking analyses focused largely on impacts over the useful lives of
                the explicitly modeled fleets, much of today's analysis considers all
                model years through 2029, as operated over their entire useful lives.
                For some impacts, such as on technology penetration rates, average
                vehicle prices, and average vehicle ownership costs, the focus was on
                the useful life of the MY 2029 fleet, as the simulation of
                manufacturers' technology application and credit use (when included in
                the analysis) continues to evolve after model year 2026, stabilizing by
                model year 2029.
                 Responding to comments recommending that the agencies present
                impacts on a calendar year basis, today's rulemaking does so, with the
                presented results extending through calendar year 2050, the last
                calendar year that includes an on-road fleet with all vehicle vintages
                represented.
                 Effects were evaluated from four perspectives: The social
                perspective, the manufacturer perspective, the private perspective, and
                the physical perspective. The social perspective focuses on economic
                benefits and costs, setting aside economic transfers such as fuel taxes
                but including economic externalities such as the social cost of
                CO2 emissions. The manufacturer perspective focuses on
                average requirements and levels of performance (i.e., average fuel
                economy level and CO2 emission rates), compliance costs, and
                degrees of technology application. The private perspective focuses on
                costs of vehicle purchase and ownership, including outlays for fuel
                (and fuel taxes). The physical perspective focuses on national-scale
                highway travel, fuel consumption, highway fatalities, and carbon
                dioxide and criteria pollutant emissions.
                 This analysis does not explicitly identify ``co-benefits,'' as such
                a concept would include all benefits other than cost savings to vehicle
                buyers. Instead, it distinguishes between private benefits--which
                include economic impacts on vehicle manufacturers, buyers of new cars
                and light trucks, and owners (or users) of used cars and light trucks--
                and external benefits, which represent indirect benefits (or costs) to
                the remainder of the U.S. economy that stem from the final rule's
                effects on the behavior of vehicle manufacturers, buyers, and users. In
                this accounting framework, changes in fuel use and safety impacts
                resulting from the final rule's effects on the number of used vehicles
                in use represent an important component of its private benefits and
                costs, despite the fact that previous analyses have failed to recognize
                these effects. The agencies' presentation of private costs and benefits
                clearly distinguishes between those that would be experienced by owners
                and users of cars and light trucks produced during previous model years
                and those that would be experienced by buyers and users of new cars and
                light trucks subject to the final standards. Moreover, it clearly
                separates these into benefits related to fuel consumption and those
                related to safety consequences of vehicle use. This is more meaningful
                and informative than simply identifying all impacts other than changes
                in fuel savings to buyers of new vehicles as ``co-benefits.''
                 For the social perspective, the following effects for model years
                through 2029 as operated through calendar year 2068 are summarized:
                 Technology Costs: Incremental cost, as expected to be paid
                by vehicle purchasers, of fuel-saving technology beyond that added
                under the no-action alternative.
                 Hybrid Vehicle Welfare Loss: Loss of value to vehicle
                owners resulting from incremental increases in the numbers of strong
                and plug-in hybrid electric vehicles (strong HEVs or SHEVs, and PHEVs)
                and/or battery electric vehicles (BEVs), beyond increases occurring
                under the no-action alternative.\2426\ The loss of value is a function
                of the factors that lead to different valuations for conventional and
                electric versions of similar-size vehicles (e.g., differences in:
                Travel range, recharging time versus refueling time, performance, and
                comfort).
                ---------------------------------------------------------------------------
                 \2426\ Through MY 2029, the ``standard setting'' analysis of
                CAFE standards sets aside the potential that manufacturers might by
                introduce new BEV (or FCV) vehicle models, but allows that the
                numbers of such vehicles produced might increase or decrease along
                with overall U.S. sales of new passenger cars and light trucks, and
                allows that additional BEV or FCV vehicle models might be intruded
                after MY 2029.
                ---------------------------------------------------------------------------
                 Pre-tax Fuel Savings: Incremental savings, beyond those
                achieved under the no-action alternative, in outlays for fuel
                purchases, setting aside fuel taxes.
                 Mobility Benefit: Value of incremental travel, beyond that
                [[Page 24905]]
                occurring under the no-action alternative.
                 Lost New Vehicle Consumer Surplus: Value of incremental
                savings to new vehicle buyers due to cheaper vehicle prices.
                 Implicit Opportunity Cost: \2427\ Value of other vehicle
                attributes forwent to apply technology to meet the standards.
                ---------------------------------------------------------------------------
                 \2427\ This value is set to ``0'' for the central analysis.
                ---------------------------------------------------------------------------
                 Refueling Benefit: Value of incremental reduction,
                compared to the no-action alternative, of time spent refueling
                vehicles.
                 Non-Rebound Fatality Costs: Social value of additional
                fatalities, beyond those occurring under the no-action alternative,
                setting aside any additional travel attributable to the rebound effect.
                 Rebound Fatality Costs: Social value of additional
                fatalities attributable to the rebound effect, beyond those occurring
                under the no-action alternative.
                 Benefits Offsetting Rebound Fatality Costs: Assumed
                further value, offsetting rebound fatality costs internalized by
                drivers, of additional travel attributed to the rebound effect.
                 Non-Rebound Non-Fatal Crash Costs: Social value of
                additional crash-related losses (other than fatalities), beyond those
                occurring under the no-action alternative, setting aside any additional
                travel attributable to the rebound effect.
                 Rebound Non-Fatal Crash Costs: Social value of additional
                crash-related losses (other than fatalities) attributable to the
                rebound effect, beyond those occurring under the no-action alternative.
                 Benefits Offsetting Rebound Non-Fatal Crash Costs: Assumed
                further value, offsetting rebound non-fatal crash costs internalized by
                drivers, of additional travel attributed to the rebound effect.
                 Additional Congestion and Noise (Costs): Value of
                additional congestion and noise resulting from incremental travel,
                beyond that occurring under the no-action alternative.
                 Energy Security Benefit: Value of avoided economic
                exposure to petroleum price ``shocks,'' the avoided exposure resulting
                from incremental reduction of fuel consumption beyond that occurring
                under the no-action alternative.
                 Avoided CO2 Damages (Benefits): Social value of
                incremental reduction of CO2 emissions, compared to
                emissions occurring under the no-action alternative.
                 Other Avoided Pollutant Damages (Benefits): Social value
                of incremental reduction of criteria pollutant emissions, compared to
                emissions occurring under the no-action alternative.
                 Total Costs: Sum of incremental technology costs, hybrid
                vehicle welfare loss, fatality costs, non-fatal crash costs, and
                additional congestion and noise costs.
                 Total Benefits: Sum of pretax fuel savings, mobility
                benefits, refueling benefits, Benefits Offsetting Rebound Fatality
                Costs, Benefits Offsetting Rebound Non-Fatal Crash Costs, energy
                security benefits, and benefits from reducing emissions of
                CO2, the CO2 equivalent of other associated
                gases, and criteria pollutants.
                 Net Benefits: Total benefits minus total costs.
                 Retrievable Electrification Costs: The portion of HEV,
                PHEV, and BEV technology costs which can be passed onto consumers,
                using the willingness to pay analysis described above.
                 Electrification Tax Credits: Estimates of the portion of
                HEV, PHEV, and BEV technology costs which are covered by Federal or
                State tax incentives.
                 Irretrievable Electrification Costs: The portion of HEV,
                PHEV, and BEV technology costs OEM's must either absorb as a profit
                loss, or cross-subsidize with the prices of internal combustion engine
                (ICE) vehicles.
                 Total Electrification Costs: Total incremental technology
                costs attributable to HEV, PHEV, or BEV vehicles.
                 For the manufacturer perspective, the following effects for the
                aggregation of model years 2017-2029 are summarized:
                 Average Required Fuel Economy: Average of manufacturers'
                CAFE requirements for indicated fleet(s) and model year(s).
                 Percent Change in Stringency from Baseline: Percentage
                difference between averages of fuel economy requirements under no-
                action and indicated alternatives.
                 Average Required Fuel Economy: Industry-wide average of
                fuel economy levels achieved by indicated fleet(s) in indicated model
                year(s).
                 Percent Change in Stringency from Baseline: Percentage
                difference between averages of fuel economy levels achieved under no-
                action and indicated alternatives.
                 Total Technology Costs ($b): Cost of fuel-saving
                technology beyond that applied under no-action alternative.
                 Total Civil Penalties ($b): Cost of civil penalties (for
                the CAFE program) beyond those levied under no-action alternative.
                 Total Regulatory Costs ($b): Sum of technology costs and
                civil penalties.
                 Sales Change (millions): Change in number of vehicles
                produced for sale in U.S., relative to the number estimated to be
                produced under the no-action alternative.
                 Revenue Change ($b): Change in total revenues from vehicle
                sales, relative to total revenues occurring under the no-action
                alternative.
                 Curb Weight Reduction: Reduction of average curb weight,
                relative to MY 2017.
                 Technology Penetration Rates: MY 2030 average technology
                penetration rate for indicated ten technologies (three engine
                technologies, advanced transmissions, and six degrees of
                electrification).
                 Average Required CO2: Average of manufacturers'
                CO2 requirements for indicated fleet(s) and model year(s).
                 Percent Change in Stringency from Baseline: Percentage
                difference between averages of CO2 requirements under no-
                action and indicated alternatives.
                 Average Achieved CO2: Average of manufacturers'
                CO2 emission rates for indicated fleet(s) and model year(s).
                 For the private perspective, the following effects for the MY 2030
                fleet are summarized:
                 Average Price Increase: Average increase in vehicle price,
                relative to the average occurring under the no-action alternative.
                 Implicit Opportunity Cost: The lost benefit of vehicle
                attributes that consumers prefer, which are sacrificed by manufacturers
                to comply with the standards.
                 Hybrid Vehicle Welfare Loss (Costs): Average loss of value
                to vehicle owners resulting from incremental increases in the numbers
                of strong HEVs, PHEVs) and/or BEVs, beyond increases occurring under
                the no-action alternative. The loss of value is a function of the
                factors that lead to different valuations for conventional and electric
                versions of similar-size vehicles (e.g., differences in: Travel range,
                recharging time versus refueling time, performance, and comfort).
                 Ownership Costs: Average increase in some other costs of
                vehicle ownership (taxes, fees, financing), beyond increase occurring
                under the no-action alternative.
                 Lost Consumer Surplus: Value of incremental savings to new
                vehicle buyers due to cheaper vehicle prices.
                 Fuel Savings: Average of fuel outlays (including taxes)
                avoided over a vehicle's expected useful lives, compared to outlays
                occurring under the no-action alternative.
                [[Page 24906]]
                 Mobility Benefit: Average incremental value of additional
                travel over average vehicles' useful lives, compared to travel
                occurring under the no-action alternative.
                 Refueling Benefit: Average incremental value of avoided
                time spent refueling over average vehicles' useful lives, compared to
                time spent refueling under the no-action alternative.
                 Total Costs: Sum of average price increase, welfare loss,
                and ownership costs.
                 Total Benefits: Sum of fuel savings, the mobility benefit,
                and the refueling benefit.
                 Net Benefits: Total benefits minus total costs.
                 For the physical perspective, the following effects for model years
                through 2029 as operated through calendar year 2068 are summarized:
                 Fuel Consumption, with rebound (billion gallons):
                Reduction of fuel consumption, relative to the no-action alternative,
                and including the rebound effect.
                 Fuel Consumption, without rebound (billion gallons):
                Reduction of fuel consumption, relative to the no-action alternative,
                and excluding the rebound effect.
                 Greenhouse Gases: Includes carbon dioxide
                (CO2), methane (CH4), and nitrous oxide
                (N2O), and values are reported separately for vehicles
                (tailpipe) and upstream processes (combining fuel production,
                distribution, and delivery) and shown as reductions in carbon dioxide
                or its equivalent relative to the no-action alternative.
                 Criteria Pollutants: Includes carbon monoxide (CO),
                volatile organic compounds (VOC), nitrogen oxides (NOX),
                sulfur dioxide (SO2) and particulate matter (PM), and values
                are shown as reductions relative to the no-action alternative.
                 Fuel Consumption: Aggregates all fuels, with electricity,
                hydrogen, and compressed natural gas (CNG) included on a gasoline-
                equivalent-gallon (GEG) basis, and values are shown as reductions
                relative to the no-action alternative.
                 VMT, with rebound (billion miles): Increase in highway
                travel (as vehicle miles traveled), relative to the no-action
                alternative, and including the rebound effect.
                 VMT, without rebound (billion miles): Increase in highway
                travel (as vehicle miles traveled), relative to the no-action
                alternative, and excluding the rebound effect.
                 Fatalities, with rebound: Increase in highway fatalities,
                relative to the no-action alternative, and including the rebound
                effect.
                 Fatalities, without rebound: Increase in highway
                fatalities, relative to the no-action alternative, and excluding the
                rebound effect.
                 Health Effects: Increase in the occurrence of a variety of
                health effects of criteria pollutant emissions, relative to the no-
                action alternative, and reported separately for tailpipe and upstream
                emissions.
                 Below, this section tabulates results for each of these four
                perspectives and does so separately for the new final CAFE and
                CO2 standards. More detailed results are presented in the
                FRIA accompanying today's rulemaking, and additional and more detailed
                analysis of environmental impacts for CAFE regulatory alternatives is
                provided in the corresponding Final Environmental Impact Statement
                (FEIS). Underlying CAFE model output files are available (along with
                input files, model, source code, and documentation) on NHTSA's
                website.\2428\ Summarizing and tabulating results for presentation here
                involved considerable ``off model'' calculations (e.g., to combine
                results for selected model years and calendar years, and to combine
                various components of social and private costs and benefits); tools
                Volpe Center staff used to perform these calculations are also
                available on NHTSA's website.\2429\
                ---------------------------------------------------------------------------
                 \2428\ Compliance and Effects Modeling System, National Highway
                Traffic Safety Administration, https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
                 \2429\ These tools, available at the same location, are scripts
                executed using R, a free software environment for statistical
                computing. R is available through https://www.r-project.org/.
                ---------------------------------------------------------------------------
                 While the National Environmental Policy Act (NEPA) requires NHTSA
                to prepare an EIS documenting estimating environmental impacts of the
                regulatory alternatives under consideration in CAFE rulemakings, NEPA
                does not require EPA to do so for EPA rulemakings. With CO2
                standards for each regulatory alternative being harmonized as practical
                with corresponding CAFE standards, environmental impacts of
                CO2 standards should be directionally identical and similar
                in magnitude to those of CAFE standards. Nevertheless, in this section,
                following the series of tables below, today's announcement provides a
                more detailed analysis of estimated impacts of the new final CAFE and
                CO2 standards. Results presented herein for the CAFE
                standards differ slightly from those presented in the FEIS; while, as
                discussed above, EPCA/EISA requires that the Secretary determine the
                maximum feasible levels of CAFE standards in manner that, as presented
                here, sets aside the potential use of CAFE credits or application of
                alternative fuels toward compliance with new standards, NEPA does not
                impose such constraints on any analysis presented in corresponding
                FEISs, and the FEIS presents results of an ``unconstrained'' analysis
                that considers manufacturers' potential application of alternative
                fuels and use of CAFE credits.
                 In terms of all estimated impacts, including estimated costs and
                benefits, the results of today's analysis are different for CAFE and
                CO2 standards. Differences arise because, even when the
                mathematical functions defining fuel economy and CO2 targets
                are ``harmonized,'' surrounding regulatory provisions may not be. For
                example, while both CAFE and CO2 standards allow credits to
                be transferred between fleets and traded between manufacturers, EPCA/
                EISA places explicit and specific limits on the use of such credits,
                such as by requiring that each domestic passenger car fleet meet a
                minimum CAFE standard (as discussed above). The CAA provides no
                specific direction regarding CO2 standards, and while EPA
                has adopted many regulatory provisions harmonized with specific EPCA/
                EISA provisions (e.g., separate standards for passenger cars and light
                trucks), EPA has not adopted all such provisions. For example, EPA has
                not adopted the EPCA/EISA provisions limiting transfers between
                regulated fleet or requiring separate compliance by domestic and
                imported passenger car fleets. Such differences introduce variance
                between impacts estimated under CAFE standards and under CO2
                standards. Also, as mentioned above, Congress has required that new
                CAFE standards be considered in a manner that sets aside the potential
                use of CAFE credits and the potential additional application of
                alternative fuel vehicles (such as electric vehicles) during the model
                years under consideration. Congress has provided no corresponding
                direction regarding the analysis of potential CO2 standards,
                and today's analysis does consider these potential responses to
                CO2 standards.
                 Tables in the remaining section summarize these estimated impacts
                for each alternative, considering the same measures as shown above for
                the final standards. For the final standards, social costs and
                benefits, private costs and benefits, and environmental and energy
                impacts were evaluated, and were done so separately for CAFE and
                CO2 standards defining each regulatory alternative. Also,
                for the final standards, the compliance-related private costs and
                [[Page 24907]]
                benefits were evaluated separately for domestic and imported passenger
                cars under CAFE standards but not under CO2 standards
                because EPCA/EISA's requirement for separate compliance applies only to
                CAFE standards.
                 Both the final standards and, all other alternatives involve
                standards less stringent than the no-action alternative. Therefore, as
                discussed above, incremental benefits and costs for each alternative
                are negative--in other words, each alternative involves forgone
                benefits and avoided costs. Environmental and energy impacts are
                correspondingly negative, involving forgone avoided CO2
                emissions and forgone avoided fuel consumption. For consistency with
                past rulemakings, these are reported as negative values rather than as
                additional CO2 emissions and additional fuel consumption.
                 Like the NPRM and PRIA (and past rulemakings), today's rulemaking
                and FRIA emphasize a ``model year'' perspective when reporting impacts.
                That is, for enough model years (here, through MY 2029) to extend
                beyond those when the estimated use of ``banked'' credits is reasonably
                likely to be sufficient to show the average manufacturer not achieving
                required CAFE or CO2 levels, the presentation of results
                mainly considers the lifetime impacts attributable to vehicles produced
                in these model years. Because standards are actually enforced on a
                model year basis, this perspective aligns well with the consideration
                of impacts on manufacturers and new vehicle buyers. However, impacts on
                national energy consumption and the natural environment will involve
                all vehicles on the road in future years, including those produced
                after MY 2029. Therefore, similar to the approach followed in recent
                and past EISs (and today's FEIS), today's rulemaking also presents
                impacts on a ``calendar year'' basis--that is, summarizing overall
                impacts (i.e., including those attributable to vehicles produced after
                MY 2029) in each calendar year through 2050. As discussed in below, the
                model year and calendar year perspectives draw on the same CAFE model
                outputs, but differ in the scope of those outputs included in
                summarized information.
                 As discussed above, more detailed results are available in the FRIA
                and FEIS accompanying today's rulemaking, as well as in underlying
                model output files posted on NHTSA's website.
                1. Average Required Fuel Economy and CO2 Standard for PCs,
                LTs, and Combined
                 The model fully represents the required CAFE and CO2
                levels for every manufacturer and every fleet. The standard for each
                manufacturer is based on the harmonic average of footprint targets (by
                volume) within a fleet, just as the standards prescribe. Unlike earlier
                versions of the CAFE model, the current version further disaggregates
                passenger cars into domestic and imported classes (which manufacturers
                report to NHTSA and EPA as part of their CAFE compliance submissions).
                This allows the CAFE model to more accurately estimate the requirement
                on the two passenger car fleets, represent the domestic passenger car
                floor (which must be exceeded by every manufacturer's domestic fleet,
                without the use of credits, but with the possibility of civil penalty
                payment), and allows it to enforce the transfer cap limit that exists
                between domestic and imported passenger cars, all for purposes of the
                CAFE program.
                 In calculating the achieved CAFE level, the model uses the
                prescribed harmonic average of fuel economy ratings within a vehicle
                fleet. Under an ``unconstrained'' analysis, or in a model year for
                which standards are already final, it is possible for a manufacturer's
                CAFE to fall below its required level without generating penalties
                because the model will apply expiring or transferred credits to
                deficits if it is strategically appropriate to do so. Consistent with
                current EPA regulations, the model applies simple (not harmonic)
                production-weighted averaging to calculate average CO2
                levels.
                 While the CAFE and CO2 standards themselves are, as
                discussed in Section VI, inputs to the agencies' analysis, because the
                standards are attribute-based standards specified separately for
                passenger car and light truck fleets and applicable to average fuel
                economy and CO2 levels, average requirements under these
                standards are analytical results, not analytical inputs. Also, because
                EPCA requires NHTSA to determine in advance minimum requirements that
                will be applicable to manufacturers' fleets of domestic passenger cars,
                these, too, are analytical results. The remainder of this section
                presents these results.
                a) Passenger Car Requirements
                 As discussed in Section V, the final standards are different from
                the preferred alternative identified in the proposal.
                 We do not know yet with certainty what CAFE and CO2
                levels will ultimately be required of individual manufacturers, because
                those levels will depend on the mix of vehicles that each manufacturer
                produces for sale in future model years. Based on the market forecast
                of future sales used to examine the final standards, the agencies
                currently estimate that the target functions shown above would result
                in the following average required fuel economy and CO2
                emissions levels for all manufacturers during MYs 2021-2026:
                [[Page 24908]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.535
                 We emphasize again that the values in these tables are estimates,
                and not necessarily the ultimate levels with which each of these
                manufacturers will have to comply, for the reasons described
                above.\2430\
                ---------------------------------------------------------------------------
                 \2430\ MY2017 values reflect the agencies' analysis, which uses
                an analysis fleet developed using MY2017 compliance data as of
                summer 2019. The analysis does not reflect subsequent updates and
                corrections to manufacturers' MY2017 compliance data.
                ---------------------------------------------------------------------------
                b) Light Truck Requirements
                 Again, while the agencies do not know yet with certainty what CAFE
                and CO2 levels will ultimately be required of individual
                manufacturers, because those levels will depend on the mix of vehicles
                that each manufacturer produces for sale in future model years, based
                on the market forecast of future sales used to examine today's proposed
                standards, the agencies currently estimate that the target functions
                shown above would result in the following average required fuel economy
                and CO2 emissions levels for individual manufacturers during
                MYs 2021-2026.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.536
                [[Page 24909]]
                 We emphasize again the values in these tables are estimates and not
                necessarily the ultimate levels with which each of these manufacturers
                will have to comply for reasons described above.\2431\
                ---------------------------------------------------------------------------
                 \2431\ MY2017 values reflect the agencies' analysis, which uses
                an analysis fleet developed using MY2017 compliance data as of
                summer, 2019. The analysis does not reflect subsequent updates and
                corrections to manufacturers' MY2017 compliance data.
                ---------------------------------------------------------------------------
                c) Average of PassengerCcar and Light Truck Requirements
                 Overall average requirements will depend, further, on the relative
                shares of passenger cars and light trucks in the new vehicle fleet. The
                agencies' analysis estimates future shifts in these shares as vehicles'
                average prices and fuel economy levels change, and as fuel prices also
                change. Resultant estimates of overall average requirements are as
                follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.537
                (d) Estimated Average Requirements for Specific Manufacturers
                 Overall average requirements (e.g., reflecting both passenger car
                and light truck fleets) applicable to each manufacturer will depend on
                the mix (i.e., footprint distribution) of vehicles produced in each
                model year, and relative production shares of passenger cars and light
                trucks. Tables appearing below summarize estimated requirements through
                model year 2029. Estimates for specific fleets (e.g., domestic
                passenger cars, imported passenger cars, light trucks) are available in
                CAFE model output files accompanying today's rulemaking, as are
                estimates for each MYs 2030-2050.\2432\
                ---------------------------------------------------------------------------
                 \2432\ The model and all inputs and outputs supporting today's
                notice are available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 24910]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.538
                [[Page 24911]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.539
                [[Page 24912]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.540
                [[Page 24913]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.541
                [[Page 24914]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.542
                [[Page 24915]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.543
                [[Page 24916]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.544
                [[Page 24917]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.545
                [[Page 24918]]
                BILLING CODE 4910-59-C
                2. Impacts on Vehicle Manufacturers
                 As mentioned above, impacts are presented from two different
                perspectives for today's final rule. From either perspective, overall
                impacts are the same. The first perspective, taken above in VII.A,
                examines overall impacts of the standards--i.e., the entire series of
                year-by-year standards--on each model year. The second perspective,
                presented here, provides a clearer characterization of the incremental
                impacts attributable to standards introduced in each successive model
                year. For example, the new final standards for MY 2023 are likely to
                impact manufacturers' application of technology in model years prior to
                MY 2023, as well as model years after MY 2023. By conducting analysis
                that successively introduces standards for each MY, in turn, isolates
                the incremental impacts attributable to new standards introduced in
                each MY, considering the entire span of MYs 1975-2029 and calendar
                years 2016-2069 included in the analysis that only considers the full
                series of successive MYs' standards. Tables appearing below summarize
                results as aggregated across these model and calendar years. Underlying
                model output files \2433\ report physical impacts and specific
                monetized costs and benefits attributable to each model year in each
                calendar (thus providing information needed to, for example,
                differentiate between impacts attributable to the MY 1975-2017 and MY
                2018-2029 cohorts). The FRIA presents costs and benefits for individual
                model years (with MY's 1975-2017 in a single bucket) for the final
                standards.
                ---------------------------------------------------------------------------
                 \2433\ Available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
                ---------------------------------------------------------------------------
                a) Industry Average Technology Penetration Rates
                 The CAFE model tracks and reports technology application and
                penetration rates for each manufacturer, regulatory class, and model
                year, calculated as the volume of vehicles with a given technology
                divided by the total volume. The ``application rate'' accounts only for
                those technologies applied by the model during the compliance
                simulation, while the ``penetration rate'' accounts for the total
                percentage of a technology present in a given fleet, whether applied by
                the CAFE model or already present at the start of the simulation.
                 In addition to the aggregate representation of technology
                penetration, the model also tracks each individual vehicle model on
                which it has operated. Accordingly, the CAFE model produces a record
                for every model year and every alternative that identifies with which
                technologies the vehicle started the simulation and which technologies
                the same vehicle had at the conclusion of each model year. Interested
                parties may use these outputs to assess how the compliance simulation
                modified any vehicle that was offered for sale in MY 2017 in response
                to a given regulatory alternative.
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.546
                [[Page 24919]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.547
                [GRAPHIC] [TIFF OMITTED] TR30AP20.548
                [[Page 24920]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.549
                [GRAPHIC] [TIFF OMITTED] TR30AP20.550
                [[Page 24921]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.551
                [GRAPHIC] [TIFF OMITTED] TR30AP20.552
                [[Page 24922]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.553
                [GRAPHIC] [TIFF OMITTED] TR30AP20.554
                [[Page 24923]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.555
                [GRAPHIC] [TIFF OMITTED] TR30AP20.556
                [GRAPHIC] [TIFF OMITTED] TR30AP20.557
                [[Page 24924]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.558
                [GRAPHIC] [TIFF OMITTED] TR30AP20.559
                [[Page 24925]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.560
                [[Page 24926]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.561
                [[Page 24927]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.562
                [[Page 24928]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.563
                [[Page 24929]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.564
                [[Page 24930]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.565
                [[Page 24931]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.566
                [[Page 24932]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.567
                [[Page 24933]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.568
                [[Page 24934]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.569
                [[Page 24935]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.570
                [[Page 24936]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.571
                [[Page 24937]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.572
                [[Page 24938]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.573
                [[Page 24939]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.574
                [[Page 24940]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.575
                [[Page 24941]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.576
                [[Page 24942]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.577
                [[Page 24943]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.578
                [[Page 24944]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.579
                [[Page 24945]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.580
                [[Page 24946]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.581
                [[Page 24947]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.582
                [[Page 24948]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.583
                [[Page 24949]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.584
                [[Page 24950]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.585
                [[Page 24951]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.586
                [[Page 24952]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.587
                [[Page 24953]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.588
                [[Page 24954]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.589
                [[Page 24955]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.590
                [[Page 24956]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.591
                [[Page 24957]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.592
                [[Page 24958]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.593
                [[Page 24959]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.594
                [[Page 24960]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.585
                [[Page 24961]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.596
                [[Page 24962]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.597
                [[Page 24963]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.598
                [[Page 24964]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.599
                [[Page 24965]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.600
                [[Page 24966]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.601
                [[Page 24967]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.602
                [[Page 24968]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.603
                [[Page 24969]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.604
                [[Page 24970]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.605
                [[Page 24971]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.606
                [[Page 24972]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.607
                [[Page 24973]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.608
                [[Page 24974]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.609
                [[Page 24975]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.610
                [[Page 24976]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.611
                [[Page 24977]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.612
                [[Page 24978]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.613
                [[Page 24979]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.614
                [[Page 24980]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.615
                [[Page 24981]]
                BILLING CODE 4910-59-C
                b) Technology Costs
                 For each technology that the model adds to a given vehicle, it
                accumulates cost. The technology costs are defined incrementally and
                vary both over time and by technology class, where the same technology
                may cost more to apply to larger vehicles as it involves more raw
                materials or requires different specifications to preserve some
                performance attributes. While learning-by-doing can bring down cost,
                and should reasonably be implemented in the CAFE model as a rate of
                cost reduction that is applied to the cumulative volume of a given
                technology produced by either a single manufacturer or the industry as
                a whole, in practice this notion is implemented as a function of time,
                rather than production volume. Thus, depending upon where a given
                technology starts along its learning curve, it may appear to be cost-
                effective in later years where it was not in earlier years. As the
                model carries forward technologies that it has already applied to
                future model years, it similarly adjusts the costs of those
                technologies based on their individual learning rates.
                BILLING CODE 4910-59-P
                [[Page 24982]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.616
                [[Page 24983]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.617
                [[Page 24984]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.618
                [[Page 24985]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.619
                [[Page 24986]]
                BILLING CODE 4910-59-C
                c) Civil Penalties
                 The other costs that manufacturers incur as a result of CAFE
                standards are civil penalties resulting from non-compliance with CAFE
                standards. The CAFE model accumulates costs of $5.50 per 1/10-MPG under
                the standard, multiplied by the number of vehicles produced in that
                fleet, in that model year. The model reports as the full ``regulatory
                cost,'' the sum of total technology cost and total fines by the
                manufacturer, fleet, and model year. As mentioned above, the relevant
                EPCA/EISA provisions do not also appear in the CAA, so this option and
                these costs apply only to simulated compliance with CAFE standards.
                d) Average Prices, Sales, and Revenue Changes
                 In all previous versions of the CAFE model, the total number of
                vehicles sold in any model year, in fact the number of each individual
                vehicle model sold in each year, has been a static input that did not
                vary in response to price increases induced by CAFE standards, nor
                changes in fuel prices, or any other input to the model. The only way
                to alter sales, was to update the entire forecast in the market input
                file. However, in the 2012 final rule, the agencies included a dynamic
                fleet share model that was based on a module in the Energy Information
                Administration's NEMS model. This fleet share model did not change the
                size of the new vehicle fleet in any year, but it did change the share
                of new vehicles that were classified as passenger cars (or light
                trucks). That capability was not included in the central analysis but
                was included in the uncertainty analysis, which looked at the baseline
                and final standards in the context of thousands of possible future
                states of the world. As some of those futures contained extreme cases
                of fuel prices, it was important to ensure consistent modeling
                responses within that context. For example, at a gasoline price of $7/
                gallon, it would be unrealistic to expect the new vehicle market's
                light truck share to be the same as the future where gasoline cost $2/
                gallon. The current model has slightly modified, and fully integrated,
                the dynamic fleet share model. Every regulatory alternative and
                sensitivity case considered for this analysis reflects a dynamically
                responsive fleet mix in the new vehicle market.
                 While the dynamic fleet share model adjusts unit sales across body
                styles (cars, SUVs, and trucks), it does not modify the total number of
                new vehicles sold in a given year. The CAFE model now includes a
                separate function to account for changes in the total number of new
                vehicles sold in a given year (regardless of regulatory class or body
                style), in response to certain macroeconomic inputs and changes in the
                average new vehicle price. The price impact is modest relative to the
                influence of the macroeconomic factors in the model. The combination of
                these two models modify the total number of new vehicles, the share of
                passenger cars and light trucks, and, as a consequence, the number of
                each given model sold by a given manufacturer. However, these two
                factors are insufficient to cause large changes to the composition of
                any of a manufacturer's fleets. In order to change significantly the
                mix of models produced within a given fleet, the CAFE model would
                require a way to trade off the production of one vehicle versus another
                both within a manufacturer's fleet and across the industry. While the
                agencies have experimented with fully-integrated consumer choice
                models, their performance has yet to satisfy the requirements of a
                rulemaking analysis.
                 Above, Section VI discusses at length the sales model the agencies
                have applied in the analysis supporting today's rulemaking.
                BILLING CODE 4910-59-P
                [[Page 24987]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.620
                [[Page 24988]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.621
                [[Page 24989]]
                BILLING CODE 4910-59-C
                e) Labor
                 As discussed in Section VI the analysis includes estimates of
                impacts on U.S. auto industry labor, considering the combined impact of
                changes in sales volumes and changes in outlays for additional fuel-
                saving technology. Note: This analysis does not consider the
                possibility that potential new jobs and plants attributable to
                increased stringency will not be located in the United States, or that
                increased stringency will not lead to the relocation of current jobs or
                plants to foreign countries. Compared to the no-action alternative
                (i.e., the baseline standards), the new final standards (alternative 1)
                and other regulatory alternatives under consideration all involve
                reduced regulatory costs expected to lead to reduced average vehicle
                prices and, in turn, increased sales. While the increased sales
                slightly increase estimated U.S. auto sector labor hours, because
                producing and selling more vehicles uses additional U.S. labor, the
                reduced outlays for fuel-saving technology slightly reduce estimated
                U.S. auto sector labor hours, because manufacturing, integrating, and
                selling less technology means using less labor to do so. Of course,
                this is technology that may not otherwise be produced or deployed were
                it not for regulatory mandate, and the additional costs of this
                technology would be borne by a reduced number of consumers given
                reduction in sales in response to increased prices. Today's analysis
                shows the negative impact of reduced mandatory technology outlays
                outweighing the positive impact of increased sales. However, both of
                these underlying factors are subject to uncertainty. For example, if
                fuel-saving technology that would have been applied under the baseline
                standards is more likely to have come from foreign suppliers than
                estimated here, less of the forgone labor to manufacture that
                technology would have been U.S. labor. Also, if sales would be more
                positively impacted by reduced vehicle prices than estimated here,
                correspondingly positive impacts on U.S. auto sector labor could be
                magnified. Alternatively, if manufacturers are able to deploy
                technology to improve vehicle attributes that new car buyers prefer to
                fuel economy improvements, both technology spending and vehicle sales
                would correspondingly increase.
                 The labor utilization analysis was focused on automotive labor
                because adjacent labor utilization factors and consumer spending
                factors for other goods and services are uncertain and difficult to
                predict. How direct labor changes may affect the macro economy and
                possibly change employment in adjacent industries were not considered.
                For instance, possible labor changes in vehicle maintenance and repair
                were not considered, nor were changes in labor at retail gas stations
                considered. Possible labor changes due to raw material production, such
                as production of aluminum, steel, copper, and lithium were not
                considered, nor were possible labor impacts due to changes in
                production of oil and gas, ethanol, and electricity considered. Effects
                of how consumers could spend money saved due to improved fuel economy
                were not analyzed, nor were effects of how consumers would pay for more
                expensive fuel savings technologies at the time of purchase analyzed;
                either could affect consumption of other goods and services, and hence
                affect labor in other industries. The effects of increased usage of
                car-sharing, ride-sharing, and automated vehicles were not analyzed.
                How changes in labor from any industry could affect gross domestic
                product and possibly affect other industries as a result were not
                estimated.
                 Also, no assumptions were made about full-employment or not full-
                employment and the availability of human resources to fill positions.
                When the economy is at full employment, a fuel economy regulation is
                unlikely to have much impact on net overall U.S. labor utilization;
                instead, labor would primarily be shifted from one sector to another.
                These shifts in employment impose an opportunity cost on society,
                approximated by the wages of the employees, as regulation diverts
                workers from other activities in the economy. In this situation, any
                effects on net employment are likely to be transitory as workers change
                jobs (e.g., some workers may need to be retrained or require time to
                search for new jobs, while shortages in some sectors or regions could
                bid up wages to attract workers). On the other hand, if a regulation
                comes into effect during a period of high unemployment, a change in
                labor demand due to regulation may affect net overall U.S. employment
                because the labor market is not in equilibrium. Schmalansee and Stavins
                point out that net positive employment effects are possible in the near
                term when the economy is at less than full employment due to the
                potential hiring of idle labor resources by the regulated sector to
                meet new requirements (e.g., to install new equipment) and new economic
                activity in sectors related to the regulated sector. In the longer run,
                the net effect on employment is more difficult to predict and will
                depend on the way in which the related industries respond to the
                regulatory requirements. For that reason, this analysis does not
                include multiplier effects but instead focuses on labor impacts in the
                most directly affected industries. Those sectors are likely to face the
                most concentrated labor impacts.
                 The tables presented below summarize these results for the final
                standards and other regulatory alternatives considered. While values
                are reported as thousands of person-years, changes in labor utilization
                would not necessarily involve the same number of changes in actual
                jobs, as auto industry employers may use a range of strategies (e.g.,
                shift changes, overtime) beyond simply adding or eliminating jobs.
                (1) CAFE Standards
                [[Page 24990]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.622
                [GRAPHIC] [TIFF OMITTED] TR30AP20.623
                (2) CO2 Standards
                [[Page 24991]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.624
                3. Impacts to Vehicle Buyers
                BILLING CODE 4910-59-P
                [[Page 24992]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.625
                [[Page 24993]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.626
                [[Page 24994]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.627
                [[Page 24995]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.628
                [[Page 24996]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.629
                [[Page 24997]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.630
                [[Page 24998]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.631
                [[Page 24999]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.632
                [[Page 25000]]
                BILLING CODE 4910-59-C
                a) Average Price Increase
                BILLING CODE 4910-59-P
                [[Page 25001]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.633
                [[Page 25002]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.634
                [[Page 25003]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.635
                [[Page 25004]]
                BILLING CODE 4910-59-C
                4. Impacts to Society
                 As the CAFE model simulates manufacturer compliance with regulatory
                alternatives, it estimates and tracks a number of consequences that
                generate social costs. The most obvious cost associated with the
                program is the cost of additional fuel economy improving/CO2
                emissions reducing technology that is added to new vehicles as a result
                of the rule. However, the model does not inherently draw a distinction
                between costs and benefits. For example, the model tracks fuel
                consumption and the dollar value of fuel consumed. This is the cost of
                travel under a given alternative (including the baseline). The ``cost''
                or ``benefit'' associated with the value of fuel consumed is determined
                by the reference point against which each alternative is considered.
                The CAFE model reports absolute values for the amount of money spent on
                fuel in the baseline, then reports the amount spent on fuel in the
                alternatives relative to the baseline. If the baseline standard were
                fixed at the current level, and an alternative achieved significantly
                greater mpg by 2025, the total expenditures on fuel in the alternative
                would be lower, creating a fuel savings ``benefit.'' This analysis uses
                a baseline that is more stringent than each alternative considered, so
                the incremental fuel expenditures are greater for the alternatives than
                for the baseline.
                 Other social costs and benefits emerge as the result of physical
                phenomena, like tailpipe emissions or highway fatalities, which are the
                result of changes in the composition and use of the on-road fleet. The
                social costs associated with those quantities represent an economic
                estimate of the social damages associated with the changes in each
                quantity. The model tracks and reports each of these quantities by:
                Model year and vehicle age (the combination of which can be used to
                produce calendar year totals), regulatory class, fuel type, and social
                discount rate.
                 The full list of potential costs and benefits is presented in Table
                VII-90 as well as the population of vehicles that determines the size
                of the factor (either new vehicles or all registered vehicles) and the
                mechanism that determines the size of the effect (whether driven by the
                number of miles driven, the number of gallons consumed, or the number
                of vehicles produced).
                BILLING CODE 4910-59-P
                [[Page 25005]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.636
                [[Page 25006]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.637
                [[Page 25007]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.638
                [[Page 25008]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.639
                [[Page 25009]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.640
                [[Page 25010]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.641
                [[Page 25011]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.642
                [[Page 25012]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.643
                [[Page 25013]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.644
                [[Page 25014]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.645
                [[Page 25015]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.646
                [[Page 25016]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.647
                [[Page 25017]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.648
                [[Page 25018]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.649
                [[Page 25019]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.650
                [[Page 25020]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.651
                [[Page 25021]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.652
                [[Page 25022]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.653
                [[Page 25023]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.654
                 The above tables summarizing estimated benefits and costs of the
                regulatory alternatives considered here exclude results of the implicit
                opportunity cost calculations discussed above and in Section
                VI.D.1.b)(8)
                [[Page 25024]]
                Implicit Opportunity Cost. The following four tables show corresponding
                benefits and costs when results of these calculations are included:
                [[Page 25025]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.655
                [[Page 25026]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.656
                [[Page 25027]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.657
                [[Page 25028]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.658
                [[Page 25029]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.659
                [[Page 25030]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.660
                [[Page 25031]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.661
                [[Page 25032]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.662
                a) Impacts on Total Fleet Size, Usage, and Safety
                [[Page 25033]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.663
                [[Page 25034]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.664
                [[Page 25035]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.665
                [[Page 25036]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.666
                [[Page 25037]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.667
                [[Page 25038]]
                BILLING CODE 4910-59-C
                (1) Total Fleet Size and VMT
                 The CAFE model carries a complete representation of the registered
                vehicle population in each calendar year, starting with an aggregated
                version of the most recent available data about the registered
                population for the first year of the simulation. In this analysis, the
                first model year considered is MY 2017, and the registered vehicle
                population enters the model as it appeared at the end of calendar year
                2016. The initial vehicle population is stratified by age (or model
                year cohort) and regulatory class--to which the CAFE model assigns
                average fuel economies based on the reported regulatory class industry
                average compliance value in each model year (and class). Once the
                simulation begins, new vehicles are added to the population from the
                market data file and age throughout their useful lives during the
                simulation, with some fraction of them being retired (or scrapped)
                along the way. For example, in calendar year 2018, the new vehicles
                (age zero) are MY 2018 vehicles (added by the CAFE model simulation and
                represented at the same level of detail used to simulate compliance),
                the age one vehicles are MY 2017 vehicles (added by the CAFE model
                simulation), and the age two vehicles are MY 2016 vehicles (inherited
                from the registered vehicle population and carried through the analysis
                with less granularity). This national registered fleet is used to
                calculate annual fuel consumption, vehicle miles traveled (VMT),
                pollutant emissions, and safety impacts under each regulatory
                alternative.
                 In support of prior CAFE rulemakings, the CAFE model accounted for
                new travel that results from fuel economy improvements that reduce the
                cost of driving. The magnitude of the increase in travel demand is
                determined by the rebound effect. In both previous versions and the
                current version of the CAFE model, the amount of travel demanded by the
                existing fleet of vehicles is also responsive to the rebound effect
                (representing the price elasticity of demand for travel)--increasing
                when fuel prices decrease relative to the fuel price when the VMT on
                which our mileage accumulation schedules were built was observed. Since
                the fuel economy of those vehicles is already fixed, only the fuel
                price influences their travel demand relative to the mileage
                accumulation schedule and so is identical for all regulatory
                alternatives.
                 While the average mileage accumulation per vehicle by age is not
                influenced by the rebound effect in a way that differs by regulatory
                alternative, three other factors influence total VMT in the model in a
                way that produces different total mileage accumulation by regulatory
                alternative. The first factor is the total industry sales response: New
                vehicles are both driven more than older vehicles and are more fuel
                efficient (thus producing more rebound miles). To the extent that more
                (or fewer) of these new models enter the vehicle fleet in each model
                year, total VMT will increase (or decrease) as a result. The second
                factor is the dynamic fleet share model. The fleet share influences not
                only the fuel economy distribution of the fleet, as light trucks are
                less efficient than passenger cars on average, but the total miles are
                influenced by fact that light trucks are driven more than passenger
                cars as well. Both of the first two factors can magnify the influence
                of the rebound effect on vehicles that go through the compliance
                simulation (MY 2017-2050) in the manner discussed above. The third
                factor influencing total annual VMT is the scrappage model. By
                modifying the retirement rates of on-road vehicles under each
                regulatory alternative, the scrappage model either increases or
                decreases the lifetime miles that accrue to vehicles in a given model
                year cohort.
                 In addition to dynamically modifying the total number of new
                vehicles sold, a dynamic model of vehicle retirement, or scrappage, has
                also been implemented. The model implements the scrappage response by
                defining the instantaneous scrappage rate at any age using two
                functions. For ages less than 30, instantaneous scrappage is defined as
                a function of vehicle age, new vehicle price, fuel prices, cost per
                mile of driving (the ratio of fuel price and fuel economy), and GDP
                growth rate. For ages greater than 30, the instantaneous scrappage rate
                is a simple exponential function of age. While the scrappage response
                does not affect manufacturer compliance calculations, it impacts the
                lifetime mileage accumulation (and thus fuel savings) of all vehicles.
                Previous CAFE analyses have focused exclusively on new vehicles,
                tracing the fuel consumption and social costs of these vehicles
                throughout their useful lives; the scrappage effect also impacts the
                registered vehicle fleet that exists when a set of standards is
                implemented.
                 For a given calendar year, the retirement rates of the registered
                vehicle population are governed by the scrappage model. To the extent
                that a given set of CAFE or CO2 standards accelerates or
                decelerates the retirement of vehicles, fuel consumption and social
                costs may change. The CAFE model accounts for those costs and benefits,
                as well as tracking all of the standard benefits and costs associated
                with the lifetimes of new vehicles produced under the rule. For more
                detail about the derivation of the scrappage functions, see Section VI.
                (2) Fuel Consumption
                 For every vehicle model in the market file, the model estimates the
                VMT per vehicle (using the assumed VMT schedule, the vehicle fuel
                economy, fuel price, and the rebound assumption). Those miles are
                multiplied by the volume for each vehicle. Fuel consumption is the
                product of miles driven and fuel economy, which can be tracked by model
                year cohort in the model. Carbon dioxide emissions from vehicle
                tailpipes are the simple product of gallons consumed and the carbon
                content of each gallon.
                 In order to calculate calendar year fuel consumption, the model
                needs to account for the inherited on-road fleet in addition to the
                model year cohorts affected by this new final rule. Using the VMT of
                the average passenger car and light truck from each cohort, the model
                computes the fuel consumption of each model year class of vehicles for
                its age in a given CY. The sum across all ages (and thus, model year
                cohorts) in a given CY provides estimated CY fuel consumption.
                 Because the model produces an estimate of the aggregate number of
                gallons sold in each CY, it is possible to calculate both the total
                expenditures on motor fuel and the total contribution to the Highway
                Trust Fund (HTF) that result from that fuel consumption. The Federal
                fuel excise tax is levied on every gallon of gasoline and diesel sold
                in the U.S., with diesel facing a higher per-gallon tax rate. The model
                uses a national perspective, where the State taxes present in the input
                files represent an estimated average fuel tax across all U.S. States.
                Accordingly, while the CAFE model cannot reasonably estimate potential
                losses to State fuel tax revenue from increasingly the fuel economy of
                new vehicles, it can do so for the HTF.
                 In addition to the tailpipe emissions of carbon dioxide, each
                gallon of gasoline produced for consumption by the on-road fleet has
                associated ``upstream'' emissions that occur in the extraction,
                transportation, refining, and distribution of the fuel. The model
                accounts for these emissions as well (on a per-gallon basis) and
                reports them accordingly.
                (3) Safety
                 Earlier versions of the CAFE model accounted for the safety impacts
                associated with reducing vehicle mass
                [[Page 25039]]
                in order to improve fuel economy. In particular, NHTSA's safety
                analysis estimated the additional fatalities that would occur as a
                result of new vehicles getting lighter, then interacting with the on-
                road vehicle population. In general, taking mass out of the heaviest
                new vehicles improved safety outcomes, while taking mass from the
                lightest new vehicles resulted in a greater number of expected highway
                fatalities. However, the change in fatalities did not adequately
                account for changes in exposure that occur as a result of increased
                demand for travel as vehicles become cheaper to operate. The current
                version of the model resolves that limitation and addresses additional
                sources of fatalities that can result from the implementation of CAFE
                or CO2 standards. These are discussed in greater detail in
                Section VI.
                 The agencies have observed that older vehicles in the population
                are responsible for a disproportionate number of fatalities, both by
                number of registrations and by number of miles driven. Accordingly, any
                factor that causes the population of vehicles to turn over more slowly
                will induce additional fatalities--as those older vehicles continue to
                be driven, rather than being retired and replaced with newer (even if
                not brand new) vehicle models. The scrappage effect, which delays (or
                accelerates) the retirement of registered vehicles, impacts the number
                of fatalities through this mechanism--importantly affecting not just
                new vehicles sold from model years 2017-2050 but existing vehicles that
                are already part of the on-road fleet. Similarly, to the extent that a
                CAFE or CO2 alternative reduces new vehicle sales, it can
                slow the transition from older vehicles to newer vehicles, reducing the
                share of total vehicle miles that are driven by newer, more
                technologically advanced vehicles. Furthermore, newer vehicles are
                equipped with technologies that make driving safer not only safer for
                the occupants of newer vehicles, but also pedestrians, cyclists, and
                even occupants of other vehicles. Accounting for the change in vehicle
                miles traveled that occurs when vehicles become cheaper to operate
                leads to a number of fatalities that can be attributed to the rebound
                effect, independent of any changes to new vehicle mass, price, or
                longevity.
                 The CAFE model estimates fatalities by combining the effects
                discussed above. In particular, the model estimates the fatality rate
                per billion miles VMT for each model year vehicle in the population
                (the newest of which are the new vehicles produced that model year).
                This estimate is independent of regulatory class and varies only by
                year (and not vehicle age). The estimated fatality rate is then
                multiplied by the estimated VMT (in billions of miles) for each vehicle
                in the population and the product of the change in curb weight and the
                relevant safety coefficient, as in the equation below.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.668
                 For the vehicles in the historical fleet, meaning all those
                vehicles that are already part of the registered vehicle population in
                CY 2017, only the model year effect that determines the
                ``FatalityEstimate'' is relevant. However, each vehicle that is
                simulated explicitly by the CAFE model, and is eligible to receive mass
                reduction technologies, must also consider the change between its curb
                weight and the threshold weights that are used to define safety
                classes. For vehicles above the threshold, reducing vehicle mass can
                have a smaller negative impact on fatalities (or even reduce
                fatalities, in the case of the heaviest light trucks). The
                ``ChangePer100Lbs'' depends upon this difference. The sum of all
                estimated fatalities for each model year vehicle in the on-road fleet
                determines the reported fatalities, which can be summarized by either
                model year or calendar year.
                b) Environmental Impacts
                 Today's final rule directly involves the fuel economy and average
                CO2 emissions of light-duty vehicles, and the final rule is
                expected directly and significantly to impact national fuel consumption
                and CO2 emissions. Fuel economy and CO2 emissions
                are closely related, so that it is expected the impacts on national
                fuel consumption and national CO2 emissions will track in
                virtual lockstep with each other.
                 Today's final rule does not directly involve pollutants such as
                carbon monoxide, smog-forming pollutants (nitrogen oxides and unburned
                hydrocarbons), fine particulate matter, or ``air toxics'' (e.g.,
                formaldehyde, acetaldehyde, benzene). While today's final rule is
                expected to impact such emissions indirectly (by reducing travel demand
                and accelerating fleet turnover to newer and cleaner vehicles on one
                hand while, on the other, increasing activity at refineries and in the
                fuel distribution system), it is expected that these impacts will be
                much smaller than impacts on fuel use and CO2 emissions
                because standards for these other pollutants are independent of those
                for CO2 emissions.
                 Following decades of successful regulation of criteria pollutants
                and air toxics, modern vehicles are already vastly cleaner than in the
                past, and it is expected that new vehicles will continue to improve.
                For example, the following chart shows trends in new vehicles' emission
                rates \2434\ for volatile organic compounds (VOCs) and nitrogen oxides
                (NOX)--the two motor vehicle criteria pollutants that
                contribute to the formation of smog.
                ---------------------------------------------------------------------------
                 \2434\ The emission rate is the rate at which a vehicle emits a
                given pollutant into the atmosphere. Tailpipe emission rates are
                expressed on a gram per mile basis. For example, driving 15,000
                miles in a year, a vehicle with a 0.4 g/mi NOX emission
                rate would emit 6,000 grams of NOX.
                ---------------------------------------------------------------------------
                [[Page 25040]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.669
                 Because new vehicles are so much cleaner than older models, it is
                expected that under any of the alternatives considered here for fuel
                economy and CO2 standards, emissions of smog-forming
                pollutants would continue to decline nearly identically over the next
                two decades. The following chart shows estimated total fuel
                consumption, CO2 emissions, and smog-forming emissions under
                the baseline and new final standards (CAFE standards--trends for
                CO2 standards would be very similar), normalized to 2017
                levels in order to allow the three to be shown together on a single
                chart:
                [[Page 25041]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.670
                 The following table summarizes relative differences between the
                baseline/augural and final standards:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.671
                 As indicated, the agencies' analysis indicates that through 2050,
                increases in annual light-duty fuel consumption and CO2
                emissions would remain below 10 percent, and increases in annual light-
                duty emissions of smog-forming pollutants would remain below 2.5
                percent.
                 As the analysis affirms, while fuel economy and CO2
                emissions are two sides (or, arguably, the same side) of the same coin,
                fuel economy and CO2 are only incidentally related to
                pollutants such as smog, and any positive or negative impacts of
                today's rulemaking on these other air quality problems would most
                likely be far too small to observe.
                [[Page 25042]]
                 The remainder of this section summarizes the impacts on fuel
                consumption and emissions for both the new final CAFE standards and the
                new final CO2 standards.
                (1) Understanding Energy and Environmental Impacts
                 Today's rulemaking and accompanying FRIA and FEIS all examine a
                range of physical impacts. These impacts reflect the combined effect of
                a range of different factors, some of which are independent of one
                another, and some of which interact. The scope and nature of this set
                of factors is such that, even among knowledgeable experts, intuition is
                often uninformative or even misleading.
                 On one hand, it is reasonable to be confident that the more CAFE
                and CO2 standards are relaxed, the more national-scale fuel
                consumption and CO2 emissions will increase, because the
                standards apply directly to the average rates at which new vehicle
                consume fuel and, in turn, emit CO2. While other factors--
                including some that work against this expectation--are involved, these
                other factors are insufficient to belie this basic expectation that
                less stringent standards will lead to increased fuel consumption and
                CO2 emissions.
                 On the other hand, while it is intuitive to expect that the
                increased fuel consumption should lead to some additional emissions to
                produce and distribute fuel, those processes are expected to become
                cleaner over time, and refineries may respond by reducing exports of
                petroleum products rather than increasing overall activity. Although
                many believe that more fuel-efficient vehicles are, by definition,
                ``cleaner,'' most pollutants impacting air quality are regulated on an
                average per-mile basis, such that vehicles' ``cleanliness'' is
                effectively independent from vehicles' fuel economy.\2435\ However,
                because emissions standards relevant to air quality are so much more
                stringent than in the past, and because some emission control
                technologies (e.g., catalytic converters) tend to deteriorate as
                vehicles age, average emission rates of vehicles are very dependent on
                when those vehicles were produced and how old they are. This means that
                total vehicular emissions of pollutants impacting air quality depend
                not directly on fuel economy, but rather on the amount of highway
                travel (since emissions are regulated on a per-mile basis) and on how
                that travel is distributed among older and newer vehicles. The agencies
                estimate that relaxing CAFE and CO2 standards will, by
                decreasing the price and fuel economy levels of vehicles produced after
                MY 2017, lead to changes in the quantities of new vehicles produced and
                sold in the U.S., as well as changes in fleet mix (i.e., the relative
                shares of passenger cars and light trucks, which are subject to
                different emissions standards), and changes in the rates at which older
                vehicles are removed from service (i.e., scrapped). Is it reasonable to
                expect that less stringent standards will necessarily accelerate the
                turnover to newer, cleaner vehicles? Does that depend on fuel prices?
                Yet another factor involves the prevalence of electric vehicles, which
                emit no air pollutants directly, but do use electricity. How might that
                electricity be generated in the future? Also, does it necessarily
                follow that less stringent CAFE and CO2 standards will
                reduce the sale of battery electric vehicles (BEVs) in the long term?
                Could less stringent standards increase long-term BEV sales if
                manufacturers are able to make early investments in BEV research and
                development, or wait for the costs of BEV systems to decline, rather
                than making larger nearer-term commitments to, say, very advanced
                engine technologies? With air quality depending on how emissions of
                various pollutants are impacted (and sometimes in different ways) by
                these factors, there is scant basis for a priori expectations regarding
                the direction, much less the magnitude of air quality impacts under the
                various regulatory alternatives.
                ---------------------------------------------------------------------------
                 \2435\ For example, in 42 U.S.C. 7521(g), the 1990 Clean Air Act
                Amendments defined specific numerical standards for passenger car
                and light truck CO, NMHC (i.e., VOC), and NOx emission rates, and
                defined them on a gram per mile basis, such that the 3-cylinder 1993
                Geo Metro and the 12-cylinder 1993 Ferrari 512 were both regulated
                to 0.4 grams per mile of NOx, even though the Metro's average fuel
                economy rating, at 47 mpg, was more than four times greater than the
                Ferrari's 11 mpg rating.
                ---------------------------------------------------------------------------
                 Although, like any other model, the CAFE model involves many
                uncertainties and does not account for every possible factor or
                interaction, the model does enable the agencies to estimate emissions
                impacts accounting for the factors mentioned above, and specific
                results can be understood through careful examination of model inputs,
                outputs, and methods. To illustrate this, the agencies consider
                estimated emissions of nitrogen oxides (NOX), a class of
                pollutants that contribute to the formation of ground-level ozone
                (i.e., smog) that is harmful to public health and welfare. The agencies
                apply the same ``unconstrained'' modeling approach as underlies the
                FEIS. Graphing estimated annual tailpipe, upstream, and combined total
                NOX emissions from passenger cars and light trucks shows
                emissions declining significantly over time, with results from the
                various action alternatives (focusing here on the least stringent,
                preferred, and most stringent alternatives, and applying the same
                vertical scale to all three charts) being virtually indistinguishable
                from the no-action alternative:
                BILLING CODE 4910-59-P
                [[Page 25043]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.672
                [[Page 25044]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.673
                [[Page 25045]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.674
                BILLING CODE 4910-59-C
                 Closer examination, though, reveals that although differences are
                very small on a relative scale, they do exhibit definitive trends.
                Reducing stringency causes total annual tailpipe NOX
                emissions to decline initially, as scrappage of older higher-emitting
                vehicles is accelerated and sales of new vehicles increase slightly
                relative to augural standards. Over time, both of these trends are
                impacted by steadily increasing fuel prices, but more important,
                reducing stringency causes the market to shift somewhat more slowly to
                electric vehicles than under the augural standards. Because electric
                vehicles emit no NOX directly, the impact on NOX
                emissions of this dampening of electric vehicle sales eventually
                outweighs the other impacts, such that by approximately 2035, less
                stringent standards begin increasing annual tailpipe NOx emissions
                rather than decreasing these emissions (relative to the augural
                standards):
                [[Page 25046]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.675
                 On the other hand, at least through 2050, less stringent standards
                show increased upstream NOX emissions. These increases
                continue to build through the late 2030s, as total fuel consumption
                under the less stringent standards continues to increase relative to
                levels under the augural standards. However, by 2040, these increases
                are steadily shrinking, due to the same delayed shift to electric
                vehicles:
                [[Page 25047]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.676
                 Model outputs indicate that on a per-mile basis, upstream
                NOX emissions beyond 2030 are 2-24 percent greater for
                electricity than for gasoline, varying over time and between regulatory
                alternatives. (Although the agencies have applied the same upstream
                emission factors to all regulatory alternatives, comparative per-mile
                upstream emissions also depend on comparative vehicle efficiency.) This
                means that, although a shift to electrification reduces tailpipe
                emissions, it also tends to increase net upstream emissions.
                 Taken together, these changes in tailpipe emissions produce very
                slight decreases in overall annual NOX emissions through
                about 2026 under each regulatory alternative. Beyond 2026, the
                regulatory action alternatives all produce increased overall annual
                NOX emissions relative to the augural standards, although
                for the most stringent regulatory alternative considered here, these
                increases plateau after about 2040:
                [[Page 25048]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.677
                 Still, although trends and differences between regulatory
                alternatives are clear on the scale of the last three of the above
                charts, the preceding three charts place these emissions changes in
                context, and show that they are barely discernable. For example, the
                largest increase shown in the last of the above charts is about 0.015
                million tons, in 2050, when total emissions are 0.33-0.35 million tons,
                down from about 1.5 million tons in 2017. In other words, the largest
                increase in overall annual NOX emissions is only about 1
                percent of recent annual NOX emissions attributable to
                passenger cars and light trucks.
                 The FEIS accompanying today's rulemaking presents tailpipe,
                upstream, and total emissions for a range of pollutants, and presents
                results of photochemical modeling to estimate corresponding changes in
                air quality, as well as results of calculations to estimate resultant
                health impacts. As indicated by the following chart, at least for the
                final standards, VOC and PM emissions follow overall trends broadly
                similar to those followed by NOX emissions, although,
                relative to recent (2017) total emissions attributable to passenger
                cars and light trucks, changes in VOC and PM emissions are not as small
                as changes in NOX emissions. Under the final standards,
                combined tailpipe and upstream CO emissions are very slightly lower
                than under the augural standards through the early 2030s, after which
                these emissions changes begin increasing at rates similar to those for
                VOC, NOX, and PM. CO2 emissions changes exhibit
                the expected trend mentioned above, with combined tailpipe and upstream
                emissions steadily increasing under the final standards. However, the
                final standards lead combined tailpipe and upstream SO2
                emissions to decrease relative to the augural standards, and as a share
                of 2017 emissions, these decreases grow from about 2 percent in 2035 to
                about 10 percent in 2050:
                [[Page 25049]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.678
                 As indicated by the following chart, changes in tailpipe
                SO2 emission follow trends nearly identical to those
                followed by changes in CO2 emissions, because both result
                directly from the quantity and composition (sulfur and carbon per
                gallon, respectively) of fuel consumed:
                [[Page 25050]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.679
                 This means that the decreases in overall SO2 emissions
                must be attributable to decreases in upstream SO2 emissions.
                The following chart shows SO2 emissions decreases becoming
                steadily larger after the mid-2030s, suggesting that, as discussed
                above, delaying the shift to electric vehicles leads to delays in
                emissions from electricity generation, and for some pollutants (notably
                below, SO2 and CO2), these emissions from
                electricity generation are large enough to reverse trends in overall
                emissions changes. For SO2, this reflects, among other
                things, the fact that, in order to enable catalytic converters to
                operate more efficiently, gasoline in sulfur is now limited to an
                average of 10 parts per million.\2436\
                ---------------------------------------------------------------------------
                 \2436\ See https://www.epa.gov/regulations-emissions-vehicles-and-engines/final-rule-control-air-pollution-motor-vehicles-tier-3.
                ---------------------------------------------------------------------------
                [[Page 25051]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.680
                 Again, the FEIS accompanying today's rulemaking further explores
                changes in emissions; the purpose of this discussion is not to
                duplicate material appearing in the FEIS, but rather to discuss some of
                the underlying factors and how they can lead to some of the trends
                reported in the FEIS.
                 Unlike the FEIS, today's rulemaking and accompanying FRIA largely
                examine impacts on a ``model year basis.'' As discussed below, while a
                calendar year basis involves considering impacts in one or a series of
                calendar years, a model year basis involves considering impacts over
                the useful lives of vehicles produced in one or over a series of model
                years. A calendar year approach answers the question ``what do we
                estimate will happen in, for example, 2035?,'' and a model year
                approach answers the question ``what impacts do we estimate will be
                attributable to vehicles produced in 2025?'' The calendar approach does
                not extend beyond 2050, the last year in which the analysis includes a
                complete on-road fleet. On the other hand, while it accounts for model
                year 2050 vehicles' fuel consumption and emissions through 2089, the
                model year approach as implemented here does not extend beyond model
                year 2029.
                 These are differences in temporal perspective that, for some types
                of impacts, lead to differences in reported trends. For example,
                returning to tailpipe NOX emissions, Figure VII-6 (using the
                calendar year perspective) shows that relaxing the stringency of CAFE
                standards leads annual tailpipe NOX emissions to increase
                starting around 2035, but leads these emissions to decrease in the
                nearer term. As discussed above, this shift can be attributed to the
                less stringent standards leading to a delayed shift toward electric
                vehicles. Because the model year perspective as implemented here
                extends through 2029, it largely sets aside this shift to electric
                vehicles, even for the ``unconstrained'' modeling underlying the FEIS
                (modeling which, unlike the ``standard setting'' type of analysis
                required by EPCA, considers that, even during 2018-2029, additional
                electric vehicles might be produced in response to standards).
                Consequently, unlike the calendar year perspective as applied beyond
                2035, the model year perspective that extends through MY 2029 always
                shows tailpipe NOX emissions decreasing as the stringency of
                CAFE standards is relaxed relative to the augural standards.
                 In addition to this difference in temporal perspective, the FEIS,
                relative to the rulemaking and FRIA, applies a perspective that is
                different in terms of how manufacturers could respond to standards. The
                ``unconstrained'' modeling underlying the FEIS allows for the potential
                that manufacturers might apply CAFE compliance credits or introduce
                additional electric vehicles in any model year. This is intended to
                reflect how manufacturers might respond to standards in the real world.
                However, EPCA requires that, for purposes of determining the maximum
                feasible standards, NHTSA set aside the potential that manufacturers
                might apply credits or increase electric vehicle offerings in the model
                years under consideration. Therefore, for CAFE, the preamble and FRIA
                use modeling that sets aside the potential use of credits
                [[Page 25052]]
                and the potential introduction of new electric vehicles through 2029
                (although, since standards prior to MY 2021 are not subject to
                reconsideration, this modeling does consider the potential use of
                credits through MY 2020). As indicated by the following chart,
                especially prior to model year 2030, this leads to significant
                differences in EV market penetration between the two types of analyses:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.681
                 Over time, these differences in EV sales lead to significant
                differences in the steadily accumulating share of overall highway
                travel powered with electricity:
                [[Page 25053]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.682
                 For most pollutants, the fact that EVs do not emit air pollutants
                outweighs the fact that combustion-based power plants do. As discussed
                above, sulfur content in gasoline is so low that the opposite is the
                case for net SO2 emissions.
                 A complete quantitative analysis of differences between calendar
                year-based emissions trends shown in the FEIS and model year-based
                emissions trends shown in the rulemaking and FRIA would involve
                examination of all of the factors mentioned above. However, considering
                the temporal difference in perspective between the two types of
                analyses, and considering the differences in the timing and pace of the
                estimated transition to electric vehicles, differences in emissions
                trends are inevitable.
                (2) CO2 Damages
                 Section V discusses, among other things, the need of the Nation to
                conserve energy, providing context for the estimated impacts on
                national-scale fuel consumption summarized below. Corresponding to
                these changes in fuel consumption, the agencies estimate that today's
                final rule will impact CO2 emissions. CO2 is one
                of several gases that absorb infrared radiation, thereby trapping heat
                and potentially making the planet warmer. The most important such gases
                directly emitted by human activities include carbon dioxide
                (CO2), methane (CH4), nitrous oxide
                (N2O), and several fluorine-containing halogenated
                substances. Although CO2, CH4, and N2O
                occur naturally in the atmosphere, human activities have changed their
                atmospheric concentrations. From the pre-industrial era (i.e., ending
                about 1750) to 2016, concentrations of these gases have increased
                globally by 44, 163, and 22%, respectively.\2437\ The FEIS accompanying
                today's rulemaking discusses the potential impacts of the emission of
                such gases at greater length, and also summaries analysis quantifying
                some of these impacts (e.g., average temperatures) for each of the
                considered regulatory alternatives.
                ---------------------------------------------------------------------------
                 \2437\ Impacts and U.S. emissions of CO2 are
                discussed at greater length in EPA's 2018 ``Inventory of U.S.
                Greenhouse Gas Emissions and Sinks,'' EPA 430-R-18-003 (Apr. 12,
                2018), available at https://www.epa.gov/sites/production/files/2018-01/documents/2018_complete_report.pdf.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 25054]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.683
                [[Page 25055]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.684
                BILLING CODE 4910-59-C
                [[Page 25056]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.685
                (3) Other Pollutant Damages--Criteria and Toxic Pollutants
                 The CAFE model uses the entire on-road fleet, calculated VMT
                (discussed above), and emissions factors (which are an input to the
                CAFE model, specified by model year and age) to calculate tailpipe
                emissions associated with a given alternative. Just as it does for
                additional CO2 emissions associated with upstream emissions
                from fuel production, the model captures criteria pollutants that occur
                during other parts of the fuel life cycle. While this is typically a
                function of the number of gallons of gasoline consumed (and miles
                driven, for tailpipe criteria pollutant emissions), the CAFE model also
                estimates electricity consumption and the associated upstream emissions
                (resource extraction and generation, based on U.S. grid mix).
                (a) Emissions Increases
                [[Page 25057]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.686
                BILLING CODE 4910-59-P
                [[Page 25058]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.687
                [[Page 25059]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.688
                [[Page 25060]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.689
                [[Page 25061]]
                BILLING CODE 4910-69-C
                (b) Air Quality Impacts of Other Pollutants
                 Although this final rule focuses on standards for fuel economy and
                CO2, it will also have an impact on criteria and air toxic
                pollutant emissions, although as discussed above, it is expected that
                incremental impacts on criteria and air toxic pollutant emissions would
                be too small to observe under any of the regulatory alternatives under
                consideration. Nevertheless, the following sections detail the criteria
                pollutant and air toxic inventory impacts of this final rule; the
                methodology used to calculate those impacts; the health and
                environmental effects associated with the criteria and toxic air
                pollutants that are being impacted by this final rule; the potential
                impact of this final rule on concentrations of criteria and air toxic
                pollutants in the ambient air; and other unquantified health and
                environmental effects.
                 Today's analysis reflects the combined result of several underlying
                impacts, all discussed above. CAFE and CO2 standards are
                estimated to impacts new vehicle prices, fuel economy levels, and
                CO2 emission rates. These changes are estimated to impact
                the size and composition of the new vehicle fleet and to impact the
                retention of older vehicles (i.e., vehicle survival and scrappage) that
                tend to have higher criteria and toxic pollutant emission rates. Along
                with the rebound effect, these lead to changes in the overall amount of
                highway travel and the distribution among different vehicles in the on-
                road fleet. Vehicular emissions depend on the overall amount of highway
                travel and the distribution of that travel among different vehicles,
                and emissions from ``upstream'' processes (e.g., petroleum refining,
                electricity generation) depend on the total consumption of different
                types of fuels for light-duty vehicles.
                (i) Impacts
                 As discussed above, in addition to affecting fuel consumption and
                emissions of carbon dioxide or its equivalent, this rule would also
                influence other pollutants, i.e., ``criteria'' air pollutants and their
                precursors, and air toxics. The final rule would affect emissions of
                carbon monoxide (CO), fine particulate matter (PM2.5),
                sulfur dioxide (SOX), volatile organic compounds (VOC),
                nitrogen oxides (NOX), benzene, 1,3-butadiene, formaldehyde,
                acetaldehyde, and acrolein. Consistent with the evaluation conducted
                for the Environmental Impact Statement accompanying today's rule, the
                agency analyzed criteria air pollutant impacts in 2025, 2035, and 2050
                (as a representation of future program impacts). Estimates of these
                other emission impacts are shown by pollutant in Table VII-124 through
                Table VII-127 and are broken down by the two drivers of these changes:
                a) ``downstream'' emission changes, reflecting the estimated effects of
                VMT rebound (discussed in Section VIII of the FRIA), changes in vehicle
                fleet age, changes in vehicle emission standards, and changes in fuel
                consumption; and b) ``upstream'' emission increases because of
                increased refining and distribution of motor vehicle gasoline relative
                to the baseline. Program impacts on criteria and toxics emissions are
                discussed below.
                 As discussed above, these changes in total annual criteria
                pollutant emissions attributable to passenger cars and light trucks
                reflect trends in both vehicular and upstream emissions, and these
                trends can either be mutually reinforcing or mutually offsetting,
                depending on the pollutant and year. Above, Figure VII-9 places these
                total changes in emissions in context, showing that, except for
                SO2, these changes in criteria pollutant emissions are very
                small. For SO2 emissions, changes are also very small
                through the late 2030s, after which reduced upstream emissions cause
                net emission reductions to exceed 10 percent of 2017 emissions by 2050.
                BILLING CODE 4910-59-P
                [[Page 25062]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.690
                [[Page 25063]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.691
                [[Page 25064]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.692
                [[Page 25065]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.693
                BILLING CODE 4910-59-C
                 As shown in Table VII-128 through Table VII-131, it is estimated
                that the new final program would result in small changes for air toxic
                emissions
                [[Page 25066]]
                compared to total U.S. inventories across all sectors. These changes
                also reflect the changing balance between vehicular and upstream
                emissions.
                BILLING CODE 4910-59-P
                [[Page 25067]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.694
                [[Page 25068]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.695
                [[Page 25069]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.696
                [[Page 25070]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.697
                [[Page 25071]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.698
                BILLING CODE 4910-59-C
                 Changes in emissions of other pollutants due to these rules will
                impact air quality. Information on current air quality and the results
                of our air quality
                [[Page 25072]]
                modeling of the projected impacts of these rules are summarized in the
                following section.
                (ii) Other Unquantified Health and Environmental Effects
                 In the proposal, the agencies sought comment on whether there are
                any other health and environmental impacts associated with advancements
                in technologies that should be considered. For example, the use of
                technologies and other strategies to reduce fuel consumption and/or
                CO2 emissions could have effects on a vehicle's life-cycle
                impacts (e.g., materials usage, manufacturing, end of life disposal),
                beyond the issues regarding fuel production and distribution (upstream)
                CO2 emissions discussed in Section VI.D.2. The agencies
                sought comment on any studies or research in this area that should be
                considered in the future to assess a fuller range of health and
                environmental impacts from the light-duty vehicle fleet shifting to
                different technologies and/or materials. At this point, the agencies
                find there is insufficient information about the lifecycle impacts of
                the myriad of available technologies, materials, and cradle-to-grave
                pathways to conduct the type of detailed assessments that would be
                needed in a regulatory context, especially considering the
                characterization of specific vehicles in the analysis fleet and the
                characterization of specific technology options.
                (c) Health Effects of Other Pollutants
                 This section presents results of the analysis showing health
                effects associated with exposure to some of the criteria and air toxic
                pollutants impacted by the new final vehicle standards. As discussed
                above, the health impacts presented here are subject to a number of
                uncertainties, some of which arise from the less complex benefits-per-
                ton approach relied on in this analysis, and some of which arise from
                the uncertainty surrounding many of the assumptions and other inputs
                relied on in the agencies' analysis. As the agencies conclude above,
                although it may seem that the agencies' estimates of increases in
                premature mortality resulting from the final standards are more likely
                to be too high than too low, it is extremely difficult to anticipate
                whether this is actually the case.
                BILLING CODE 4910-59-P
                [[Page 25073]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.699
                [[Page 25074]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.700
                [[Page 25075]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.701
                [[Page 25076]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.702
                [[Page 25077]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.703
                [[Page 25078]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.704
                [[Page 25079]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.705
                [[Page 25080]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.706
                [[Page 25081]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.707
                [[Page 25082]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.708
                [[Page 25083]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.709
                [[Page 25084]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.710
                [[Page 25085]]
                BILLING CODE 4910-59-C
                B. Impacts on Calendar Year Basis
                 As with the NPRM, the agencies' analysis primarily examines
                regulatory impacts on a model year basis, accounting for the physical
                impacts and monetized costs and benefits attributable to vehicles
                produced prior to model year 2030 and occurring throughout these
                vehicles' useful lives. EDF submitted comments arguing that the
                agencies should examine impacts on a calendar year basis, as discussed
                above in VI.A.\2438\ CAFE analysis has historically examined effects of
                the standards on a model year basis, because CAFE (and CO2)
                standards are enforced on a model year basis, and manufacturers'
                responses to these standards (i.e., their costs), which are the direct
                effects of the standards, occur on a model year basis. On the other
                hand, overall impacts on national energy consumption and the
                environment result from the evolution and operation of the overall on-
                road fleet, and this motivates consideration of results on a calendar
                year basis. As also discussed in VI.A., the agencies have expanded the
                presentation of results in today's rulemaking and FRIA by presenting
                some impacts for each of CYs 2017-2050 and, to enable doing so, have
                extended the analysis to cover model years through 2050.
                ---------------------------------------------------------------------------
                 \2438\ EDF, NHTSA-2018-0067-12108, Appendix A at 9, et seq., and
                Appendix B at 11-14.
                ---------------------------------------------------------------------------
                 For this analysis, the CAFE model reports impacts for each model
                year through 2050, and, to capture the entire useful lives of these
                vehicles, for each of calendar years 2017-2089.\2439\ One way to
                illustrate the model's outputs is to consider three cohorts of model
                years: MYs 1978-2017 (MYs to which the analysis applies no additional
                fuel-saving technology), MYs 2018-2029 (MYs included in both the ``MY
                basis'' and ``CY basis'' approaches), and MYs 2030-2050 (MYs included
                only the ``CY basis'' approach). On a calendar year basis, impacts of
                the final standards on annual CO2 emissions (impacts on fuel
                consumption would follow essentially the same trends) may be attributed
                to these cohorts as follows:
                ---------------------------------------------------------------------------
                 \2439\ As for the NPRM, DOT has made the model and all inputs
                and outputs for today's analysis available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system. The model documentation available at the same location
                explains, among other things, the structure and contents of each
                type of input and output file. The
                ``annual_societal_effects_report.csv ``and
                ``annual_societal_costs_report.csv'' reports contain, respectively,
                estimates of physical impacts and monetized costs and benefits
                attributable to each model year in each calendar years. Other output
                file types contain corresponding aggregations either all calendar
                years, or across all model years.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.711
                BILLING CODE 4910-59-P
                 Here, the large lower area of the chart shows annual CO2
                emissions estimated to occur under the baseline/augural CAFE standards,
                through calendar year
                [[Page 25086]]
                2089, which is the last year any MY 2050 vehicles are estimated still
                to be on the road. The steady declines through 2050 reflect turnover to
                more efficient vehicles produced under either regulatory alternative,
                and the steep decline after 2050 reflects vehicles included in the
                analysis being removed from service. Of the increased annual emissions
                under the final standards, the black area shows the portion
                attributable to vehicles produced during MYs 2018-2029, and the topmost
                area shows the portion attributable to vehicles produced during MYs
                2030-2050. The final standards are estimated to reduce emissions from
                vehicles produced during MYs 1978-2017 by accelerating scrappage of
                these vehicles, but these changes are too small to be visible in this
                chart.
                 The bulk of the reporting of results here and in the FRIA examines
                impacts over the useful lives of vehicles produced prior to MY 2030. In
                terms of the above chart, this means excluding the topmost area,
                producing the following:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.712
                 On the other hand, calendar year accounting, as considered for this
                analysis, includes all model years included in the analysis (i.e.,
                through MY 2050), and examines impacts in all calendar years for which
                a full on-road fleet is simulated. In terms of the first of the above
                charts, this means ``cutting off'' results at calendar year 2050:
                [[Page 25087]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.713
                 Here, the horizontal axis extends through 2089 to make clear that
                this calendar year accounting involves excluding emissions impacts over
                most of the useful lives of the latest model years included in the
                analysis. On a scale covering just those calendar years included in the
                calendar year analysis, the same chart appears as follows:
                [[Page 25088]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.714
                 Viewed on the same calendar year basis, technology costs appear as
                follows, with differences between costs under the baseline/augural
                standards and under the final standards shown as amounts by which the
                former exceed the latter (e.g., in 2025, the final standards are
                estimated to avoid about $19 billion in technology costs that would
                have been incurred under the baseline/augural standards):
                [[Page 25089]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.715
                 Present value analysis considered involves discounting all
                estimated future costs and benefits to 2019. At a 7 percent discount
                rate, the undiscounted technology costs shown above correspond to
                discounted costs shown in the following chart:
                [[Page 25090]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.716
                 Without discounting, therefore, the final standards avoid $457
                billion in technology costs through 2050, each additional year of
                analysis after 2036 adding about $14 billion to that total. At a 7
                percent discount rate, the final standards still avoid $183 billion in
                technology costs, while incremental amounts attributable to each
                additional year of analysis are (of course) lower than the undiscounted
                amounts--declining to about $5 billion during 2035-2036 and, by 2045,
                about $2 billion.
                 For each of the regulatory alternatives considered here, the
                following tables summarize results of such aggregations for each
                reported category of monetized costs and benefits. The first three
                tables focus on the final CAFE standards, presenting total amounts
                through 2050 at 3 percent and 7 percent discount rates. The second
                three tables show results for corresponding CO2 standards.
                [[Page 25091]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.717
                [[Page 25092]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.718
                [[Page 25093]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.719
                [[Page 25094]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.720
                [[Page 25095]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.721
                [[Page 25096]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.722
                [[Page 25097]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.723
                [[Page 25098]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.724
                 As illustrated above, the model year analysis answers the question
                ``what impacts do we think might eventually be attributable to vehicles
                produced before 2030?,'' and the calendar year analysis answers the
                question ``what do
                [[Page 25099]]
                we think might happen between now and 2050?'' Again, CAFE and
                CO2 standards are enforced on a model year basis, and the
                agencies accordingly simulate manufacturers' responses to these
                standards--and estimate manufacturers' corresponding costs--on a model
                year basis. This motivates consideration of results on a model year
                basis. On the other hand, overall impacts on national energy
                consumption and the environment result from the evolution and operation
                of the overall on-road fleet, and this motivates consideration of
                results on a calendar year basis.
                 These different perspectives produce results that, without careful
                consideration, appear to conflict. The model year perspective as
                applied through MY 2029 shows less stringent standards producing
                environmental benefits (compared to the augural standards) attributable
                to the aggregate of vehicles produced prior to MY 2030. While the
                calendar year perspective also shows similar trends prior to (calendar
                year) 2035, with the estimated transition to electric vehicles
                accelerating over time, the calendar year perspective shows less
                stringent standards mostly increasing emissions (SO2 being
                an exception) relative to the augural standards.
                 Still, some important aspects of estimated social benefits and
                costs are common to both the model year and calendar year perspectives.
                For each of the regulatory action alternatives, the magnitude of total
                incremental benefits (relative to the baseline augural standards) is
                similar to the magnitude of total incremental costs. This stands in
                marked contrast to the agencies' 2012 rulemaking announcing the augural
                standards, and finding of estimated benefits that were 3-4 times larger
                than costs.\2440\ Under today's analysis, estimated benefits and costs
                are instead of similar magnitude, with estimated net benefits, by
                comparison, small enough to be even directionally uncertain, such that
                an alternative estimated to produce small positive net benefits under
                one perspective and applying a 7 percent discount rate might be
                estimated to produce small negative net benefits under the other
                perspective and/or applying a 3 percent discount rate. While the
                agencies obviously must consider benefits, costs, and net benefits, our
                decisions are based on wider considerations. Consistent with the
                agencies' 2012 final rule, today's final rule finds--from both the
                model year and calendar year perspectives--that forgone fuel savings
                (forgone because today's final rule involves relaxing rather than
                increasing the stringency of CAFE and CO2 standards) account
                for the bulk of estimated forgone social benefits. These are private
                benefits, which raises a significant question of whether there is a
                meaningful market failure that needs to be addressed by more stringent
                regulation.
                ---------------------------------------------------------------------------
                 \2440\ 77 FR at 62629 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 Section VI contains an extensive discussion and analysis of the
                existence and nature of various market failures related to fuel economy
                standards. These potential market failures include the well-established
                externalities of environmentally harmful emissions, congestion, and
                safety; as well the debatable and hypothetical market failures related
                to the ``energy paradox.'' The energy paradox refers to an observation
                that some consumers appear voluntarily to forgo investments in energy
                conservation even when those initial investments appear to repay
                themselves--in the form of savings in energy costs--over the relatively
                near term. Section VI.D.1 discussion casts doubt on the theoretical
                underpinnings that the energy paradox represents a market failure,
                discusses recent research that suggests the extent consumers are
                undervaluing fuel economy has been overstated, and suggests the
                analysis supporting claims of an energy paradox overlooks the
                opportunity costs of other vehicle attributes that consumers and
                manufacturers trade off with fuel efficiency technology. As stated in
                Section VI, while the agencies have reservations about the extent to
                which a market failure capable of driving very large net private
                financial harm to consumers exists, the agencies do not take a position
                on the existence of an energy paradox in this rulemaking.
                 The primary analysis shows that the CAFE final rule would generate
                $12.9 billion in total social net benefits using a 7 percent discount
                rate, but without the large net private loss of $26.4 billion, the net
                social benefits would equal the external net benefits, or $39.3
                billion. Therefore, given significant questions about whether
                government action to impose restrictions in private markets could
                improve net social benefits absent a market failure, if no market
                failure exists to motivate the $26.4 billion in private losses to
                consumers, the net benefits of these final standards would be $39.3
                billion. The CY analysis produces similar results, though the estimated
                private losses are exacerbated relative to the external gains. The CY
                analysis shows the CAFE final rule would generate -$6 billion in total
                net social benefits using a 7 percent discount rate, but without the
                large net private loss of $65 billion, the net social benefits would
                equal the external net benefits of $59 billion.
                 One commenter suggested that the agencies should elect to use CY
                accounting in the primary analysis because the MY accounting approach
                resulted in an inconsistent accounting of costs and benefits owing to
                the scrappage effect. While the CY accounting approach does reduce non-
                rebound safety benefits from $9 billion to $8 billion (combined fatal
                and non-fatal benefits), the total external net benefits of the rule
                actually increase by $20 billion using the CY approach. This result is
                driven primarily by a significant increase in congestion cost savings
                from less rebound driving, from $44 billion to $69 billion. Any changes
                in the net benefits in the opposite direction using CY accounting
                result from increased net private costs to consumers own financial
                wellbeing from allowing more consumer choice. These increased net
                private costs occur because the CY analysis captures model years far
                into the future, which are more uncertain and not subject to today's
                CAFE final rule. Therefore, the agencies see little evidence that the
                inconsistency suggested by the commenter is important, or that the
                primary conclusions of the analysis are meaningfully influenced by it.
                Sensitivity Analysis
                 As discussed at the beginning of this section, results presented
                today reflect the agencies' best judgments regarding many different
                factors. Based on analyses in past rulemakings, the agencies recognize
                that some analytical inputs are especially uncertain, some are likely
                to exert considerable influence over specific types of estimated
                impacts, and some are likely to do so for the bulk of the analysis. To
                explore the sensitivity of estimated impacts to changes in model
                inputs, analysis was conducted using alternative values for a range of
                different inputs. Results of this sensitivity analysis are summarized
                in the Final Regulatory Impact Analysis (FRIA) accompanying today's
                rulemaking, and detailed model inputs and outputs are available on
                NHTSA's website.\2441\ The following table lists the cases included in
                the sensitivity analysis.
                ---------------------------------------------------------------------------
                 \2441\ The CAFE model and all inputs and outputs supporting
                today's rulemaking are available at https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 25100]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.725
                [[Page 25101]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.726
                [[Page 25102]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.727
                BILLING CODE 4910-59-C
                VIII. How do the final standards fulfill the agencies' statutory
                obligations?
                A. How Does the technical assessment support the final CO2
                standards as compared to the alternatives that EPA has considered?
                1. Introduction
                 Title II of the Clean Air Act provides for comprehensive regulation
                of mobile sources, authorizing EPA to regulate emissions of air
                pollutants from all mobile source categories. Under Section 202(a) and
                relevant case law, as discussed below, EPA considers such issues as
                technology emission reduction effectiveness, its cost (both per
                vehicle, per manufacturer, and per consumer), the lead time necessary
                to implement the technology, and based on this the feasibility of
                potential standards; the impacts of potential standards on emissions
                reductions of both GHGs and non-GHGs; the impacts of standards on oil
                conservation and energy security; the impacts of standards on fuel
                savings by consumers; the impacts of standards on the auto industry;
                other energy impacts; as well as other relevant factors such as impacts
                on safety.
                 EPA is afforded considerable discretion under section 202(a) when
                assessing issues of technical feasibility and availability of lead time
                and in weighing these factors. In light of its consideration of the
                relevant factors, EPA has concluded, for the reasons discussed below,
                that the previous standards (which increase stringency at a rate of
                about 5% per year) are not appropriate, and the best action is to
                revise the standards to increase stringency by 1.5% per year. Beginning
                in 2009, EPA and NHTSA have worked together jointly to establish fuel
                economy and tailpipe CO2 emission standards for light duty
                vehicles. The first rulemaking, finalized in 2010, established
                standards for the 2012 through 2016 model years. Shortly thereafter, in
                2012, the agencies established standards for the 2017 through 2025
                model years--but given the limitation in EPCA that only allows for
                standards to be set five years at a time, the 2022-2025 model year
                standards were only final for EPA's tailpipe CO2 emissions
                regulation. This rapid period of rulemaking to establish standards over
                a decade in advance may have marked a departure for NHTSA, but it
                followed EPA's longstanding
                [[Page 25103]]
                approach when regulating vehicular criteria pollutant emissions to
                provide a significant period of time for the industry to develop
                technologies to achieve standards.
                 While EPA had decades of experience regulating light duty vehicle
                emissions, it did not previously have experience regulating tailpipe
                CO2 emissions. And regulating CO2 emissions is
                quite different from regulating criteria pollutant emissions. With
                criteria pollutants, technological emission controls exist primarily in
                the form of engine controls and catalytic conversion. Today's emission
                controls for criteria pollutants have only a de minimis effect on
                performance or functionality of the vehicle.
                 Controlling tailpipe CO2 emissions for an internal
                combustion engine requires controlling the amount of energy used to
                propel the vehicle. All else being equal, better performance (in
                acceleration or passing speed) requires more energy. Similarly,
                vehicles with more storage capacity tend to be larger, and moving an
                object with larger mass requires more energy than objects with smaller
                mass. Vehicles with greater towing performance likewise require more
                energy. Maintaining utility and performance requires sophisticated and
                expensive technological solutions, such as reducing mass through
                advanced materials, changing engine combustion cycles, increasing
                compression ratios, or turbo-charging the engine. Consumers often can
                feel the difference in vehicle performance as a result of these
                controls, and as will be discussed herein.
                 As discussed when issuing the 2012 Final Rule, the economic and
                market assumptions underlying the standards the agencies finalized were
                crucial, and long-term projections are inherently uncertain. Upon
                review of those assumptions, such as the price of gas and the sales mix
                of pick-up trucks and sport-utility vehicles as compared to passenger
                cars, the agencies have now concluded that many of these assumptions
                have not proven to be accurate and therefore have been updated. Given
                the uncertainty about the 2012 assumptions at the time of that
                rulemaking, the agencies incorporated a mid-term evaluation process for
                EPA's 2022-2025 model year standards that would be ``collaborative,
                robust and transparent,'' and ``based on information available at the
                time of the mid-term evaluation and an updated assessment of all the
                factors considered in setting the standards and the impacts of those
                factors on the manufacturers' ability to comply.'' \2442\
                ---------------------------------------------------------------------------
                 \2442\ 77 FR at 62633.
                ---------------------------------------------------------------------------
                 While that process was expected to take place throughout 2017, and
                a final determination issued in the Spring of 2018, this process was
                expedited. On July 27, 2016, the agencies published a Federal Register
                notice making the public aware of the availability of a draft Technical
                Assessment Report, with comments due at the end of September 2016. On
                December 6, 2016, EPA published a notice in the Federal Register making
                the public aware of its proposed Final Determination and extensive
                Technical Support Document to keep the standards set in 2012 in place
                through the 2025 model year without change. The public was given until
                December 30, 2016 to comment on the proposed determination. Less than
                two weeks later, on January 12, 2017, EPA finalized its determination.
                 Industry commenters stated that the 2017 Final Determination ``is
                the product of egregious procedural and substantive defects and EPA
                should withdraw it,'' that EPA had ``fail[ed] to provide an adequate
                period for meaningful notice and comment,'' that EPA had
                ``acknowledg[ed] that the Proposed Determination adjusted a number of
                EPA assumptions in response to commenters who pointed out errors at
                earlier stages'' while stating that ``there was no need for more time
                because [it] did not include much new material,'' and that ``EPA [had]
                underestimated the burden [of the standards],'' ``EPA [made] cursory
                assertions that downplayed the impact of its mandate on auto sales and
                employment,'' and ``EPA refused to consider many of the [industry's]
                technical concerns even when supported by an outside consultant,
                asserted [industry] provided insufficient data, and then refused
                further meetings for clarification.'' \2443\
                ---------------------------------------------------------------------------
                 \2443\ Alliance letter to Administrator Pruitt, Feb. 21, 2017,
                available at https://autoalliance.org/wp-content/uploads/2017/02/Letter-to-EPA-Admin.-Pruitt-Feb.-21-2016-Signed.pdf.
                ---------------------------------------------------------------------------
                 In light of commenters' concerns about EPA's 2017 final
                determination, in March 2017, EPA announced its intent to reconsider
                the final determination in order to allow additional opportunity to
                hear from the public, and additional consultation and coordination with
                NHTSA in support of a national harmonized program. In August 2017, EPA
                published a notice in the Federal Register requesting comment on its
                reconsideration of the initial determination, and held a public hearing
                on the matter in September 2017. Then, in April 2018, EPA issued a
                revised final determination finding that the 2022-2025 model year GHG
                standards set in 2012 were not appropriate and a rulemaking should be
                initiated to revise the standards, as appropriate.
                 In this proceeding, in order to determine what standards are
                appropriate, EPA and NHTSA sought comment on a wide range of potential
                standards--ranging from holding the 2020 standards flat through the
                2026 model year to retaining the standards finalized in 2012. Similar
                to the 2012 rulemaking, EPA considered a number of different
                alternatives--ranging from the standards finalized in 2012, to holding
                the 2020 MY standards flat through MY 2026. As in 2012, the manner in
                which different factors are weighed can yield very different result--
                more stringent standards would improve CO2 emissions, reduce
                energy consumption, and save consumers fuel. Less stringent standards
                would reduce technology costs for manufacturers and save consumers in
                upfront purchase prices, enabling the fleet to turnover more quickly.
                While weighing these factors, EPA has considered compliance results
                that have been observed throughout the fleet. While the agencies have
                seen extraordinary reductions in tailpipe CO2 emissions
                since EPA has begun regulation in this area, manufacturers are
                increasingly falling short of meeting their performance targets, and
                are increasingly using acquired or earned credits to comply with
                requirements. For the 2016 model year, the overall fleet failed, for
                the first time in regulation history, to meet emission targets--
                achieving 272 grams per mile, when the standard was 263 grams per
                mile.\2444\ The 2016 model year saw only five major manufacturers
                perform at or better than their CO2 footprint standards--
                Honda, Hyundai, Mazda, Nissan, and Subaru. For the 2017 model year,
                only three major manufacturers--BMW, Honda, and Subaru--performed
                better than their CO2 standards, and the total fleet
                underperformed compared to the standards--achieving 263 grams per mile,
                when the fleetwide standard was 258 grams per mile.\2445\ The emissions
                averaging, credit banking and trading system was established to allow
                [[Page 25104]]
                manufacturers greater flexibility and lead time to address technical
                feasibility and cost without sacrificing effectiveness of the
                standards, but widespread reliance upon credits across the industry may
                raise concerns about compliance in future years, particularly since the
                more significant increases in stringency in the 2012 rulemaking have
                yet to be effective. Taken together, the agencies now believe this
                information supports the conclusion that the lead time EPA estimated
                would be sufficient to achieve compliance with the previous standards
                for MYs 2021-26, was not sufficient.
                ---------------------------------------------------------------------------
                 \2444\ EPA Greenhouse Gas Emission Standards for Light-Duty
                Vehicles: Manufacturer Performance Report for the 2016 Model Year.
                EPA-420-R-18-002 (January 2018).
                 \2445\ 2018 EPA Automotive Trends Report: Greenhouse Gas
                Emissions, Fuel Economy, and Technology since 1975, available at:
                https://www.epa.gov/automotive-trends/download-automotive-trends-report.
                ---------------------------------------------------------------------------
                 In this action, EPA is reducing the rate of stringency increases
                from those adopted in the 2012 rulemaking in part to ensure that the
                standards remain reasonable and appropriate. As in 2012, EPA is
                deciding against selecting alternatives that are more stringent or less
                stringent than appropriate. The final rule analysis projects that the
                1.5 percent alternative would result in less significant shortfalls
                compared to more stringent alternatives, which will ease compliance
                burdens while nonetheless pushing the market beyond what it would
                demand in the absence of standards or what would be achieved with less
                stringent standards. The standards finalized today will result in
                continuing improvements compared to the 2020 model year, and are best
                viewed in the context of the larger rulemaking, as shown in the chart
                below:
                BILLING CODE 4910-59-P
                [GRAPHIC] [TIFF OMITTED] TR30AP20.728
                BILLING CODE 4910-59-C
                2. Basis for the CO2 Standards Under Section 202(a) of the
                Clean Air Act
                 Title II of the Clean Air Act (CAA) provides for comprehensive
                regulation of mobile sources, authorizing EPA to regulate emissions of
                air pollutants from all mobile source categories. This rule implements
                a specific provision from Title II, section 202(a).\2446\ Section
                202(a)(1) states that ``[t]he Administrator shall by regulation
                prescribe (and from time to time revise) . . . standards applicable to
                the emission of any air pollutant from any class or classes of new
                motor vehicles or new motor vehicle engines, which in his judgment
                cause, or contribute to, air pollution which may reasonably be
                anticipated to endanger public health or welfare.'' If EPA makes the
                appropriate endangerment and cause or contribute findings, then section
                202(a) directs EPA to issue standards applicable to emissions of those
                pollutants.
                ---------------------------------------------------------------------------
                 \2446\ 42 U.S.C. 7521(a).
                ---------------------------------------------------------------------------
                [[Page 25105]]
                 Any standards under CAA section 202(a)(1) ``shall be applicable to
                such vehicles and engines for their useful life.'' Emission standards
                set by the EPA under section 202(a)(1) are technology-based, as the
                levels chosen must be premised on a finding of technological
                feasibility. Thus, standards promulgated under section 202(a) are to
                take effect only after ``such period as the Administrator finds
                necessary to permit the development and application of the requisite
                technology, giving appropriate consideration to the cost of compliance
                within such period.'' \2447\ EPA must consider costs to those entities
                which are directly subject to the standards.\2448\ Thus, ``the
                [s]ection 202(a)(2) reference to compliance costs encompasses only the
                cost to the motor-vehicle industry to come into compliance with the new
                emission standards.'' \2449\ EPA is afforded considerable discretion
                under section 202(a) when assessing issues of technical feasibility and
                availability of lead time to implement new technology. Such
                determinations are ``subject to the restraints of reasonableness,''
                which ``does not open the door to `crystal ball' inquiry.'' \2450\ In
                developing such technology-based standards, EPA has the discretion to
                consider different standards for appropriate groupings of vehicles
                (``class or classes of new motor vehicles''), or a single standard for
                a larger grouping of motor vehicles.\2451\
                ---------------------------------------------------------------------------
                 \2447\ CAA section 202 (a)(2); see also NRDC v. EPA, 655 F.2d
                318, 322 (DC Cir. 1981).
                 \2448\ Motor & Equipment Mfrs. Ass'n Inc. v. EPA, 627 F. 2d
                1095, 1118 (DC Cir. 1979).
                 \2449\ Coalition for Responsible Regulation, 684 F.3d at 128;
                see also id. at 126-27 (rejecting arguments that EPA was required to
                consider or should have considered costs to other entities, such as
                stationary sources, which are not directly subject to the emission
                standards).
                 \2450\ NRDC, 655 F.2d at 328 (quoting International Harvester
                Co. v. Ruckelshaus, 478 F.2d 615, 629 (DC Cir. 1973)).
                 \2451\ NRDC, 655 F.2d at 338.
                ---------------------------------------------------------------------------
                 Although standards under CAA section 202(a)(1) are technology-
                based, they are not based exclusively on technological capability. EPA
                has the discretion, and in some instances has been specifically
                directed by Congress, to consider and weigh various factors along with
                technological feasibility, such as the cost of compliance, \2452\ lead
                time necessary for compliance, \2453\ safety,\2454\ other impacts on
                consumers,\2455\ and energy impacts associated with use of the
                technology.\2456\
                ---------------------------------------------------------------------------
                 \2452\ See section 202(a)(2).
                 \2453\ Id.
                 \2454\ See NRDC, 655 F.2d at 336 n. 31.
                 \2455\ Since its earliest Title II regulations, EPA has
                considered the safety of pollution control technologies. See 45 FR
                14496, 14503 (March 5, 1980). (``EPA would not require a particulate
                control technology that was known to involve serious safety
                problems. If during the development of the trap-oxidizer safety
                problems are discovered, EPA would reconsider the control
                requirements implemented by this rulemaking.'').
                 \2456\ See George E. Warren Corp. v. EPA, 159 F.3d 616, 623-624
                (DC Cir. 1998) (ordinarily permissible for EPA to consider factors
                not specifically enumerated in the CAA).
                ---------------------------------------------------------------------------
                 Unlike standards set under provisions such as section 202(a)(3) and
                section 213(a)(3), EPA is not required to set technology-forcing
                standards when such standards would not be appropriate. EPA has
                interpreted a similar statutory provision, CAA section 231,\2457\ as
                follows:
                 \2457\ Section 231(a)(2)(A) of the CAA provides: ``The
                Administrator shall, from time to time, issue proposed emission
                standards applicable to the emission of any air pollutant from any
                class or classes of aircraft engines which in his judgment causes,
                or contributes to, air pollution which may reasonably be anticipated
                to endanger public health or welfare.'' Section 231(a)(3) provides
                in part: ``Within 90 days after the issuance of such proposed
                regulations, he shall issue such regulations with such modifications
                as he deems appropriate. Such regulations may be revised from time
                to time.'' Sectiion 231(b) provides: ``Any regulation prescribed
                under this section (and any revision thereof) shall take effect
                after such period as the Administrator finds necessary (after
                consultation with the Secretary of Transportation) to permit the
                development and application of the requisite technology, giving
                appropriate consideration to the cost of compliance within such
                period.''
                ---------------------------------------------------------------------------
                 While the statutory language of section 231 is not identical to
                other provisions in title II of the CAA that direct EPA to establish
                technology-based standards for various types of engines, EPA
                interprets its authority under section 231 to be somewhat similar to
                those provisions that require us to identify a reasonable balance of
                specified emissions reduction, cost, safety, noise, and other
                factors. See, e.g., Husqvarna AB v. EPA, 254 F.3d 195 (D.C. Cir.
                2001) (upholding EPA's promulgation of technology-based standards
                for small non-road engines under section 213(a)(3) of the CAA).
                However, EPA is not compelled under section 231 to obtain the
                ``greatest degree of emission reduction achievable'' as per sections
                213 and 202 of the CAA, and so EPA does not interpret the Act as
                requiring the agency to give subordinate status to factors such as
                cost, safety, and noise in determining what standards are reasonable
                for aircraft engines. Rather, EPA has greater flexibility under
                section 231 in determining what standard is most reasonable for
                aircraft engines, and is not required to achieve a ``technology
                forcing'' result.\2458\
                ---------------------------------------------------------------------------
                 \2458\ 70 FR 69664, 69676 (Nov. 17, 2005).
                 This interpretation was upheld as reasonable in NACAA v. EPA.\2459\
                CAA section 202(a), as with section 231, does not specify the degree of
                weight to apply to each factor, and EPA accordingly interprets its
                authority under section 202(a) similarly to its interpretation of
                section 231 as set forth above: EPA has discretion in choosing an
                appropriate balance among the statutory factors.\2460\
                ---------------------------------------------------------------------------
                 \2459\ 489 F.3d 1221, 1230 (DC Cir. 2007).
                 \2460\ See Sierra Club v. EPA, 325 F.3d 374, 378 (D.C. Cir.
                2003) (even where a provision is technology-forcing, the provision
                ``does not resolve how the Administrator should weigh all [the
                statutory] factors in the process of finding the 'greatest emission
                reduction achievable'''); see also Husqvarna AB v. EPA, 254 F. 3d
                195, 200 (D.C. Cir. 2001) (great discretion to balance statutory
                factors in considering level of technology-based standard, and
                statutory requirement ``[to give] appropriate consideration to the
                cost of applying . . . technology'' does not mandate a specific
                method of cost analysis); Hercules Inc. v. EPA, 598 F. 2d 91, 106-07
                (D.C. Cir. 1978) (``In reviewing a numerical standard, we must ask
                whether the agency's numbers are within a `zone of reasonableness,'
                not whether its numbers are precisely right''); Permian Basin Area
                Rate Cases, 390 U.S. 747, 797 (1968) (same); Federal Power
                Commission v. Conway Corp., 426 U.S. 271, 278 (1976) (same); Exxon
                Mobil Gas Marketing Co. v. FERC, 297 F. 3d 1071, 1084 (D.C. Cir.
                2002) (same).
                ---------------------------------------------------------------------------
                 As noted above, EPA has found that the elevated concentrations of
                greenhouse gases in the atmosphere may reasonably be anticipated to
                endanger public health and welfare.\2461\ EPA defined the ``air
                pollution'' referred to in CAA section 202(a) to be the combined mix of
                six long-lived and directly emitted GHGs: carbon dioxide
                (CO2), methane (CH4), nitrous oxide
                (N2O), hydrofluorocarbons (HFCs), perfluorocarbons (PFCs),
                and sulfur hexafluoride (SF6). The EPA further found under
                CAA section 202(a) that emissions of the single air pollutant defined
                as the aggregate group of these same six greenhouse gases from new
                motor vehicles and new motor vehicle engines contribute to air
                pollution. As a result of these findings, section 202(a) requires EPA
                to issue standards applicable to emissions of that air pollutant. New
                motor vehicles and engines emit CO2, CH4,
                N2O, and HFC. EPA has established standards and other
                provisions that control motor vehicle emissions of CO2,
                HFCs, N2O, and CH4. EPA has not set any standards
                for PFCs or SF6 as they are not emitted by motor vehicles.
                ---------------------------------------------------------------------------
                 \2461\ 74 FR 66496 (Dec. 15, 2009).
                ---------------------------------------------------------------------------
                3. EPA's Conclusion That the Final CO2 Standards Are
                Appropriate and Reasonable
                 In this section, EPA discusses the factors, data and analysis the
                Administrator has considered in the selection of the EPA's revised
                CO2 emission standards for MYs 2021 and later and the
                comments received on EPA's consideration of these factors (see further
                discussion below on EPA's summary and analysis of comments).
                 As discussed in Section VIII.A.1 above, the primary purpose of
                Title II of the Clean Air Act is the protection of public health and
                welfare, and GHG
                [[Page 25106]]
                emissions from light-duty vehicles have been found by EPA to endanger
                public health and welfare.\2462\ The goal of the light-duty vehicle GHG
                standards is to reduce these emissions which cause or contribute to air
                pollution which may reasonably be anticipated to endanger public health
                or welfare, while taking into account other factors as discussed above.
                ---------------------------------------------------------------------------
                 \2462\ Id.
                ---------------------------------------------------------------------------
                 CAA section 202(a)(2) states when setting emission standards for
                new motor vehicles, the standards ``shall take effect after such period
                as the Administrator finds necessary to permit the development and
                application of the requisite technology, giving appropriate
                consideration to the cost of compliance within such period.'' 42 U.S.C.
                7521(a)(2). That is, when establishing emission standards, the
                Administrator must consider both the lead time necessary for the
                development of technology that can be used to achieve the emission
                standards and the resulting costs of compliance on those entities that
                are directly subject to the standards. In previous rulemakings,
                including the rulemaking that established the current standards, EPA
                considered lead time-related elements, including comparative per-
                vehicle cost increases by manufacturer for both cars and trucks,
                comparative penetration rates of advanced technologies by manufacturers
                for both cars and trucks, and lead time concerns about increasing
                technology penetration rates for these advanced technologies beyond
                current levels. EPA also considered comparative industry-wide costs and
                differences between alternatives, framed in terms of total costs and
                percentage differences between alternatives. These elements are
                discussed in detail throughout the analysis. As mentioned previously,
                however, the performance of the fleet in recent years indicates that
                the lead time deemed as adequate in the 2012 rulemaking was not
                sufficient.
                 EPA is not limited to consideration of the factors specified in CAA
                section 202(a)(2) when establishing standards for light-duty vehicles.
                In addition to feasibility and cost of compliance, EPA may (and
                historically has) considered such factors as safety, energy use and
                security, degree of reduction of both GHG and non-GHG pollutants,
                technology cost-effectiveness, and costs and other impacts on
                consumers.
                 EPA also considers relevant case law. Critical to this series of
                joint rulemakings with NHTSA, the Court in Massachusetts v. EPA,\2463\
                recognized EPA's argument that ``it cannot regulate carbon dioxide
                emissions from motor vehicles'' without ``tighten[ing] mileage
                standards . . . .''--a task assigned to DOT. The Court found that
                ``[t]he two obligations may overlap, but there is no reason to think
                the two agencies cannot both administer their obligations and yet avoid
                inconsistency.'' \2464\ Accordingly, the agencies have worked closely
                together in setting standards, and many of the factors that NHTSA
                considers to set maximum feasible standards overlap with factors that
                EPA considers under the Clean Air Act. Just as EPA considers energy use
                and security, NHTSA considers these factors when evaluating the need of
                the nation to conserve energy, as required by EPCA. Just as EPA
                considers technological feasibility, the cost of compliance,
                technological cost-effectiveness and cost and other impacts upon
                consumers, NHTSA considers these factors when weighing the
                technological feasibility and economic practicability of potential
                standards. EPA and NHTSA both consider implications of the rulemaking
                on CO2 emissions as well as criteria pollutant emissions.
                And, NHTSA's role as a safety regulator inherently leads to the
                consideration of safety implications when establishing standards. The
                balancing of competing factors by both EPA and NHTSA are consistent
                with each agency's statutory authority and recognize the overlapping
                obligations the Supreme Court pointed to in directing collaboration.
                ---------------------------------------------------------------------------
                 \2463\ 549 U.S. 497, 531 (2007).
                 \2464\ Id. at 532.
                ---------------------------------------------------------------------------
                 As discussed in prior rulemakings setting GHG standards,\2465\ EPA
                may establish technology-forcing standards under section 202(a), but it
                must provide a rationale for concluding that the industry can develop
                the needed technology in the available time. However, EPA is not
                required to set technology-forcing standards under section 202(a).
                Rather, because section 202(a), unlike the text of section 202(a)(3)
                and section 213(a)(3),\2466\ does not specify that standards shall
                obtain ``the greatest degree of emission reduction achievable,'' EPA
                retains considerable discretion under section 202(a) in deciding how to
                weigh the various factors, consistent with the language and purpose of
                the Clean Air Act, to determine what standards are appropriate.
                ---------------------------------------------------------------------------
                 \2465\ See, e.g., 77 FR 62624, 62673 (Oct. 15, 2012), EPA and
                NHTSA final rule for 2017 and later model year light-duty GHG
                emissions and CAFE standards.
                 \2466\ Section 202(a)(3) provides that regulations applicable to
                emissions of certain specified pollutants from heavy-duty vehicles
                or engines ``shall contain standards which reflect the greatest
                degree of emission reduction achievable through the application of
                technology which the Administrator determines will be available . .
                . giving appropriate consideration to cost, energy, and safety
                factors associated with the application of such technology.'' 42
                U.S.C. 7521(a)(3). Section 213(a)(3) contains a similar provision
                for new nonroad engines and new nonroad vehicles (other than
                locomotives or engines used in locomotives). 42 U.S.C. 7547(a)(3).
                ---------------------------------------------------------------------------
                 The proposed rule presented an analysis of alternatives, in support
                of the Administrator's consideration of a range of alternative
                CO2 standards as potential revisions of the existing
                standards for model years 2021 and later, from the previous standards
                (representing an increase in stringency of approximately 5 percent per
                year from MY 2021 through MY 2025) to several less stringent
                alternatives. These alternatives ranged from a zero percent increase in
                stringency to a stringency increase for passenger cars of 2 percent per
                year and for light trucks of 3 percent per year, in addition to the
                baseline alternative consisting of the previous standards.\2467\ The
                analysis supported the range of alternative standards based on factors
                relevant to the EPA's exercise of its section 202(a) authority, such as
                emissions reductions of GHGs and other air pollutants, the necessary
                technology and associated lead-time, the costs of compliance for
                automakers, the impact on consumers with respect to cost and vehicle
                choice, and effects on safety. The proposed rule identified the
                alternative composed of a zero percent increase in stringency as the
                preferred alternative.
                ---------------------------------------------------------------------------
                 \2467\ 83 FR 42990, Table I-4 (August 24, 2018).
                ---------------------------------------------------------------------------
                 EPA received numerous public comments on the range of stringency
                alternatives in the proposed rule and the Administrator's consideration
                of various factors in determining appropriate GHG standards under
                section 202(a) of the CAA. Below EPA responds to comments on these
                issues. EPA notes that many comments concerned the technical foundation
                and analysis upon which EPA was basing its regulatory decisions, such
                as the modeling of emission control technologies and costs, the safety
                analysis, and consumer issues. Comments specific to these analyses are
                discussed elsewhere in this preamble. The section below addresses
                comments specifically addressing EPA's considerations in finalizing
                appropriate CO2 emissions standards under the CAA.
                 EPA's conclusion, after consideration of the factors described
                below, public comments, and other information in the administrative
                record for this action is that holding CO2 emissions
                standards for MY 2020 flat through MY 2026 is not appropriate or
                reasonable. EPA
                [[Page 25107]]
                concludes steady stringency increases year over year are warranted, but
                that the MY 2021-2026 standards first established in 2012 are not
                appropriate taking into account lead time and the various factors
                described below. Accordingly, the Administrator has concluded that 1.5
                percent annual increases in stringency from the MY 2020 standards
                through MY 2026 (Alternative 3 of this final rule analysis) \2468\ are
                reasonable and appropriate.
                ---------------------------------------------------------------------------
                 \2468\ The numbered Alternatives presented in the SAFE proposed
                rule (see Table I-4 at 83 FR 42990, August 24, 2018) were in some
                cases defined differently than those presented in this final rule
                (see Section V). Unless otherwise stated, the Alternatives described
                in this section refer to those presented in this final rule.
                ---------------------------------------------------------------------------
                a) Consideration of the Development and Application of Technology To
                Reduce CO2 Emissions
                 When EPA establishes emission standards under CAA section 202, it
                considers both what technologies are currently available and what
                technologies under development may become available. For today's final
                rule, EPA considered the analysis of the potential penetration into the
                future vehicle fleet of a wide range of technologies that both reduce
                CO2 and improve fuel economy (see FRIA Chapter X). The
                majority of these technologies have already been developed, have been
                commercialized, and are in-use on vehicles today. These technologies
                include, but are not limited to, engine and transmission technologies,
                vehicle mass reduction technologies, technologies to reduce aerodynamic
                drag, and a range of electrification technologies. The electrification
                technologies include 12-volt stop-start systems, 48-volt mild hybrids,
                strong hybrid systems, plug-in hybrid electric vehicles, and dedicated
                electric vehicles.
                 This consideration is especially important given current
                projections about relatively lower fuel prices than what was projected
                in 2012. In that rulemaking, EPA expressed concern that some
                alternatives may require too much advanced technologies (including
                electrification) in light of uncertain consumer acceptance of added
                costs, as well as the technologies themselves.\2469\ There, EPA
                concluded that more stringent increases in technology penetration rates
                raise serious concerns about the ability and likelihood that
                manufacturers can smoothly implement additional technologies to meet
                requirements.\2470\
                ---------------------------------------------------------------------------
                 \2469\ 77 FR 62879.
                 \2470\ See 77 FR at 62875, discussion about certain alternatives
                may require too much electrification and ``may well be overly
                aggressive in the face of uncertain consumer acceptance of both the
                added costs and the technologies themselves. EPA continues to
                believe these technology penetration rates are inappropriate given
                the concerns just voiced.'' At 62877, ``This increase in tech
                penetration rates raises serious concerns about the ability and
                likelihood manufacturers can smoothly implement. . . .''
                ---------------------------------------------------------------------------
                 As shown in Section VII of this preamble and in FRIA Section VII,
                the projected penetration of technologies varies across the
                Alternatives considered for this final rule. In general, the baseline
                alternative consisting of the previous EPA standards as finalized in
                2012 was projected to result in the highest penetration of advanced
                technologies into the vehicle fleet, in particular mild hybrids at 7.1
                percent penetration and strong hybrids at 9 percent penetration by MY
                2030. By contrast, the revised final standards adopted today (1.5
                percent per year stringency improvement from MY 2021 through MY 2026)
                are projected to result in a significantly lower level of mild and
                strong hybrids used to meet the standards, at 1.6 percent mild hybrids
                and 2.2 percent strong hybrids by MY 2030. Further, the final rule
                analysis indicates that the previous CO2 standards would
                have led to a projected 5.7 percent penetration of dedicated electric
                vehicles (EV), with 0.4 percent penetration of plug-in hybrid electric
                vehicles (PHEV); the revised final standards reduce this projected
                level to 3.7 percent EV penetration (with 0.2 percent PHEV
                penetration), which again is more in line with what the EPA believes is
                a more appropriate projected level of market penetration.
                 The technology penetration rates in the analysis for the final rule
                are changed since EPA's prior analysis. These changes in the estimated
                penetrations in this rulemaking are due to changes in the model that
                are meant to reflect consumer response to the standards, as well as
                changes to estimates for technology costs and effectiveness. In the
                2017 Final Determination on Model Year 2022-2025 standards, where EPA
                found there was available and effective technology to meet the MY 2022-
                2025 standards, the technology was available at reasonable cost to the
                vehicle manufacturers and consumers, there was adequate lead time, and
                the standards were feasible and practicable. EPA also found that the
                previous MY 2022-2025 standards could be met largely through advanced
                gasoline vehicle technologies, with low levels of electrified
                vehicles.\2471\ The levels of electrified vehicle technologies
                projected in this final rule to meet the baseline Alternative (the
                previous GHG standards) differ slightly from those projected in the
                2017 Final Determination. In this final rule, EPA projects a combined
                strong and mild hybrid penetration of 16 percent (compared to 20
                percent in the 2017 Final Determination), with the share of mild
                hybrids somewhat lower (7 percent compared to 18 percent in the 2017
                Final Determination) and the share of strong hybrids higher (9 percent
                compared to 2 percent in in the 2017 Final Determination). EPA projects
                a total level of plug-in vehicles of 6 percent, similar to the 5
                percent total projected in the 2017 Final Determination, but with a
                slightly different mix of plug-in hybrid electric vehicles (0.4 percent
                compared to 2 percent in the 2017 Final Determination) and dedicated
                electric vehicles (5.7 percent compared to 3 percent in the 2017 Final
                Determination).
                ---------------------------------------------------------------------------
                 \2471\ ``Final Determination on the Appropriateness of the Model
                Year 2022-2025 Light-Duty Vehicle Greenhouse Gas Emissions Standards
                under the Midterm Evaluation,'' EPA-420-R-17-001, January 2017. See
                Table ES-1, page 4-5, and Section II (i), (ii), and (iii), pages 28-
                24. Hereafter ``2017 Final Determination.''
                ---------------------------------------------------------------------------
                 Another aspect of the analysis that EPA considered related to
                technology development and application is manufacturers' projected
                level of over-compliance under the alternatives considered for the
                final rule. Under the least stringent Alternatives (Alternative 1, zero
                percent stringency improvement, and Alternative 2, 0.5 percent per year
                stringency improvement), manufacturers overall are projected to over-
                comply with those levels of stringency. For example, under Alternative
                1, manufacturers are projected to achieve a CO2 level of 206
                g/mi in MY 2029, 16 g/mi below (more stringent than) the required
                target level of 222 g/mi. Similarly, for Alternative 2, manufacturers
                are projected to achieve a CO2 level of 205 g/mi in MY 2029,
                10 g/mi below the required target level of 215 g/mi. Thus, the industry
                is projected to considerably over-comply with the Alternative 1 and 2
                standards. Under the final standards, the projected level of over-
                compliance is much narrower, only 4 g/mi (198 g/mi by MY 2029 compared
                to a 202 g/mi target), and for other alternatives that are more
                stringent than the final standards, that gap is similar or even more
                narrow as shown in Table VII-7. This is an indication that the
                standards in Alternatives 1 and 2 may not represent
                [[Page 25108]]
                an appropriate level of stringency when compared to the pace at which
                manufacturers would be applying technologies. While some level of over-
                compliance is expected so that manufacturers retain a reasonable
                compliance margin, Alternatives 1 and 2 would, based on the final rule
                analysis, result in manufacturers retaining a compliance margin more
                than 2-3 times that of the other alternatives. The Administrator has
                rejected those lower stringency Alternatives in part for this reason
                and believes that the final standards (Alternative 3, 1.5 percent per
                year stringency improvement) represent an appropriate margin of
                compliance that can be attained given the projected pace of
                manufacturers' application of technologies.
                 EPA received several comments regarding its consideration of the
                development and application of GHG reducing technologies. The
                California Air Resources Board (CARB) commented that, despite what they
                characterize as evidence of widely available technology, EPA has
                proposed to promulgate emission standards that are less stringent than
                existing standards and that would lead to increased emissions of GHGs.
                The New York State Department of Environmental Conservation commented
                that the proposal did not ``appropriately value, or consider,
                technology advancement and innovation by OEMs and automotive parts
                suppliers'' and noted the role of technology innovation in reducing
                technology costs. EPA notes that the agencies specifically considered
                technology cost-savings attributable to experience with technology--in
                other words, the analysis provides that technology costs reduce over
                time.
                 The Center for Biological Diversity (CBD) et al. commented that
                since technologies exist today that can achieve the current standards,
                reducing the standards to the level proposed in the NPRM is contrary to
                the objectives of the Clean Air Act. These parties further commented
                that EPA failed to make a proposed finding that additional lead-time is
                necessary, as they argue is required by Section 202(a)(2). The Green
                Energy Institute at Lewis and Clark Law School and others similarly
                commented that EPA lacks a reasonable justification for extending the
                phase-in period for the current standards because compliant
                technologies currently exist and are already commercially available.
                 The Attorney General of California and others commented that EPA
                acknowledges that most or all technology necessary to meet the current
                standards is available, and does not provide evidence to support how
                additional lead time is ``necessary to permit the development and
                application of the requisite technology.''
                 In response to the public comments, and as EPA indicated in the
                proposal and in the 2012 Final Rule establishing the previous
                standards, the technologies projected to be used to meet the GHG
                standards, including the alternatives in the proposal as well as the
                final standards, are currently available and in production. If the
                appropriateness of the standards were based solely on an assessment of
                technology availability, and lead time considerations were limited to
                the development of such technology, EPA might consider more stringent
                CO2 standards to be potentially appropriate. But this is not
                the sole or predominant factor to be weighed. In 2012, EPA had to
                balance this issue as well. As in 2012, manufacturers today are capable
                of building vehicles that can meet the standards that any of the
                regulatory alternatives evaluated in the final rule would require.
                However, greater uncertainty about consumer acceptance of those
                technologies (as compared to what EPA believed was likely in 2012)
                means that providing more lead time is appropriate.\2472\
                ---------------------------------------------------------------------------
                 \2472\ See 77 FR at 62871 (``As stated above, EPA's analysis
                indicates that there is a technology pathway for all manufacturers
                to build vehicles that would meet their final standards as well as
                the alternative standards. The differences between the final
                standards and these analyzed alternatives lie in the per-vehicle
                costs and the associated technology penetration rates.'').
                ---------------------------------------------------------------------------
                 As in 2012, EPA disagrees with commenters that a finding that
                necessary technology is available is, by itself, determinative of the
                appropriate emission standard under CAA section 202(a). As described in
                the proposed rule and in this section of the final rule, the
                Administrator weighs technology availability and lead time along with
                several other factors, including costs, emissions impacts, safety, and
                consumer impacts in determining the appropriate standards under section
                202(a) of the CAA.
                 Under this analysis, given the factors discussed later in this
                Section, the previous standards would yield technology penetration
                rates for advanced technologies beyond what is appropriate and
                reasonable. By contrast, the final standards are projected to result in
                more modest penetration rates for advanced technologies that
                nonetheless will achieve an increased level of technology penetration
                compared to the standards applicable for MY 2020. For example, the
                final rule analysis projects that dynamic cylinder deactivation
                penetration for MY 2030 would be 39.2 percent under the previous
                standards for, but 34.4 percent under today's final standards.
                Similarly, turbocharged engine penetration would be a projected 48
                percent by MY 2030 under the previous standards, compared to 36.4
                percent under the final standards. In addition, mild hybrids are
                projected to change from 7.1 percent to 1.6 percent, strong hybrids
                from 9 percent to 2.2 percent, and dedicated electric vehicles from 5.7
                percent to 3.7 percent (all for MY 2030) under the final standards
                instead of the previous standards. The Administrator believes that the
                level of technology development and application for the final standards
                is an appropriate balance, in light of the relevant factors considered
                as a whole, as discussed below.
                (b) Consideration of the Cost of Compliance
                 EPA is required to consider costs of compliance when setting
                standards under section 202(a). The standards finalized today would
                reduce required technology costs for the industry by an estimated $108
                billion for the vehicles produced from MY 2017 through MY 2029 (at 3
                percent discount rate, see Section VII) compared to the EPA standards
                established in 2012. While less-stringent increases would result in
                additional technology cost savings ($129 billion and $126 billion for
                Alternatives 1 and 2, respectively), technology cost savings are only
                one element that EPA considers.
                 In addition to capital cost savings, the final standards would
                reduce the per-vehicle costs by $1,250 per vehicle in MY 2030, compared
                to the standards set in 2012, as shown in Table VII-77. While less-
                stringent increases would result in greater per-vehicle technology
                cost-savings, cost-savings alone do not dictate the appropriate
                standards. For example, Alternatives 1 and 2 would save manufacturers
                $1,218 and $1,181 in per-vehicle costs in MY 2030 compared to the
                previously issued standards. Alternatives more stringent than the final
                standards would be more burdensome to manufacturers, with Alternatives
                4 through 8 ranging from a cost savings to manufacturers of $927 to
                $351 per-vehicle compared to the previous standards.
                 The costs to comply projected in this final rule are higher than
                those previously projected by EPA in the 2017 Final Determination: In
                2017 EPA projected that the per-vehicle cost to meet the MY 2025
                standards would be $875 on average, with a range of $800 to $1,115
                considering a range of
                [[Page 25109]]
                sensitivities (in 2015 dollars).\2473\ The costs to the auto industry
                for complying with the previous MY 2022-2025 standards projected in the
                2017 Final Determination were $24 billion to $33 billion (in 2015$ at 7
                percent and 3 percent discount rates, respectively).\2474\ Again, EPA
                notes that the values in this final rule analysis and the values in the
                2017 Final Determination have different points of reference making them
                not directly comparable, as discussed above.
                ---------------------------------------------------------------------------
                 \2473\ See 2017 Final Determination Table ES-1, page 4-5, and
                II(v), page 24-26.
                 \2474\ Id. at Table ES-4, page 7.
                ---------------------------------------------------------------------------
                 Several public comments addressed EPA's consideration of costs of
                compliance in setting the revised standards. The Alliance of Automobile
                Manufacturers (Alliance) commented that the proposal's cost estimates
                for the current MY 2021 and later standards differed from what EPA
                projected in 2012 when setting those standards. The Alliance argued
                that that those changes in the expected costs of the previously issued
                standards provide significant reasoned support for EPA's view that the
                existing standards should be reduced.
                 The Association of Global Automakers (Global Automakers) commented
                on the importance of lead time for technology investment. While it
                agreed that the existing standards are too stringent, it stated that
                vehicle manufacturers and suppliers have invested $76 billion in
                manufacturing facilities, and that much of that was for improvement in
                CO2 emission reductions and fuel economy improvements. At
                least some of that investment, according to Global Automakers, was made
                to meet the standards set in 2012. Global Automakers expressed concern
                with an abrupt halt to gradual fuel economy improvements, as such an
                approach could result in stranded capital investments for automakers
                and suppliers.
                 CBD and others disagreed with EPA's conclusion that the cost of
                broader adoption of technologies is unreasonable in light of other
                factors considered by EPA. CBD and others claimed that the Clean Air
                Act narrowly allows for consideration of cost only as a question of
                whether costs of compliance make it infeasible for manufacturers to
                meet standards within the relevant period. They argue that this
                consideration relates to lead time, and not to a broader consideration
                of costs. They assert that broader compliance cost considerations apply
                only to the motor vehicle industry. They also claim that compliance
                costs to meet the standards set in 2012 for the 2017-2025 model years
                are not challenging to the industry.
                 These commenters also state that the costs to industry to meet the
                standards are not high enough to require reducing standards, to permit
                development and application of the required technology. They claim that
                the only burden that Congress intended to impose as a constraint on
                emission reduction requirements are costs that are ``so severe as to
                preclude the deployment of required technology during the relevant
                period.''
                 The New York State Department of Environmental Conservation
                commented on the role of technology innovation in considering
                technology feasibility, while acknowledging that the feasibility
                analysis allows for consideration of numerous factors argues that since
                technology exists today to meet the standards for MY 2026, no lead time
                is necessary. It further states that EPA did not appropriately balance
                or consider in the proposal future technological advancements and OEM
                innovation that will further constrain the costs of new technology.
                 In response to the Alliance's comment that the projected compliance
                costs have changed significantly from EPA's 2012 rule, EPA agrees.
                Indeed, this is a significant factor in EPA's conclusion that the
                previous standards were too stringent. EPA notes that the projected
                difference between the cost to comply with the previous standards and
                the costs to comply with the standards established today is lower in
                this final rule analysis as compared to the projected difference
                between the proposal's preferred alternative and the previous
                standards. EPA concludes that the final standards nevertheless result
                in significant reductions in required technology costs for auto
                manufacturers compared to the previous standards.
                 EPA also considered the Global Automakers' concern that freezing
                the standards from MY 2021-2026 as proposed could result in stranded
                capital for the auto industry and automotive suppliers who have
                invested significantly in meeting the previous standards. The standards
                EPA is finalizing today, unlike the proposed preferred alternative,
                will require the gradual increase in CO2 improvements across
                the fleet, at a rate of 1.5 percent per year stringency improvement,
                thus supporting investments in GHG-reducing technologies, at a pace
                that EPA believes is more reasonable than that of the previous
                standards.
                 EPA disagrees with CBD et al.'s comments that the agency's
                consideration of costs is inappropriate or not supported by the record.
                EPA disagrees that Congress intended section 202(a)(2)'s requirement to
                give ``appropriate consideration to the cost of compliance within such
                period'' to mean that the agency ``only consider compliance costs if
                they are so severe as to preclude deployment of the requisite
                technology during the period.'' EPA does not interpret the Clean Air
                Act as limiting EPA's consideration of costs to manufacturers only to
                the question of whether such costs are so high that a manufacturer
                could not afford to deploy the technology in question for a given model
                year--that would be tantamount to suggesting that EPA must always set a
                standard to achieve ``the greatest degree of emission reduction
                achievable through the application of technology,'' which as discussed
                above is not EPA's approach to setting standards such as these under
                section 202(a). And this is particularly important when setting
                CO2 standards, which, as described above, have a significant
                impact on vehicle utility and performance that differs from other
                standards established under Section 202. As discussed above, Congress
                specified such technology-forcing standards elsewhere in section 202
                and could have done so here (or otherwise specified that standards
                shall take effect ``as soon as practicable'' while taking into
                consideration costs and other factors)--but did not do so. Section
                202(a) prevents EPA from implementing standards sooner than feasible,
                taking into account lead time considerations and the cost of
                compliance, but does not require standards be implemented as soon as
                feasible or at the limit of feasibility, taking into account the cost
                of compliance. EPA notes that it received numerous comments on the
                analysis underlying the proposed rule, and the analysis for this final
                rule in fact was changed from the proposal in consideration of these
                comments, as discussed in Section VI.B. Nevertheless, the projected
                costs to comply with the previous MY 2021-2026 standards remain
                significant as discussed above, and EPA has considered these costs
                along with other factors under the CAA in determining the final
                standards, as discussed in Section VIII.A.3.h) below.
                (c) Consideration of Costs to Consumers
                 In this section EPA considers the cost impacts on consumers. First,
                the initial up-front costs to consumers are discussed, then the costs
                associated with fuel expenditures, and finally the total ownership
                costs to consumers over the life of the vehicles.
                [[Page 25110]]
                 In addition to the $1,250 per-vehicle technology costs to the
                automotive industry described above, which EPA expects could, and
                likely would, be passed on to consumers, the analysis estimates other
                per-vehicle costs that could be borne by consumers, specifically costs
                attributed to changes in financing, insurance, taxes, and other fees,
                as shown in Section VII. Considering these additional costs, EPA's
                final standards (Alternative 3) would result in reduced costs to
                consumers of $1,385 in MY 2029 (at a 3 percent discount rate) compared
                to EPA's previously issued standards. While alternatives lower in
                stringency than the final standards would save consumers more (i.e.,
                Alternatives 1 and 2 would save consumers $1,665 and $1,637,
                respectively, in MY 2029 at 3 percent discount rate), while
                alternatives more stringent than the final standards would save
                consumers less (i.e., Alternatives 4 through 7 would save consumers a
                range of from $1,329 to $620, for MY 2029 at 3 percent discount rate),
                this is only one of the factors EPA considers in setting standards. On
                balance, EPA believes that further increases in stringency, compared to
                the proposal, are appropriate and reasonable.
                 Compared to the previously issued CO2 standards, the
                standards finalized today will result in increased fuel consumption and
                associated expenditures for consumers. The analysis detailed in the
                Final RIA and summarized in Section VII of this preamble projects the
                increased fuel consumption for owners of the vehicle over the projected
                life of the vehicle, up to 39 years, as compared to the previously
                issued standards as the baseline. For example, as shown in Table VII-84
                (at a 3 percent discount rate), consumers will spend $1,461 more in
                fuel costs over the vehicle lifetime, which the analysis assumes can be
                up to 39 years,\2475\ under today's final standards (Alternative 3)
                compared to the previously issued standards.
                ---------------------------------------------------------------------------
                 \2475\ For further information of on the modeled distribution of
                registrations by age see, e.g., Table VI-238--Registrations, Total
                VMT, and Proportions of Total VMT by Vehicle Age (in Section
                VII.D.2.b).2.(d)) which shows the distribution of registrations by
                vehicle age.
                ---------------------------------------------------------------------------
                 EPA notes that, when comparing lifetime fuel savings for all owners
                of a vehicle to the upfront additional ownership costs--generally borne
                by the initial purchaser, a net reduction in benefits of $175 is seen
                under the final standards. That said, as noted by several commenters,
                consumers keep vehicles for a much shorter period of time prior to
                trading the vehicle in for another or selling the vehicle.\2476\ CFA,
                for instance mentioned that consumers retain vehicles for more than
                five years, and a group of State Comptrollers and Treasurers referred
                to an IHS Markit report that the average length of time a consumer
                keeps a new car is approximately 6.6 years. Accordingly, such a
                simplistic comparative approach would anticipate that a consumer
                account for fuel savings over a much longer period of time than would
                be rational. Further, it is important to note that consumers are
                informed of estimated average annual fuel costs for the vehicle, as
                well as a comparison of the difference between five years'-worth of
                fuel costs or savings compared to an average new vehicle on the
                Monroney label that must be posted on every new vehicle offered for
                sale.
                ---------------------------------------------------------------------------
                 \2476\ It should be noted, however, that, all else being equal,
                improved fuel economy can improve resale value of a vehicle. That
                said, it is not at all clear that consumers generally anticipate
                potential future incremental trade-in value attributable to improved
                fuel economy when making a decision as to which new vehicle to
                purchase.
                ---------------------------------------------------------------------------
                 In the 2017 Final Determination, EPA projected that the previous MY
                2022-2025 standards compared to the MY 2021 standards would provide
                fuel savings of $52 billion to $92 billion and total net benefits of
                $59 billion to $98 billion (in 2015 dollars and at 7 percent and 3
                percent discount rates, respectively, and based on AEO2016 reference
                case fuel prices). The up-front vehicle costs to consumers were
                projected to be approximately $926 per vehicle, including the vehicle
                technology costs, taxes and insurance.\2477\ EPA projected that
                consumers would realize net savings of $1,650 over the lifetime of a
                new MY 2025 vehicle (net of increased lifetime costs and lifetime fuel
                savings).\2478\ Under the final standards, vehicle sales are expected
                to increase by 2.2 million vehicles over MY 2017-2029 compared to
                projected sales under the previous standards. EPA views this projection
                of vehicle sales increases resulting from the final standards as
                important in facilitating the turnover of the fleet to newer, safer
                vehicles, all of which will be subject to increasingly stringent
                criteria pollutant emission requirements as federal Tier 3 emission
                standards continue to phase in from MY 2017 through MY 2025.
                 Below the major comments are summarized regarding EPA's
                consideration of the impact of the revised standards on consumers.
                Securing America's Future Energy (SAFE) commented that vehicle prices
                are influenced by many factors beyond the GHG standards, and that costs
                to improve fuel economy make up only a portion of the vehicle price.
                SAFE notes that fuel savings from efficient vehicles offsets increase
                ownership costs. SAFE further claims, without support, that standards
                ``do not have a major role in creating higher vehicle prices, or in
                suppressing sales.'' Accordingly, SAFE argues that pausing fuel economy
                increases, as proposed in the NPRM, is not justified. SAFE suggests
                that fuel savings impacts should be discussed along with technology
                cost increases.
                 CBD and others commented that EPA's consideration of consumer
                costs, including finance and insurance costs, cannot outweigh its
                public health mandate. Such commenters noted that some of the options
                analyzed in the notice showed that fuel savings of the lifetime of the
                vehicle outweighed upfront vehicle price increases, and that not
                choosing such an alternative is not justified. CBD then goes on to
                argue that the analysis inflates technology costs and undercounts fuel
                savings.
                 The California Attorney General and others claim that EPA's
                consideration of the potential increased costs for consumers related to
                maintenance, financing, insurance, taxes, and other fees is
                unjustified, unlawful, and contrary to its prior position that
                compliance cost considerations include only costs to the motor-vehicle
                industry.
                 EPA notes that fuel efficiency and GHG standards affect labor and
                materials costs, technology add-ons, and sales mix, and expects the
                estimated cost decrease from these final standards to have a positive
                effect on the auto market and vehicle buyers. As described in the
                notice and throughout this preamble, EPA disagrees that standards have
                no major impact on increasing prices or suppressing sales. Fuel-saving
                technology adds costs, and as prices increase, fewer consumers can
                afford to buy new cars--either because they cannot afford a new car, or
                because they decide to purchase an older vehicle, or because they
                decide to keep their existing vehicle. EPA also notes that both the
                notice and this preamble discusses fuel savings from the various
                alternatives analyzed. Some commenters suggest EPA calculate and
                consider fuel savings, spread over the lifetime of the vehicle up to 39
                years and experienced by multiple owners--compared to the upfront
                vehicle costs, which are generally paid for by the original purchaser
                either in cash or through additional finance costs over a much shorter
                period of time. This approach, which would yield a projected $175 in
                additional costs (additional lifetime outlays for fuel minus avoided
                upfront vehicle costs)
                [[Page 25111]]
                over the multi-owner, lifetime of a vehicle beyond the initial
                ownership savings, distorts the comparison. Instead, EPA concludes that
                the upfront vehicle technology costs (and associated financing costs)
                are a more important factor. In other words, a consumer is more likely
                to buy a new vehicle at a lower up-front price even if that vehicle
                will incur a more-than offsetting level of fuel costs over its lifetime
                that will be borne by the first and all subsequent owners of the
                vehicle.\2479\ By reducing upfront costs, more consumers will be able
                to afford new vehicles, which will result in a quicker fleet turnover
                to safer, more efficient vehicles that emit lower amounts of criteria
                pollutants than the existing fleet. In fact, the agencies project that
                the revised standards will result in 2.2 million additional new
                vehicles sold--all of which would meet the latest safety standards and
                be subject to the phase-in of the Tier 3 criteria pollutant emission
                standards.
                ---------------------------------------------------------------------------
                 \2479\ For further discussion regarding consumers valuation of
                fuel economy, see preamble section VI.D.1.b).(2) (sales), preamble
                section VI.D.1.b).(8), and Final Regulatory Impact Analysis section
                III.C.
                ---------------------------------------------------------------------------
                 With respect to the comments that consideration of costs to
                consumers is contrary to CAA section 202(a)(2), EPA disagrees. As
                discussed above, section 202(a)(2) requires EPA to consider the cost of
                compliance, which EPA has done, and it allows EPA to consider other
                costs, including costs to consumers, which EPA also have done, in this
                rule and past rules setting standards under section 202(a). The statute
                sets some minimum requirements for EPA's consideration, but permits a
                wider range of concerns to be considered, including public health and
                welfare but also safety, costs to consumers, and other factors
                discussed herein. As discussed above, and below, EPA has considered the
                effects of a range of potential standards across this entire set of
                factors. The agency is permitted to take all of these factors into
                account, and that is what it has done in selecting the final standards.
                d) Consideration of GHG Emissions and Other Air Pollutant Emissions
                 As discussed above, the purpose of GHG standards established under
                CAA section 202 is to reduce GHG emissions, which EPA has found to
                endanger public health and welfare, in an appropriate manner that takes
                into account other factors as directed by Congress and in the
                reasonable exercise of EPA's discretion under the statute. Today's
                final standards are projected to increase CO2 emissions
                compared to the previously issued standards, by a total of 867 million
                metric tons (MMT) over the lifetime of MY 1977 through MY 2029 vehicles
                (see Section VII of this preamble)--i.e., by 2.9% of the amount
                projected to be attributable to passeners cars and light trucks under
                the baseline/augural standards. Of this CO2 emissions
                increase, 731 MMT would come from tailpipe emissions, and an additional
                136 MMT from upstream sources, both being nearly 3% greater than
                projected to occur under the baseline/augural standards. The analysis
                projects that Alternatives more stringent than the final standards
                would result in smaller increases in CO2 emissions. Also
                compared to the baseline/augural standards, and also over the lifetime
                of MY 1977-2029 vehicles, Alternatives 4 through 7 are projected to
                increase CO2 emissions by 826 MMT (2.8%) to 361 MMT (1.2%).
                Alternatives less stringent than the final standards would increase
                CO2 emissions by a greater amount, 1,074 MMT (3.5%) and
                1,044 MMT (3.6%), for Alternatives 1 and 2 respectively.\2480\
                ---------------------------------------------------------------------------
                 \2480\ This preamble and the FRIA document estimate annual GHG
                emissions from light-duty vehicles under the baseline CO2
                standards, the final standards, and the standards defined by each of
                the other regulatory alternatives considered. For the final rule
                issued in 2012, EPA estimated changes in atmospheric CO2,
                global temperature, and sea level rise using GCAM and MAGICC with
                outputs from its OMEGA model. Because the agencies are now using the
                same model and inputs, outputs from NHTSA's EIS (that used more
                recent versions of GCAM and MAGICC) were analyzed. Today's analysis
                estimates that annual GHG emissions from light-duty vehicles under
                the CO2 standards and corresponding CAFE standards, which
                are very similar. Especially considering the uncertainties involved
                in estimating future climate impacts, the very similar estimates of
                future GHG emissions under CO2 standards and
                corresponding CAFE standards means that climate impacts presented in
                NHTSA's EIS represent well the climate impacts of the CO2
                standards.
                ---------------------------------------------------------------------------
                 In addition to GHG emissions, EPA has considered the change in
                criteria air pollutant emissions impacts due to the revised
                CO2 standards. EPA has considered both tailpipe emissions
                and upstream emissions associated with increased fuel consumption.
                Unlike with CO2 emissions, which EPA found to be a long-
                lived greenhouse gas well-mixed throughout the global atmosphere,
                criteria pollutant emissions contribute primarily to local and regional
                air pollution. Generally, tailpipe emissions for volatile organic
                compounds (VOC), nitrogen oxides (NOX), and particulate
                matter (PM) decrease under the final standards compared to the previous
                standards, leading to improvements in human health in areas where air
                quality improves. Upstream emissions attributable to refining and
                transportation of the additional fuel needed under less stringent
                standards increase under the final standards, leading to adverse
                impacts on public health in locations where air quality worsens. The
                additional upstream emissions generally exceed the reduced tailpipe
                emissions, leading to net increases in these pollutants and net
                increases in adverse health effects. Under the model year analysis
                (changes in pollutants summed over the lifetimes of MY 1977-2029
                vehicles for calendar year 2017 and later), and relative to total
                emissions projected to be attributable to passenger cars and light
                trucks under the baseline/augural standards, these increases range from
                0.1% (for NOX) to 0.7% (for SO2 and PM). On the
                other hand, projected net emissions of carbon monoxide (CO) are 0.4%
                lower under the final standards than under the baseline/augural
                standards, and emissions of air toxics (e.g., benzene) are 0.1-0.4%
                lower under the final standards, varying among different toxic
                compounds.
                 In addition to evaluating emissions impacts under the model year
                analysis described above, EPA has considered the emissions impacts
                under a calendar year analysis, which provides information over a
                longer time horizon about the interactions between all vehicle model
                years on the road in any given calendar year--that is, considering the
                effects of the revised MY 2021 and later standards on fleet turnover
                and utilization from calendar year 2017 out to 2050. Both the model
                year analysis and the calendar year analysis provide relevant
                information about the impacts of EPA's standards. When viewed from the
                calendar year analysis perspective that extends through 2050, the
                emissions impacts of the revised MY 2021 and later standards compared
                to the baseline/augural standards vary over time, with cumulative
                differences generally being greater in magnitude than under the model
                year analysis: EPA's analysis shows cumulative VOC emissions through
                2050 under the final standards increasing by a total of nearly 575
                thousand tons (1.9%) relative to the cumulative amount projected to
                accrue through 2050 under the baseline/augural standards. On the same
                basis, estimated NOX and PM emissions increase by about 173
                thousand tons (0.8%) and 16.5 thousand tons (1.7%), respectively. On
                the other hand, also on the same basis, estimated CO and SO2
                emissions decrease by about 278 thousand tons (0.1%) and 38 thousand
                tons (0.8%), respectively.
                 As shown in the NHTSA Final Environmental Impact Statement (FEIS),
                [[Page 25112]]
                NHTSA's analysis indicates small air quality improvements in some areas
                and small decrements in others which could help or hinder individual
                areas' efforts to attain the NAAQS in the future.
                 EPA has also considered the health effects of air pollution
                associated with today's final standards. As discussed above, it is the
                cumulative contribution of the lower projected vehicle tailpipe
                emissions with the higher projected upstream emissions (primarily from
                the production and distribution of gasoline) which impact air quality.
                As noted above and presented in detail elsewhere in this preamble and
                the Final RIA, vehicle emissions are generally reduced due to the SAFE
                final rule.
                 Due largely to the projected increase in upstream emissions
                resulting from the increased production and transportation of gasoline
                resulting from the standards finalized today compared to the previous
                EPA standards, the Final Rule analysis projects increases in premature
                deaths, asthma exacerbation, respiratory symptoms, non-fatal heart
                attacks, and a wide range of other health impacts. While these health
                impacts are presented in detail elsewhere in this preamble and in the
                Final RIA, two factors suggest that the forgone premature mortality
                benefits are overstated. First, in the last year, EPA has completed
                analysis that demonstrated the likelihood that the air quality modeling
                approach used here (i.e., benefits per ton) overestimates foregone PM
                premature mortality benefits. Second, the 2012 rulemaking significantly
                overestimated gasoline price projections in its baseline, predicting
                lower fuel consumption, thus overestimating the premature mortality
                benefits in that rule. While gasoline price projections in this
                rulemaking have been updated to reflect recent data, the potential for
                this kind of unanticipated fluctuation in gasoline prices remains, thus
                estimates of fuel consumption and the correlated foregone premature
                mortality benefits may not capture actual market outcomes.
                 The valuation of premature mortality effects rely on the results of
                ``benefits per ton'' approach (BPT). This approach is a reduced form
                approach, which is less complex than full-scale air quality modeling,
                requiring less agency resources and time. Based on EPA's work to
                examine reduced form approach, the BPT may yield estimates of
                PM2.5-benefits for the mobile sector that are as much as 10
                percent greater than those estimated when using full air quality
                modeling.
                 The EPA is currently working on a systematic comparison of results
                from its BPT technique and other reduced-form techniques with results
                from full-form photochemical modelling. While this analysis employed
                photochemical modeling simulations, we acknowledge that the Agency has
                elsewhere applied reduced-form techniques. The summary report from the
                ``Reduced Form Tool Evaluation Project'', which has not yet been peer
                reviewed, is available on EPA's website at https://www.epa.gov/benmap/reduced-form-evaluation-project-report. Under the scenarios examined in
                that report, EPA's BPT approach in the 2012 rule (which was based off a
                2005 inventory) may yield estimates of PM2.5-benefits for
                the mobile sector that are as much as 10 percent greater than those
                estimated when using full air quality modeling. The estimate increases
                to 30 percent greater for the electricity sector. The EPA continues to
                work to develop refined reduced-form approaches for estimating
                PM2.5 benefits.
                 Also, in this regulation, a key projection that influences the
                estimation about car purchase and driving behavior is the gasoline
                price projection. From 2008 through 2018, the average monthly gasoline
                price ranged from less $1/gallon to $4/gallon.\2481\ The gasoline price
                level and the volatility of price changes are major drivers of car
                purchasing behavior thereby gasoline consumption and the resulting
                criteria pollutant emissions. If gasoline prices are lower than
                projected in an analysis, consumers are more likely to purchase less
                fuel efficient cars, resulting in more emissions and vice versa.
                ---------------------------------------------------------------------------
                 \2481\ https://www.eia.gov/energyexplained/gasoline/price-fluctuations.php.
                ---------------------------------------------------------------------------
                 With a lower fuel price projection and an expectation that new
                vehicle buyers respond to fuel prices, the 2012 rule would have shown
                much smaller fuel savings attributable to the more stringent standards.
                Projected fuel prices are considerably lower today than in 2012. The
                agencies now understand new vehicle buyers to be at least somewhat
                responsive to fuel prices, and the agencies have therefore updated
                corresponding model inputs to produce an analysis the agencies consider
                to be more realistic.
                 The first of these assumptions, fuel prices, was simply an artifact
                of the timing of the rule. Following recent periodic spikes in the
                national average gasoline price and continued volatility after the
                great recession, the fuel price forecast then produced by EIA (as part
                of AEO 2011) showed a steady march toward historically high, sustained
                gasoline prices in the United States. However, the actual series of
                fuel prices has skewed much lower. As it has turned out, the observed
                fuel price in the years between the 2012 final rule and this rule has
                frequently been lower than the ``Low Oil Price'' sensitivity case in
                the 2011 AEO, even when adjusted for inflation. The discrepancy in fuel
                prices is important to the discussion of differences between the
                current rule and the 2012 final rule, because that discrepancy leads in
                turn to differences in analytical outputs and thus to differences in
                what the agencies consider in assessing what levels of standards are
                reasonable, appropriate, and/or maximum feasible. Long-term predictions
                are challenging and the fuel price projections in the 2012 rule were
                within the range of conventional wisdom at the time. However, it does
                suggest that fuel economy and tailpipe CO2 regulations set
                almost two decades into the future are vulnerable to surprises, in some
                ways, and reinforces the value of being able to adjust course when
                critical assumptions are proven inaccurate. This value was codified in
                regulation when EPA bound itself to the mid-term evaluation process as
                part of the 2012 final rule.\2482\
                ---------------------------------------------------------------------------
                 \2482\ See 40 CFR 86-1818-12(h).
                ---------------------------------------------------------------------------
                 Because of these uncertainties surrounding air quality modeling of
                premature mortality effects, the projections of foregone PM premature
                mortality benefits are uncertain and may be over-stated. Fluctuations
                in gasoline prices contribute to this uncertainty, making it difficult
                to accurately project gasoline consumption and its related premature
                mortality benefits.
                 The analysis projects that the air pollution emission increases
                associated with the revised standards will lead to an increase of 440
                to 1,000 premature deaths--deaths that occur before the normally
                expected life span--0.5% more than the number of such deaths projected
                to occur under the baseline/augural standards and over the lifetime of
                the MY 1977-MY 2029 vehicles. In addition, a wide range of health
                impacts are projected to increase by 0.4-0.6% under the final standards
                compared to occurrences projected to occur the standards established in
                2012, as summarized in Table VII-132 et seq.
                 When quantified using the calendar year (CY) analysis perspective
                (CYs 2018-2050), under the revised final standards (compared to the
                previous standards), premature mortality is expected to increase from
                460 to 1,010 deaths (i.e., by 0.4%), upper and lower respiratory
                symptoms are expected to increase by 22,000 cases (0.4%), asthma
                exacerbations are projected to increase by 16,000 cases (0.4%), acute
                bronchitis
                [[Page 25113]]
                cases are projected by increase by 720 (0.4%), non-fatal heart attacks
                are projected to increase by 450 (0.4%), hospital admissions for
                cardiovascular and respiratory issues are projected to increase by 225
                (0.4%) cases, and emergency room visits for respiratory issues are
                projected to increase by 260 (0.4%). In addition, these additional
                health impacts are expected to result in an additional 61,000 work loss
                days (0.3% of the number projected under the baseline/augural
                standards) and 355,000 minor restricted activity days (0.4% more than
                under that baseline/augural standards) for the public. Compared to the
                baseline/augural standards, the agencies estimate that the final
                standards rule will increase by 0.3-0.4% each of the various health
                impacts accumulated through 2050 (e.g., premature deaths, upper and
                lower respiratory symptoms, asthma exacerbations, acute bronchitis
                cases, hospital admissions for cardiovascular and respiratory issues,
                emergency room visits for respiratory issues).
                 In the 2017 Final Determination, EPA projected GHG emissions
                reductions of 540 million metric tons over the lifetimes of MY 2022-
                2025 vehicles.\2483\ EPA also projected criteria pollutant emission
                reductions for CY2040 of 97,000 tons of VOC, 24,000 tons of
                NOX, 3,600 tons of PM2.5, and 15,000 tons of
                SO2.\2484\ EPA projected that these emissions reductions
                would result in positive health benefits through CY2050.\2485\ In this
                final rule, the revised final standards compared to the previous
                standards are projected to result in an increase in emissions and
                health incidences, as discussed above, resulting in $5 billion or $3
                billion (in 2018 $, and reflecting, respectively, either a 7 percent or
                3 percent discount rate) in foregone public health benefits (see Table
                VII-103 and Table VII-104).
                ---------------------------------------------------------------------------
                 \2483\ 2017 Final Determination at Table ES-3, page 6, and
                Section II (iv), page 24.
                 \2484\ 2016 Proposed Determination at Appendix C, Table C.54,
                page A-163.
                 \2485\ Id. at Table C.87, page A-183.
                ---------------------------------------------------------------------------
                 In public comments on these topics, the Attorney General of
                California and others commented that, in adopting the previous
                standards, EPA focused on obtaining significant CO2 emission
                reductions, but now proposed to increase emissions relative to the
                previous standards without sufficient justification. They claim that
                EPA offered no justification of acknowledgement of a change in
                position, stating that none of the alternatives further the goal of
                CO2 emission reductions. They argue that EPA justifies its
                proposal on the limited impact of the rule on global climate change,
                and that failing to seek incremental improvements is contrary to the
                EPA's duties under the Clean Air Act.
                 The United States Conference of Catholic Bishops commented that
                considering public safety of any set of standards requires giving
                significant weight to the effect of air pollution, and that the
                proposal failed to promote public health and safety.
                 The Chesapeake Bay Foundation (CBF) claims that the proposal would
                have significant health consequences that disproportionately impact
                minority and low-income communities in the Chesapeake Bay. They discuss
                general impacts of climate change CBF argues that criteria pollutant
                health impacts of the proposal, should be more heavily weighed against
                safety impacts of the rule.
                 The State of Washington commented that the agencies did not analyze
                public health effects from increased criteria pollutant emissions
                arising from increased petroleum consumption or environmental justice
                concerns. They claim that the NPRM's discussion of the negligible
                impact of the rulemaking on global climate change is ``deeply
                concerning.''
                 As noted above, EPA agrees that the purpose of Title II emission
                standards is to protect the public health and welfare from air
                pollution, and in establishing emission standards, the agency is
                cognizant of the importance of this goal. At the same time, EPA
                balances multiple factors in determining what standards are reasonable
                and appropriate. And, contrary to some commenters' views, unlike other
                provisions in Title II, section 202(a) does not require the
                Administrator to set standards which result in the greatest degree of
                emissions control achievable. Thus, in setting these standards, the
                Administrator has taken into consideration other factors discussed
                above and below, including not only technological feasibility, lead-
                time, and the cost of compliance, but also potential impacts of vehicle
                emission standards on safety and other impacts on consumers.
                 Several commenters claimed that the agencies did not analyze health
                impacts of the various alternatives, but this is not accurate. First,
                the notice and PRIA included this information in monetized terms to
                facilitate the balancing of various factors. Further, NHTSA conducted a
                comprehensive Draft Environmental Impact Statement, which discussed
                these effects in detail. For this final rule, these health impacts have
                been separately itemized, as summarized above. Other commenters claimed
                that the agencies did not sufficiently consider environmental justice
                elements in the proposal. This, too, is inaccurate, as discussed
                elsewhere in this preamble.
                 In response to comments of the California Attorney General and
                others, that the Clean Air Act cannot allow for increases in a
                regulated emission, EPA notes that the 2012 Final Rule specifically
                called for a Mid Term Evaluation process that envisioned the potential
                for an adjustment of the standards in case the stringency increases
                established in 2012 were no longer reasonable and appropriate. As
                discussed above, the increases in stringency of the standards for MY
                2021-2025 are, on balance, not reasonable and appropriate based on a
                consideration of the factors described in this preamble. EPA now
                recognizes based on updated information and analysis that industry
                should be provided additional lead time to meet the later model years
                of standards set in the 2012 rule, and, as discussed in this preamble,
                industry is having unanticipated difficulties complying with earlier
                years of the standards, with fleetwide performance failing to meet
                CO2 emission targets in MY 2016 and MY 2017. That is not to
                say that CO2 and criteria pollutant emissions are not
                significant factors in this rulemaking. Indeed, they are weighed
                heavily along with other important factors considered by EPA, which has
                led to increasing stringency on a 1.5 percent annual basis for the
                2021-2026 model years. Importantly, the agencies project that the
                revised standards will result in an additional 2 million new vehicles
                sold before 2030 compared to under the baseline/augural standards. This
                means that an additional 2 million vehicles will be produced during the
                phase-in of the Tier 3 emission standards, which implement more
                stringent tailpipe standards for criteria pollutants, displacing
                greater numbers of higher-emitting older vehicles and providing
                significant health benefits. As discussed, when finalizing the Tier 3
                standards in 2014, ``[t]he final Tier 3 vehicle and fuel standards
                together will reduce dramatically emissions of NOX, VOC,
                PM2.5, and air toxics.'' \2486\
                ---------------------------------------------------------------------------
                 \2486\ 79 FR 23425.
                ---------------------------------------------------------------------------
                 Although GHG emissions reductions would be lessened under the
                standards finalized today compared to the previously issued EPA
                standards, in light of this assessment indicating higher vehicle costs
                and associated impacts on consumers, EPA believes that, on balance, the
                final standards
                [[Page 25114]]
                (Alternative 3) are justified and appropriate.
                (e) Consideration of Consumer Choice
                 EPA believes that consumer demand is an important consideration in
                setting CO2 emission standards, because one of EPA's goals
                in setting the standards has been and continues to be to allow
                manufacturers to provide, and consumers to purchase, vehicles with
                varying attributes and functionality rather than to shift demand to
                certain vehicle types or sizes. Societal and economic trends play a
                role in this area as well--if fuel prices are relatively high, demand
                for fuel-efficient vehicles increase and, as a result, compliance with
                standards is easier to achieve. If fuel prices are relatively low--as
                they are now and are projected to be in the mid-term--consumer demand
                for fuel-efficiency is less strong, making it harder for manufacturers
                to comply with the standard. While manufacturer difficulty in complying
                due to lack of consumer demand may not be the deciding factor in
                determining the appropriate levels of stringency for standards, it is
                relevant to understanding lead time difficulties, which EPA is required
                to consider under Section 202(a)(2).
                 As discussed previously, the EPA CO2 standards are based
                on vehicle footprint, and in general smaller footprint vehicles have
                individual CO2 targets that are lower (more stringent) than
                larger footprint vehicles. The passenger car fleet has footprint curves
                that are distinct from the light-truck fleet. One of EPA's goals in
                designing the footprint-based standards, in considering the shape,
                slope, and stringency of the footprint standard curves, and in adopting
                various compliance flexibilities (e.g., emissions averaging, banking,
                and trading, air-conditioning credits, off-cycle credits) was to
                maintain consumer choice. The EPA standards are designed to require
                reductions of CO2 emissions over time from the vehicle fleet
                as a whole, but also to provide sufficient flexibility to the
                automotive manufacturers so that firms can produce vehicles that serve
                the needs of their customers. The past several model years in the
                marketplace show that, while this approach reduces the impact of
                increased fuel economy on consumer choice, it does not adequately
                account for changes in consumer preference. As a result, as discussed
                throughout this preamble, manufactures are struggling to meet
                CO2 emission standards based upon their fleet performance.
                In fact, the 2017 model year saw that only three major manufacturers
                had fleets that met the standards. One reason behind these challenges
                is that, while the footprint-based attribute standards account for
                vehicle length and width, they do not account for vehicle height or
                weight. And, since many crossovers sold today are classified as
                passenger cars and not light trucks, the additional weight of such
                vehicles to provide for requisite ride height puts pressure on
                CO2 emission compliance for automaker passenger car fleets.
                Similarly, large SUVs are subject to the same footprint-based standards
                as lighter trucks, putting pressure on CO2 emission standard
                compliance. For the 2017 model year, 12 percent of the fleet consisted
                of car-based SUVs, and 32 percent of the fleet consisted of truck-based
                SUVs.\2487\ Taller and heavier vehicles, including crossovers and SUVs,
                are more popular today than was expected at the time the standards were
                set. While automobile manufacturers have continued to offer a broad
                range of vehicles (e.g., full-size pick-up trucks with high towing
                capabilities, minivans, cross-over vehicles, SUVs, and passenger cars;
                vehicles with off-road capabilities; luxury/premium vehicles,
                supercars, performance vehicles, entry level vehicles, etc.) despite
                continuing required increases in fuel economy stringency, this has
                largely been possible because of well-stocked over-compliance credit
                banks from when standards were less stringent and the ability to
                acquire credits from other manufacturers. As mentioned earlier, the
                agencies have concerns whether this is sustainable. Automotive
                companies have been able to reduce their fleet-wide CO2
                emissions while continuing to produce and sell the many diverse
                products that serve the needs of consumers in the market. The agencies
                recognize that automotive customers are diverse, that automotive
                companies do not all compete for the same segments of the market, and
                that increasing stringency in the standards can be expected to have
                different effects not only on certain vehicle segments but also on
                certain manufacturers that have developed market strategies around
                those vehicle segments. Taking into consideration this diversity of the
                automotive customer base, and of the strategies which have developed to
                meet specific segments, EPA concludes that the previous standards are
                not reasonable or appropriate.
                ---------------------------------------------------------------------------
                 \2487\ 2018 EPA Automotive Trends Report: Greenhouse Gas
                Emissions, Fuel Economy, and Technology since 1975, available at:
                https://www.epa.gov/automotive-trends/download-automotive-trends-report.
                ---------------------------------------------------------------------------
                 In the initial determination, EPA assessed several factors related
                to consumer choice, including the costs to consumers of new vehicles
                and fuel savings to consumers, as described above under Section
                VII.A.2.c). In 2017, EPA found that the previous standards would
                increase the upfront costs of vehicles but overall would have positive
                net benefits because lifetime fuel savings outweighed the lifetime
                vehicle costs for consumers. As discussed above, the costs of
                technology to comply with the standards are generally borne by the
                initial purchaser, with understanding of fuel cost implication given
                statutorily required disclosures. In contrast, the fuel savings are
                realized by many subsequent owners over the vehicles' lifetime, which
                this analysis assumes can be up to 39 years. New vehicle purchasers are
                not likely to place as much weight on fuel savings that will be
                realized by subsequent owners. Accordingly, EPA is placing greater
                weight on the up-front vehicle cost savings to consumers in light of
                the goal of accelerating the turnover of the motor vehicle fleet to
                safer cars that emit fewer criteria pollutants.
                 EPA received many comments regarding the agency's consideration of
                consumer choice in determining appropriate standards under section
                202(a) of the CAA. The Alliance commented that EPA's concerns regarding
                consumer choice are well founded, stating ``in the years since 2012
                (and in part due to the unexpected decrease in fuel prices), consumers
                have demonstrated less interest in high-efficiency/low-emission
                vehicles than EPA and NHTSA projected in issuing the 2012 Final Rule.
                As such, compliance with the existing standards would require a
                substantially greater variance than EPA expected from the vehicle fleet
                that consumers would otherwise choose.''
                 Global Automakers agreed that consumer acceptance is an important
                factor, but does not justify holding standards flat through the 2026
                model year. Global Automakers further commented that ``[f]uel economy
                remains a factor in vehicle purchase decisions, though perhaps not a
                dominant one.''
                 CBD and others commented that the Clean Air Act does not allow EPA
                to reduce stringency based upon consumer choice factors. They point to
                the diversity of the vehicle fleet and argue that EPA's consideration
                of projected tech levels and associated costs as ``speculative'' and
                not grounded in fact.
                 U.S. Congressman Mark DeSaulnier claimed that the justification for
                the proposal appeared to be consumer
                [[Page 25115]]
                willingness to buy new vehicles. He claimed that absent any standards
                whatsoever, automakers could produce more vehicles that consumers would
                want to purchase. He stated that the standards require all vehicles to
                become more efficient and that EPA has an overly simplistic
                understanding of American consumers, who, according to him, are ``wary
                of the price tag'' when shopping, but, nonetheless, ``overwhelmingly
                want more efficient vehicles, and they want to reduce the health burden
                of air pollution.''
                 The Institute for Policy Integrity (IPI) claims, without support,
                that as fuel efficiency technology is introduced and becomes
                widespread, consumer attitudes will change and will start focusing on
                such technology. IPI also claims that manufacturers can change consumer
                preference through advertising. IPI implies that manufactures play a
                larger role in shaping consumer options of their needs that consumers
                do themselves. IPI also comments that academic literature relating to
                demand- and supply-side obstacles to fuel economy indicates that the
                proposal's justification runs counter to available evidence.
                 The University of California Berkeley Environmental Law Clinic
                (Berkeley) argued against EPA's consideration of consumer choice in
                setting standards, claiming that low-income households bear exposure to
                operating costs, fuel price fluctuations, and environmental impacts.
                Berkeley also claimed that EPA's purported list of features consumers
                may favor over fuel economy is not supported by evidence, and, in any
                event, should be categorized into lists of ``needs'' versus ``wants.''
                 Consumer choice is a complex consideration when setting standards.
                As Congressman DeSaulnier correctly notes, EPA cannot disregard its
                consideration of public health and welfare based upon the agency-
                projected whims of consumers. At the same time, the willingness of
                consumers to pay for fuel economy improvements, which as described
                above affects vehicle performance and utility in a manner
                distinguishable from criteria pollutant emissions, has a direct effect
                upon the ability of manufacturers to sell their product. And as
                consumers demand vehicles with increased ride height (which, all else
                being equal, increases CO2 emissions), establishing
                standards that account for this--but still require manufacturers to
                focus on improving emission performance, is reasonable and appropriate.
                 In response to Global Automakers' comment that consumers do not
                heavily focus on fuel economy in making purchase decisions, EPA agrees,
                but notes that this is a consumer's choice, as federal law requires
                that consumers are made aware of fuel economy impacts, pursuant to 49
                U.S.C. 32908. EPA also agrees that the willingness to pay for fuel
                economy improvements is ``not zero.''
                 EPA agrees with the Global Automakers comment that while consumer
                choice is an important consideration in determining the appropriate
                level of the revised standards, the final rule analysis does not
                support holding the standards constant. Although EPA proposed standards
                at the level of 0 percent increase in stringency from MY 2021 and
                later, after considering the comments received and based on the updated
                analysis for this final rule, EPA is finalizing standards with a 1.5
                percent per year improvement in stringency from MY 2021 to MY 2026. As
                indicated in the comments on this topic, there is a range of views and
                relevant information concerning the extent of consumers' interest in
                fuel economy and on the role fuel savings plays in consumer purchase
                decisions.\2488\ EPA's understanding is that some consumers value fuel
                economy more than others, and EPA finds it unnecessary to identify the
                precise role of fuel economy in consumer purchase decisions because the
                Administrator believes that the standards should encourage a range of
                vehicles meeting a range of consumer preferences. Further, as described
                above, consumers are made aware of the relative fuel price impacts of
                new vehicles, given the required information label on new vehicles,
                thus indicating that, in all likelihood, consumers do take fuel
                expenses into account when making new vehicle purchase decisions.
                ---------------------------------------------------------------------------
                 \2488\ Studies of the role of fuel economy in consumer purchase
                decisions have found a wide range of values (Greene, D., A. Hossain,
                J. Hofmann, G. Helfand, and R. Beach. ``Consumer Willingness to Pay
                for Vehicle Attributes: What Do We Know?'' Transportation Research
                Part A 118 (2018), p. 258-79). The National Academy of Sciences in
                2015 judged that ``there is a good deal of evidence that the market
                appears to undervalue fuel economy relative to its expected present
                value, but recent work suggests that there could be many reasons
                underlying this, and that it may not be true for all consumers.''
                National Research Council of the National Academies (2015). Cost,
                Effectiveness, and Deployment of Fuel Economy Technologies for
                Light-Duty Vehicles. Washington, DC: National Academies Press, p. 9-
                16.
                ---------------------------------------------------------------------------
                 EPA disagrees with Congressman DeSaulnier's assertion that EPA
                seeks to set standards that do not affect what manufacturers produce--
                instead, the agencies examine what consumers are purchasing in the
                market to determine what standards are appropriate. The agency's
                assumptions in 2012--that consumers would gravitate toward the purchase
                of compact sedans and coupes in response to exceedingly high fuel
                prices--have proved incorrect. Fuel prices have fallen and remained
                relatively low, and are projected to remain relatively low throughout
                the period covered by this rulemaking. EPA seeks to achieve
                improvements in CO2 emissions, but it is not realistic to
                expect the high demand for crossover vehicles to abate, or for those
                vehicles to meet more-stringent standards set for compact sedans. That
                said, EPA agrees with Congressman DeSaulnier that American consumers
                are wary of the price of vehicles--popular reporting that consumers may
                reference explain affordability concerns in crisis terms--even
                indicating that the average price of a vehicle is now beyond that which
                is affordable to the median household income of every city outside of
                Washington, DC \2489\ This results in significant adverse economic
                impacts--higher finance charges, taxes, registration fees, and
                insurance costs, all of which result in challenges qualifying for
                financing and longer finance terms, which increase the likelihood of
                negative equity scenarios. EPA also agrees with Congressman DeSaulnier
                that consumers want increased fuel efficiency and to reduce the impacts
                of harmful air pollution. These are all true. But direct health impacts
                of vehicles emissions stem more from criteria pollutant emissions than
                from CO2 emissions. And CO2 emission technology
                has a significant relationship to the price of vehicles for which
                consumers are so wary. EPA, with this rulemaking, is attempting to
                strike the correct balance between a number of factors, including
                improving efficiency and affordability, which should yield additional
                sales and an improved rate of fleet turnover to vehicles that have
                better criteria pollutant emissions--particularly since the vehicles
                sold subject to this rulemaking will be sold during the phase-in of
                Tier 3 criteria pollutant emission standards.
                ---------------------------------------------------------------------------
                 \2489\ See., e.g., Car and Driver, ``For Middle-Class Shoppers,
                New Cars Are Moving out of Reach'' November 30, 2019. Available at:
                https://www.caranddriver.com/news/a30061910/middle-class-car-shoppers-priced-out/; New York Times, ``New Cars Are Too Expensive
                for the Typical Family, Study Finds'' July 2, 2016. Available at:
                https://www.nytimes.com/2016/07/02/your-money/new-cars-are-too-expensive-for-the-typical-family-study-finds.html.
                ---------------------------------------------------------------------------
                 In response to Berkeley, low-income consumers are even more
                sensitive to upfront vehicle purchase prices than they are to the
                smaller delta between weekly or monthly fuel costs
                [[Page 25116]]
                experienced over time between the previous standards and the standards
                finalized today--they may well take note of the fact that one cannot
                pay today's bills with tomorrow's savings. They may also want to take
                note that the standards finalized today are projected to improve fleet
                turnover into newer vehicles that emit reduced criteria pollutants.
                 EPA disagrees with the assertion by CBD and others that the agency
                has not provided a rationale for its consideration of consumer choice
                in determining the appropriate standards. EPA notes that despite a
                variety of vehicles on the market today and over the past several
                years, the fleet has failed to comply with standards based upon
                performance beginning with the 2016 model year, and has fallen further
                behind in the 2017 model year, when only three major automakers
                complied with CO2 emission standards based upon performance
                alone.
                 In response to IPI's comment that the deployment of more fuel-
                efficient technologies, combined with manufacturer advertising, will
                change consumer preference, this runs counter to historical trends.
                Manufacturers have continuously deployed additional fuel efficiency
                technology in each model year--which is why EPA continues to see
                fleetwide improvements in CO2 emissions on new vehicles. And
                manufacturers have consistently advertised the fuel economy performance
                of their vehicles. Federal law requires the physical posting fuel
                economy performance, as well as estimated and comparative fuel cost
                information, on every new vehicle offered for sale. Notwithstanding
                this activity, consumer demand, and willingness to pay for technology
                that reduces CO2 emissions and improves fuel economy, has not matched
                required standards--which is one of the reasons that EPA is revising
                the standards today. As discussed in the proposal, EPA recognizes that
                the diversity in the automotive customer base, combined with the facts
                and analysis developed by the agency in this rulemaking, raises
                concerns that the previous standards, if they are not adjusted, may not
                continue to fulfill the agency's goal of providing sufficient
                manufacturer flexibility to meet consumer needs and consumer choice
                preferences in their vehicle purchasing decisions. In the 2012 Final
                Rule and the Initial Determination, EPA expected that consumers would
                readily accept fuel-saving technologies in their new vehicles, despite
                the agency's uncertainty about the role of fuel savings in consumers'
                purchase decisions. Given low fuel prices and the pronounced market
                shift to crossovers and SUVs, notwithstanding required disclosers of
                fuel costs and relative fuel economy performance, EPA now concludes
                that it is appropriate to account for the shift in consumer preference
                in concluding that the standards set in 2012 did not provide sufficient
                lead time for manufacturers to achieve the standards set at that time.
                EPA remains concerned that the projected level of hybridization and
                other advanced technologies and the associated vehicle costs necessary
                to achieve the previous standards are too high from a consumer-choice
                perspective, and not sufficiently account for consumer acceptance of
                such technology. While consumers have benefited from improvements over
                several decades in traditional vehicle technologies, such as
                advancements in transmissions and internal combustion engines,
                electrification technologies are a departure from what consumers have
                traditionally purchased. Strong hybrid and other advanced
                electrification technologies have been available for many years (20
                years for strong hybrids and eight years for plug-in and all electric
                vehicles), and sales levels have been relatively low, in the 2-3
                percent range.\2490\ As discussed above, the analysis projects that the
                2012 EPA standards would be projected to require a significant increase
                in hybridization (up to 8 percent for mild hybrids and 10 percent for
                strong hybrids in MY 2030). This large increase in technology demand
                over the next decade could lead to automotive companies needing to
                change the choice of vehicle types they are able to offer to consumers,
                compared to what the companies would otherwise have offered in the
                absence of the previously issued standards. As discussed above,
                manufacturers are, by and large, not meeting existing standards based
                upon actual fleet performance in CO2 emissions and are
                instead relying upon the use of earned or acquired credits. As the
                previous standards were set to increase significantly through MY 2020
                and thereafter, reducing the rate of increase is appropriate and
                reasonable. Doing so will provide manufacturers with sufficient lead
                time to meet the standards being set today.
                ---------------------------------------------------------------------------
                 \2490\ For instance, the 2019 calendar year saw only a 1.4%
                penetration of battery electric vehicles in the light duty fleet,
                following 1.2% for 2018, 0.6% for 2017, 0.5% for 2016, and 0.4% for
                2015. Wards Auto Monthly Sales reports, available at https://wardsintelligence.informa.com/.
                ---------------------------------------------------------------------------
                 EPA recognizes that one possibility for automotive companies who
                wish to retain their current vehicle offerings, but face compliance
                challenges is to purchase GHG emissions credits. In EPA's annual
                Automotive Trends Report, EPA has reported that credit trading has
                occurred frequently in the past several years to achieve compliance
                with the GHG standards.\2491\ Credit trading can lower a manufacturer's
                costs of compliance, both for those selling and those purchasing
                credits, and this program compliance flexibility is another tool
                available to auto firms to allow them to continue offering the types of
                vehicles that customers want. Between MY 2010 and MY 2017, these trades
                have included 11 firms, with five firms selling CO2 credits
                to seven firms.\2492\ The number of firms participating in the GHG
                credits market represents about one-half of the automotive companies
                selling vehicles in the U.S. market, but since several of these firms
                are small players, they represent less than half of the vehicle
                production volume. In total, approximately 48 million Megagrams of
                CO2 credits have been traded between firms, which represents
                19 percent of the MY 2017 industry-wide bank of credits. That said,
                more manufacturers have relied upon previously earned credits to
                achieve compliance. Between MY 2010 and MY 2017, 80% of firms applied
                previously earned credits. However, long-term planning is an important
                consideration for automakers, and an automaker who may need to purchase
                credits as part of a future compliance strategy is not guaranteed to
                find credits. The automotive industry is highly competitive, and firms
                may be reluctant to base their future product strategy on an uncertain
                future credit availability, but face struggles in achieving
                CO2 emission reductions in a manner that meets consumer
                expectations for cost, utility, and performance. Also, pools of
                available credits continue to decline over time as the standards become
                more stringent and previously banked credits are either used or expire;
                indeed, this has happened in recent years.\2493\ EPA's views on the
                availability of the credit market to aid in manufacturers' compliance
                have changed since the Initial Determination. Based upon the
                information available to the EPA in early January 2017, the auto
                industry had outperformed its standards in the four previous compliance
                years (MYs 2012-2015) and EPA had viewed that as
                [[Page 25117]]
                a positive trend.\2494\ Since then, however, overall manufacturer
                performance failed to meet the standard fleetwide, and many
                manufacturers relied on credits to meet their individual compliance
                targets. Furthermore, recent experience suggests that availability of
                the credit bank is becoming a more uncertain means to achieve
                compliance.\2495\ Thus, while credit trading may be a useful
                flexibility to reduce the overall costs of the program and to smooth
                the pathway to compliance realizing necessary transitions from vehicle
                redesign cycles, EPA believes it is important to set standards that
                preserve consumer choice without relying on credit purchasing
                availability as a compliance mechanism. As discussed in Section VII,
                the agencies project that the EPA final standards (Alternative 3, 1.5
                percent year over year stringency improvement), will require more
                realistic penetration of advanced CO2 emission technologies
                such as electrification--better ensuring that manufacturers will be
                able to provide vehicles that meet consumer demand.
                ---------------------------------------------------------------------------
                 \2491\ 2018 EPA Automotive Trends Report at Figures 5.15 and
                5.17.
                 \2492\ EPA Greenhouse Gas Emission Standards for Light-Duty
                Vehicles: Manufacturer Performance Report for the 2016 Model Year.
                EPA-420-R-18-002. January 2019.
                 \2493\ 2018 EPA Automotive Trends Report at Figure 5.17 and
                Table 5.17.
                 \2494\ See Initial Determination at page 7-8.
                 \2495\ Id. at Figure ES-8.
                ---------------------------------------------------------------------------
                (f) Consideration of Safety
                 As discussed above, EPA has long considered the safety implications
                of its emission standards.\2496\ More recently, EPA has considered the
                potential impacts of emission standards on safety in past rulemakings
                on GHG standards, including the 2010 rule which established the 2012-
                2016 light-duty vehicle GHG standards, and the 2012 rule which
                previously established 2017-2025 light-duty vehicle GHG standards.
                Indeed, section 202(a)(4)(A) specifically prohibits the use of an
                emission control device, system or element of design that will cause or
                contribute to an unreasonable risk to safety.\2497\ The relationship
                between CO2 emissions and safety is more nuanced. Safety
                impacts relate to changes in the use of vehicles in the fleet, relative
                mass changes, and the turnover of fleet to newer and safer vehicles.
                ---------------------------------------------------------------------------
                 \2496\ See, e.g., 45 FR 14496, 14503 (1980) (``EPA would not
                require a particulate control technology that was known to involve
                serious safety problems.'').
                 \2497\ 42 U.S.C. 7521(a)(4)(A).
                ---------------------------------------------------------------------------
                 The analysis for the final rule projects that there will be a
                change in vehicle miles traveled (VMT) under the final standards,
                specifically 607 billion less miles traveled compared to the previous
                standards case. Based on these projections about reduced VMT in the
                light-duty fleet, the analysis estimates that fatalities will be
                reduced by 2584 (out of a total impact of 3269) over the lifetime of MY
                1977-2029 vehicles compared to the previous CO2
                standards.\2498\ In other words, the reduction in fatalities under the
                final standards compared to the previous standards is primarily driven
                by the modeling's projected changes in VMT and associated changes in
                mobility (i.e., people driving less). The details of the safety
                assessment are discussed in Section VI of this preamble and in Section
                VI of the FRIA. Under alternatives with stringency levels lower than
                the final standards, the analysis projects greater reductions in VMT,
                and thus projects somewhat greater reductions in fatalities based on
                these VMT changes. Under alternatives with stringency levels higher
                than the final standards, the analysis projects lower reductions in
                VMT, and thus projects fewer fatalities reduced, See Table VI-271.
                ---------------------------------------------------------------------------
                 \2498\ The number of fatalities projected is a product of two
                contributing factors: the number of miles driven (VMT) and the risk
                of driving (i.e., fatalities per mile). Overall in this final rule
                analysis, the change in fatalities projected is primarily caused by
                the changes in VMT.
                ---------------------------------------------------------------------------
                 EPA notes that the magnitude of the changes in fatalities stemming
                from changes in mobility projected in this final rule is less than what
                was presented in the proposed rule. In response to comments, the
                agencies took a conservative approach to modeling the effects of
                standard stringency upon safety. The agencies held VMT constant across
                alternatives. The reasons for the differences in fatality estimates in
                the final rule compared to the proposed rule, including changes to the
                modeling inputs and projections based on the agencies' assessment of
                public comments.
                 The approach for reporting fatality impacts for this final rule is
                different than the previous analyses for the Initial Determination and
                the 2012 rulemaking. First, the analysis quantifies the number of
                fatalities caused by changes in VMT between each Alternative and the
                previous standards, whereas previous analyses did not. Second, the
                safety analysis itself is different from previous analyses that assumed
                that automakers would not reduce the weight of approximately the
                lightest half of passenger cars--discounting the safety impacts of mass
                reduction. Third, while the agencies qualitatively discussed the effect
                of price increases attributable to increased stringency on vehicle
                sales, fleet turnover, and the improved safety of newer vehicles, the
                agencies never attempted to quantify these impacts.
                 With respect to public comments, the Alliance commented that ``EPA
                has discretion to consider all the relevant factors in setting
                appropriate emissions standards under Sec. 202(a)(1), including
                vehicle safety. Moreover, given NHTSA's greater expertise in evaluating
                motor vehicle safety, it is appropriate for EPA to respect the views of
                its companion agency on those issues.'' The Alliance commented that
                ``[t]he new safety analysis likewise provides support for EPA's
                conclusion that the MY 2021-2025 GHG standards are not appropriate and
                should be reduced in stringency. Indeed, given that the `primary
                purpose' of Sec. 202(a)(1) is `the protection of public health and
                welfare,' EPA would be abdicating its statutory duty if it ignored
                these concerns.''
                 Global Automakers commented that safety impacts due to the rebound
                effect should not be attributed to the standards and should not serve
                as a basis for keeping the standards flat. They further argued that the
                dynamic scrappage model is flawed and should be removed from the
                modeling for purposes of the final rule. They also argued, that
                Congress expressed interest in improving efficiency, emissions, and
                safety (without no recognition of cost as a factor), and that
                therefore, improvement in all such areas should provide that
                improvements in efficiency would not lead to negative safety impacts.
                 CBD and others commented that safety concerns should not be
                considered because the record does not indicate that vehicles must be
                unsafe to meet the previous standards. They further commented that EPA
                cannot justify reduced stringency upon ``rebound'' fatalities, and they
                argue that those fatalities cannot be considered by EPA, since they
                ``stem from voluntary choices by individuals to drive more--not the
                `operation or function'of the technologies at issue'' (quoting CAA
                Section 202(a)(4)(A)).
                 Environmental Defense Fund (EDF) similarly commented that the
                estimates of fatalities are unsound, as is considering total fatalities
                resulting from increased stringency, rather than fatality rates. They
                added that the projected fatalities stem from consumer and manufacture
                behaviors that are removed from the stringency requirements. They
                further argue that considering fatalities that are attributable to the
                standards--particularly rebound fatalities--are inappropriate. EDF,
                UCS, and Consumers Union argue that fatalities attributable to
                increased driving are not relevant to agency decisions.
                 In response to the Alliance comments, EPA has considered safety, as
                described in this section, and agrees that the
                [[Page 25118]]
                potential impacts of emission standards on safety is an important
                consideration in determining appropriate standards under CAA section
                202(a). In response to comments from Global Automakers that the safety
                analysis in the proposed rule did not support freezing the standards,
                EPA agrees that safety considerations alone do not justify such an
                approach, and notes that the safety analysis performed for this final
                rule has changed from the analysis for the proposed rule based on
                consideration of public comments. EPA is finalizing standards that are
                more stringent (1.5 percent per year stringency improvement for MY
                2021-2026) than the proposed rule's preferred alternative (0 percent
                stringency improvement).
                 Several commenters argued that the proposal's claims of reduced
                fatalities were based upon projected changes in driving, arguing that
                that EPA should not decide the level of the standards based on these
                assumed changes in travel. As discussed above, EPA acknowledged that
                the reduction in fatalities under the final standards compared to the
                previous standards are in large part driven by projected changes in
                driving behavior (i.e., people driving less). While EPA is not seeking
                to restrict mobility or driving, ignoring impacts associated with this
                rule would be inappropriate. Moreover, the provisions of Section
                202(a)(4) do not preclude EPA from considering such impacts. While EPA
                has considered the safety assessment for this final rule, as discussed
                in the following section below, safety was one of several factors
                considered in deciding on the level of today's final standards.
                g) Consideration of Energy Security Impacts
                 Among other factors EPA considered in selecting the previous
                standards in the 2012 Final Rule was the effect of the standards on
                U.S. petroleum imports and energy security.\2499\ As discussed in the
                PRIA, Final RIA and in Section Energy Security, the energy security
                position of the United States has changed dramatically since 2012. The
                U.S. has become a net exporter of petroleum and additional payments by
                United States consumers resulting from upward pressure on oil price due
                to additional demand are a transfer that occurs within the United
                States economy.\2500\ Additional petroleum use necessarily increases
                demand and thus subjects the nation to additional risk of price shocks,
                but this risk is significantly reduced as the United States has
                dramatically increased domestic petroleum production and has additional
                capacity to do so. Accordingly, energy security concerns are reduced
                compared to the assessment in the 2012 rulemaking and do not alter
                EPA's selection of final revised standards in this rule.
                ---------------------------------------------------------------------------
                 \2499\ See 77 FR 62938, et seq.
                 \2500\ The U.S. Energy Information Administration EIA estimates
                that the United States exported more total crude oil and petroleum
                products in September and October 2019, and expects the United
                States to continue to be a net exporter. See Short Term Energy
                Outlook November 2019, available at https://www.eia.gov/outlooks/steo/archives/nov19.pdf.
                ---------------------------------------------------------------------------
                (h) Balancing of Factors and EPA's Revised Standards for MY 2021 and
                Later
                 As discussed in this section, the Administrator is required to
                consider a number of factors when establishing emission standards under
                section 202(a)(2) of the Clean Air Act: The standards ``shall take
                effect after such period as the Administrator finds necessary to permit
                the development and application of the requisite technology, giving
                appropriate consideration to the cost of compliance within such
                period.'' \2501\ For this Final Rule, the Administrator has considered
                a wide range of potential emission standards (Baseline/No Action
                Alternative and Alternatives 1 through 7), ranging from the previous
                EPA standards (Baseline/No Action Alternative), through a number of
                less stringent alternatives, including the proposed preferred
                alternative (Alternative 1, 0 percent per year stringency improvement)
                and what has been chosen as the final standards (Alternative 3, 1.5
                percent per year stringency improvement). The Administrator has
                determined that the revised final standards, which would increase the
                stringency of the MY 2020 standards by 1.5 percent per year for both
                passenger cars and light-trucks from MY 2021 through 2026, are
                appropriate under section 202(a) of the CAA. In addition to
                technological feasibility, lead-time, and the costs of compliance, the
                Administrator has also considered the impact of the standards on GHG
                and non-GHG emissions reductions, the costs to consumers, and vehicle
                safety.
                ---------------------------------------------------------------------------
                 \2501\ 42 U.S.C. 7521(a)(2).
                ---------------------------------------------------------------------------
                 In addition to comments on each of the factors the Administrator
                considered discussed above, comments also were received on how the
                Administrator should balance these factors in determining the
                appropriate final standards.
                 The Alliance commented that the CAA provides EPA with significant
                latitude to exercise its expert judgment in determining the level at
                which emissions standards should be set. The Alliance commented further
                that unlike other CAA provisions, Sec. 202(a)(1) does not require EPA
                to set standards that will result in the greatest degree of emission
                reduction achievable. Instead, the statute leaves EPA flexibility to
                decide what factors are relevant, and how to weigh those factors, in
                its decision-making process. The Alliance also commented ``EPA also has
                'significant latitude' regarding the 'coordination of its regulations
                with those of other agencies,' '' ``EPA has discretion to defer to the
                judgment of other agencies regarding issues within their areas of
                expertise,'' and the CAA ``gives the agency authority to engage in
                reasoned decision-making, balancing all of the relevant factors in
                light of the available facts. EPA has done that here and has provided a
                reasoned explanation of its determination that the environmental
                benefits of the existing MY 2021-2025 GHG standards are outweighed by
                their negative effects on costs and safety.''
                 The American Iron and Steel Institute commented that it favors the
                general direction taken in the SAFE proposal, including the preferred
                option for CO2 standards, and that it believes a final SAFE
                rule that ``balances the priorities of costs to consumers, safety
                design considerations, employment impacts and total GHG emissions will
                result in the best outcome.''
                 CBD and others claimed that the justifications EPA offered in the
                notice are untethered from the statute, and that EPA used a flawed
                analysis. Further, they claim that EPA did not exercise its own
                judgment and delegated its responsibilities impermissibly to NHTSA,
                failing to consider ``relevant EPA information.''
                 EPA's analysis is described in detail in this preamble. EPA decided
                to use the CAFE model for a number of reasons, described in more detail
                in Section IV, including that using two models results in an
                inefficient use of resources, the CAFE model can analyze both EPA's and
                NHTSA's statutory programs, the CAFE model is capable of modeling
                incremental improvements of discrete technologies, and EPA believes
                that the CAFE model provides reasonable results. Merely because EPA has
                a set of its own analytical tools that model similar effects does not
                mean that it must use those tools to perform the analysis, and doing so
                would create unnecessary complication and lead to potential
                inconsistencies. Since the agencies are establishing standards jointly
                and seeking to avoid
                [[Page 25119]]
                inconsistencies in a manner consistent with Supreme Court direction,
                using the same model for the analysis is reasonable. Nonetheless, EPA
                has exercised its own judgment in this final rule.
                 The California Attorney General and others claim that EPA failed
                adequately to acknowledge, explain, or justify its departure from the
                prior determination. They claim that EPA failed to propose or make a
                finding required by Section 202(a)(2) relating to adequate lead time,
                inconsistent with EPA's prior explanation that it is provided with
                limited flexibility in making such a determination.
                 The California Attorney General and others also claim that EPA's
                analysis improperly weighs the factors it considers, and that it
                insufficiently weighed certain factors required under the Clean Air
                Act, including air pollution. In response, EPA notes that the Clean Air
                Act does not specify how the Administrator should weigh the factors
                considered, as discussed elsewhere in this section.
                 The California Attorney General and others further noted that the
                purpose of the Clean Air Act is to is to ``protect and enhance the
                quality of the Nation's air resources so as to promote the public
                health and welfare and the productive capacity of its population.''
                 The Institute for Policy Integrity claimed that the agencies
                balanced the factors in a way that conflicts with their controlling
                statutes and weighed the statutory factors without regard for the
                accuracy of the accompanying cost-benefit analysis.
                 The National Coalition for Advanced Transportation claimed that the
                proposal appeared to be based on heightened concerns with cost,
                consumer acceptance, and safety, and insufficiently on technology
                availability and emissions reductions. As discussed in this section,
                EPA is neither relying solely on cost or safety nor ignoring any
                factors, but rather is balancing a number of factors.
                 Green Energy Institute at Lewis and Clark Law School et al.
                commented that the Clean Air Act does not authorize the weakening or
                freezing of existing standards due to industry costs or consumer
                preferences. While EPA has broad discretion to revise standards based
                upon a balancing of factors, the final rule will provide for increasing
                stringency of 1.5 percent per year from MY 2021 through MY 2026.
                 Motor & Equipment Manufacturers Association (MEMA) commented that
                the technology costs from their preferred alternative (Alternative 8 in
                the notice) were not significant and did not justify holding MY 2020
                standards flat in light of other elements, such as preserving
                investments in fuel saving technology. EPA disagrees, and considers the
                reductions in costs resulting from the revised final standards, $1,250
                per vehicle by MY 2029, to be one important aspect of the justification
                of these standards.
                 EPA believes the previously issued standards for MY 2021 and later,
                considered as a whole, are too stringent. Factors in favor of reduced
                stringency include manufacturer compliance costs, and the related per-
                vehicle cost savings. As described above, the agencies project that the
                final CO2 standards will reduce manufacturers' MY 2018-2029
                compliance costs by $108 billion (when applying a 3% discount rate),and
                will reduce average MY 2030 vehicle prices $977 (also applying a 3%
                discount rate). Including other costs, such as financing and insurance,
                consumers the standards finalized today will result in reduced costs of
                $1,286 per-vehicle for a MY 2030 vehicle. EPA expects that the final
                standards will not impede consumers from being able to purchase a new
                vehicle of their choice or require significant changes in product lines
                for any manufacturer. In fact, under the final standards, vehicle sales
                are expected to increase by 2.2 million vehicles over MY 2017-2029
                compared to projected sales under the augural standards, a significant
                increase in vehicles sold over this timeframe see Table VI-155. EPA
                views this projection of vehicle sales increases resulting from the
                final standards as important in facilitating the turnover of the fleet
                to newer, safer vehicles, all of which will be subject to increasingly
                stringent criteria pollutant emission requirements as federal Tier 3
                emission standards continue to phase in from MY 2017 through MY 2025.
                 Another factor weighing toward reduced stringency is safety. As
                discussed previously, reduced stringency results in less pressure on
                manufacturers to reduce mass in vehicles, which, for smaller passenger
                cars has negative safety implications when involved in accidents with
                heavier vehicles. Further, as vehicle prices decrease compared to the
                previous standards, more consumers will be able to afford newer
                vehicles, which are significantly safer. Lastly, as vehicles will not
                be required to be as fuel efficient as under the previous standards,
                ``rebound'' driving will be reduced. The agencies project a reduction
                in 605 billion miles traveled by light-duty vehicles produced through
                MY 2029, and project that this reduced VMT will lead to 2,584 fewer
                highway fatalities under the final standards compared to the previous
                CO2 standards (i.e., people are projected to drive less
                under the final standards with an associated reduction in driving-
                related fatalities). While, notwithstanding EPA's involvement with
                State and local Transportation Control Measures (TCMs), the
                Administrator does not seek to change the way people drive--EPA's
                intention is not to restrict mobility, or to discourage driving, based
                on the level of the standards--EPA nonetheless believes it is
                appropriate to consider this projection.\2502\ The agencies also
                project that accelerated fleet turnover attributable to the change in
                standards will lead to the avoidance of a further 447 fatalities, and
                that the reduced need for reductions of vehicle mass will lead to the
                avoidance of a further 238 fatalities. In other words, the agencies
                project that the change in CO2 standards will lead to 3,269 fewer
                fatalities over the useful lives of vehicles produced through MY 2029.
                ---------------------------------------------------------------------------
                 \2502\ Information regarding TCMs is available at https://www.epa.gov/statelocalenergy/transportation-control-measures.
                ---------------------------------------------------------------------------
                 Factors that weigh in favor of increased stringency options are
                increased upstream criteria pollutant emissions attributable to
                additional refining and other fuel-related activities, as well as
                increased CO2 emissions and consumer fuel expenditures.
                 As described above, the agencies project that the revised final
                standards will have a negative impact on air quality health outcomes,
                including a projected increase of 444 to 1,000 premature deaths from
                increased air pollution over the lifetime of the MY 1977-2029 vehicles
                on the road after calendar year 2017 cumulative through CY 2068, under
                EPA's CO2 program.\2503\ EPA recognizes that the final
                standards are projected to increase CO2 emissions compared
                to the previous EPA standards. However, EPA notes that, unlike other
                provisions in Title II referenced above, section 202(a) does not
                require EPA to set standards for light-duty vehicles which result in
                the ``greatest degree of emission reduction achievable.'' EPA has not
                chosen the standard that has the highest estimated net social benefits.
                However, as discussed elsewhere in this preamble, from a cost benefit
                perspective, the differences among the various alternatives are
                relatively narrow. EPA believes consideration of costs and benefits is
                certainly relevant to its
                [[Page 25120]]
                exercise of discretion in selecting appropriate standards, but also
                recognizes that some costs and benefits are difficult to quantify, and
                additional factors can prove material under the Clean Air Act as well
                in those policy decisions. For example, EPA notes that the agency
                decided against pursuing more stringent alternatives analyzed in both
                the rulemaking establishing 2012-2016 standards and the rulemaking
                establishing 2017-2025 standards.
                ---------------------------------------------------------------------------
                 \2503\ The agencies believe that these premature mortality
                estimates may be over-estimated. Please see more detailed
                discussions in Sections VI.D.3.d) and VIII.A.3.d) in this preamble,
                and similar discussions in the final regulatory impact analysis.
                ---------------------------------------------------------------------------
                 EPA has also given weight to the policy goal of establishing
                CO2 standards which are coordinated with NHTSA's CAFE
                standards. While not a statutory requirement, EPA has considered the
                importance of having coordinated and harmonized EPA CO2 and
                CAFE programs, while recognizing the different statutory authorities
                for those programs, since the establishment of the EPA CO2
                program. The agencies discussed the importance of having one national
                program in the SAFE Vehicles Part 1 joint action.\2504\ In today's
                joint final rule, DOT is establishing CAFE standards for MY 2021-2026
                which increase in stringency at a level of 1.5 percent per year. The
                revised EPA standards will also increase in stringency at a rate of 1.5
                percent per year. Coordinating revisions to the GHG and CAFE standards
                in order to maintain one national program is a factor the Administrator
                has consideration in determining the revised GHG standards.
                ---------------------------------------------------------------------------
                 \2504\ 84 FR 51,310 (Sept. 27, 2019).
                ---------------------------------------------------------------------------
                 In light of available statutory discretion and the range of factors
                that the statute authorizes and permits the Administrator to consider,
                and his consideration of the factors discussed above, the EPA concludes
                that reducing the stringency of the MY 2021-2026 standards is an
                appropriate approach under section 202(a). Therefore, based on the data
                and analysis detailed in this final rule, the Administrator concludes
                that the previous MY 2021 and later CO2 standards are too
                stringent, and is establishing revised standards for MY 2021 through MY
                2026 at a level of 1.5 percent per year improvement in stringency.
                 In response to comments concerned about EPA's proposal to freeze
                the MY 2021-2026 standards at MY 2020 levels, EPA notes that it is
                finalizing the 1.5 percent per year improvement in stringency level and
                not the 0 percent improvement level proposed, after considering the
                somewhat higher costs to industry and up-front vehicle costs to the
                consumer and slightly lower GHG emissions and health-related impacts
                compared to the proposed preferred alternative. The Administrator has
                taken these tradeoffs into account in his balancing of factors under
                section 202(a) of the CAA.
                 While the set of factors considered by EPA under section 202(a) of
                the CAA in today's final rule and under the midterm evaluation
                regulations \2505\ in the Initial Determination are similar and
                overlapping, the Administrator recognizes that he is balancing these
                factors differently in this final rule than in the Initial
                Determination. In the Initial Determination, EPA's decision that the
                previous MY 2022-2025 standards were appropriate was based on
                conclusions that the standards were feasible within the lead time
                provided at reasonable costs, the standards would result in significant
                reductions in GHG emissions and oil consumption and associated fuel
                savings for consumers, and the standards would yield significant
                benefits to public health and welfare and positive net benefits
                overall, without adverse impacts on industry, safety, or
                consumers.\2506\
                ---------------------------------------------------------------------------
                 \2505\ 40 CFR 86.1818-12(h).
                 \2506\ Initial Determination, Section III, page 29-30.
                ---------------------------------------------------------------------------
                 Since the Initial Determination, EPA has completed its compliance
                review of the first two model years covered by the 2012 final rule.
                Notwithstanding widespread availability of vehicles that meet or exceed
                their CO2 emission targets, consumers are not expressing
                sufficient interest in fuel economy in their purchasing decisions to
                enable manufacturers to meet the standards based upon fleet
                performance. Although manufacturers earned significant credits in the
                early years of the agency's CO2 regulation history, these
                credits are being applied broadly across the industry and well in
                advance of the more aggressive model year stringency increases. While
                some manufacturers, including alternative fuel automakers are earning
                significant tradable credits, they do not have to trade them. And
                building a program around the potential for acquiring credits from
                competing manufacturers is not the intention of this action. While EPA
                is analyzing the differences between these standards and the previous
                standards for this rulemaking, EPA cannot ignore that this rulemaking
                was foreseen in the 2012 rulemaking. The prospect of revising the
                standards was expressly envisioned in that rulemaking based upon the
                uncertainty in the assumptions and future projections at that time.
                When viewed from the perspective of the larger set of MY 2017 through
                MY 2026 standards rulemakings, the standards finalized today fit the
                pattern of gradual, tough, but feasible stringency increases that take
                into account real world performance, shifts in fuel prices, and changes
                in consumer behavior toward crossovers and SUVs and away from more
                efficient sedans. This approach ensures that manufacturers are provided
                with sufficient lead time to achieve standards, considering the cost of
                compliance.
                 In this final rule, the EPA is placing greater weight on the costs
                to industry and the up-front vehicle costs to consumers. EPA believes
                that the costs to both industry and automotive consumers would have
                been too high under the previous standards, and that the standards
                should be revised to be less stringent to lower these costs. EPA
                believes that by lowering the auto industry's costs to comply with the
                program, with a commensurate reduction in per-vehicle costs to
                consumers, the final rule is enhancing the ability of the fleet to turn
                over to newer, cleaner and safer vehicles.
                 EPA believes that the characteristics and impacts of these and
                other alternative standards generally reflect a continuum in terms of
                technical feasibility, cost, lead time, consumer impacts, emissions
                reductions, and oil savings, and other factors evaluated under section
                202(a). In determining the appropriate standard to adopt in this
                context, EPA judges that the final standards are appropriate and
                preferable to more stringent alternatives based largely on
                consideration of cost--both to manufacturers and to consumers--and the
                potential for overly aggressive penetration rates for advanced
                technologies relative to the penetration rates seen in the final
                standards, especially in the face of an unknown degree of consumer
                acceptance of both the increased costs and of the technologies
                themselves--particularly given current projections of fuel prices
                during that timeframe. At the same time, the final rule helps to
                address these issues by maintaining incentives to promote broader
                deployment of advanced technologies, and so provides a means of
                encouraging their further penetration while leaving manufacturers
                alternative technology choices. EPA thus judges that more stringent
                alternatives, which would necessitate even more technology and more
                cost, would not be appropriate. Instead, EPA is adopting a more gradual
                increase in stringency to ensure that the benefits of reduced GHG
                emissions are achieved without the potential for disruption to
                automakers or consumers.
                [[Page 25121]]
                B. NHTSA's Statutory Obligations and Why the Selected Standards Are
                Maximum Feasible as Determined by the Secretary
                 In this section, NHTSA discusses the factors, data and analysis
                that the agency has considered in the selection of the CAFE standards
                for MYs 2021 and later and the comments received on NHTSA's
                consideration of these factors (see further discussion below on NHTSA's
                summary and analysis of comments).
                 As discussed in more detail below, the primary purpose of EPCA, as
                amended by EISA, and codified at 49 U.S.C. chapter 329, is energy
                conservation, and fuel economy standards help to conserve energy by
                requiring automakers to make new vehicles travel a certain distance on
                a gallon of fuel.\2507\ The goal of the CAFE standards is to conserve
                energy, while taking into account the statutory factors set forth at 49
                U.S.C. 32902(f), as discussed below.
                ---------------------------------------------------------------------------
                 \2507\ While individual vehicles need not meet any particular
                mpg level, as discussed extensively elsewhere in this preamble, it
                is broadly true that fuel economy standards require vehicle
                manufacturers' fleets to meet certain fuel economy levels as set
                forth by NHTSA in regulation.
                ---------------------------------------------------------------------------
                 49 U.S.C. 32902(f) states when setting maximum feasible CAFE
                standards for new vehicles, the Secretary of Transportation \2508\
                ``shall consider technological feasibility, economic practicability,
                the effect of other motor vehicle standards of the Government on fuel
                economy, and the need of the United States to conserve energy.'' In
                previous rulemakings, including the 2012 final rule that established
                CAFE standards for MY 2021 and set forth augural standards for MYs
                2022-2025, NHTSA considered technological feasibility, including the
                availability of various fuel-economy-improving technologies to be
                applied to new vehicles in the timeframe of the standards depending on
                the ultimate stringency levels, and also considered economic
                practicability, including the differences between a range of regulatory
                alternatives in terms of effects on per-vehicle costs, industry-wide
                costs, the ability of both the industry and individual manufacturers to
                comply with standards at various levels, as well as effects on vehicle
                sales, industry employment, and consumer demand. NHTSA also considered
                how compliance with other motor vehicle standards of the Government
                might affect manufacturers' ability to meet CAFE standards represented
                by a range of regulatory alternatives, and how the need of the U.S. to
                conserve energy could be more or less met under a range of regulatory
                alternatives, in terms of considerations like costs to consumers, the
                national balance of payments, environmental implications like climate
                and smog effects, and foreign policy effects like the likelihood that
                U.S. military and other expenditures could change as a result of more
                or less oil consumed by the U.S. vehicle fleet. These elements are
                discussed in detail throughout this analysis. As will be discussed in
                greater detail below, while NHTSA is considering all of the same
                factors in setting today's CAFE standards that it considered in
                previous rulemakings, and in many instances in a similar way as it
                considered those factors in previous rulemakings, the facts on the
                ground have changed and NHTSA is therefore choosing to set CAFE
                standards at a different level from what the 2012 final rule set forth.
                ---------------------------------------------------------------------------
                 \2508\ By delegation, NHTSA.
                ---------------------------------------------------------------------------
                 NHTSA is not limited to consideration of the factors specified in
                49 U.S.C. 32902(f) when establishing CAFE standards for passenger cars
                and light trucks. In addition to the factors enumerated above, NHTSA
                may (and historically has) considered such factors as safety and the
                environment.
                 NHTSA also considers relevant case law. Critical to this series of
                joint rulemakings with EPA, the Court in Massachusetts v. EPA,\2509\
                recognized EPA's argument that ``it cannot regulate carbon dioxide
                emissions from motor vehicles'' without ``tighten[ing] mileage
                standards . . . .''--a task assigned to DOT. The Court found that
                ``[t]he two obligations may overlap, but there is no reason to think
                the two agencies cannot both administer their obligations and yet avoid
                inconsistency.'' \2510\ Accordingly, the agencies have worked closely
                together in setting standards, and many of the factors that NHTSA
                considers to set maximum feasible standards overlap with factors that
                EPA considers under the Clean Air Act. Just as EPA considers energy use
                and security, NHTSA considers these factors when evaluating the need of
                the nation to conserve energy, as required by EPCA. Just as EPA
                considers technological feasibility, the cost of compliance,
                technological cost-effectiveness and cost and other impacts upon
                consumers, NHTSA considers these factors when weighing the
                technological feasibility and economic practicability of potential
                standards. EPA and NHTSA both consider implications of the rulemaking
                on CO2 emissions as well as criteria pollutant emissions.
                And, NHTSA's role as a safety regulator inherently leads to the
                consideration of safety implications when establishing standards. The
                balancing of competing factors by both EPA and NHTSA are consistent
                with each agency's statutory authority and recognize the overlapping
                obligations the Supreme Court pointed to in directing collaboration.
                NHTSA also considers the Ninth Circuit's decision in Center for
                Biological Diversity v. NHTSA \2511\ which remanded NHTSA's 2006 final
                rule establishing standards for MYs 2008-2011 light trucks and
                underscored that ``the overarching purpose of EPCA is energy
                conservation.''
                ---------------------------------------------------------------------------
                 \2509\ 549 U.S. 497, 531 (2007).
                 \2510\ Id. at 532.
                 \2511\ 538 F.3d 1172 (9th Cir. 2008).
                ---------------------------------------------------------------------------
                 The proposed rule presented an analysis of a wide range
                alternatives as potential revisions of the existing standards for model
                year 2021 and new standards for model years 2022-2026. These
                alternatives ranged from a zero percent increase in stringency to a
                stringency increase for passenger cars of 2 percent per year and for
                light trucks of 3 percent per year, in addition to the baseline
                alternative consisting of the augural standards.\2512\ The analysis
                supported the range of alternative standards based on factors relevant
                to NHTSA's exercise of its 49 U.S.C. 32902(f) authority, such as fuel
                saved and emissions reduced, the technologies available to meet the
                standards, the costs of compliance for automakers and their abilities
                to comply by applying technologies, the impact on consumers with
                respect to cost and vehicle choice, and effects on safety. The proposed
                rule identified the alternative composed of a zero percent increase in
                stringency as the preferred alternative.
                ---------------------------------------------------------------------------
                 \2512\ 83 FR 42990, Table I-4 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 NHTSA received numerous public comments on the range of stringency
                alternatives in the proposed rule and NHTSA's consideration of various
                factors in determining maximum feasible CAFE standards under 49 U.S.C.
                chapter 329. Below NHTSA responds to comments on these issues. NHTSA
                notes that many comments concerned the technical foundation and
                analysis upon which NHTSA was basing its regulatory decisions, such as
                the modeling of fuel economy-improving technologies and costs, the
                safety analysis, and consumer issues. Comments specific to these
                analyses are discussed elsewhere in this preamble. The section below
                addresses comments specifically addressing NHTSA's considerations in
                finalizing maximum
                [[Page 25122]]
                feasible CAFE standards under 49 U.S.C. chapter 329.
                 NHTSA's conclusion, after consideration of the factors described
                below, public comments, and other information in the administrative
                record for this action is that 1.5 percent annual increases in
                stringency from the MY 2020 standards through MY 2026 (Alternative 3 of
                this final rule analysis) \2513\ are maximum feasible. Holding CAFE
                standards for MY 2020 flat through MY 2026, as proposed, would unduly
                weigh economic practicability concerns more heavily than the need of
                the United States to conserve energy, while finalizing the MY 2021 and
                augural standards first established and set forth in 2012 would place
                undue weight on the need of the U.S. to conserve energy while being
                beyond economically practicable, as described in more detail below.
                ---------------------------------------------------------------------------
                 \2513\ The numbered Alternatives presented in the SAFE proposed
                rule (see Table I-4 at 83 FR 42990, August 24, 2018) were in some
                cases defined differently than those presented in this final rule
                (see Section V). Unless otherwise stated, the Alternatives described
                in this section refer to those presented in this final rule.
                ---------------------------------------------------------------------------
                 The following sections discuss in more detail the statutory
                requirements and considerations involved in NHTSA's determination of
                maximum feasible CAFE standards, comments received on those issues, and
                NHTSA's explanation of its balancing of factors for this final rule.
                1. EPCA, as Amended by EISA
                 EPCA, as amended by EISA, contains a number of provisions regarding
                how to set CAFE standards. DOT (by delegation, NHTSA) \2514\ must
                establish separate CAFE standards for passenger cars and light trucks
                \2515\ for each model year,\2516\ and each standard must be the maximum
                feasible that the Secretary (again, by delegation, NHTSA) believes the
                manufacturers can achieve in that model year.\2517\ In determining the
                maximum feasible level achievable by the manufacturers, EPCA requires
                that NHTSA consider four statutory factors of technological
                feasibility, economic practicability, the effect of other motor vehicle
                standards of the Government on fuel economy, and the need of the United
                States to conserve energy.\2518\ In addition, NHTSA has the authority
                to consider (and traditionally does) other relevant factors, such as
                the effect of the CAFE standards on motor vehicle safety and consumer
                preferences.\2519\ The ultimate determination of what standards can be
                considered maximum feasible involves a weighing and balancing of
                factors, and the balance may shift depending on the information before
                NHTSA about the expected circumstances in the model years covered by
                the rulemaking. The agency's decision must also support the overarching
                purpose of EPCA, energy conservation, while balancing these
                factors.\2520\
                ---------------------------------------------------------------------------
                 \2514\ EPCA and EISA direct the Secretary of Transportation to
                develop, implement, and enforce fuel economy standards (see 49
                U.S.C. 32901 et. seq.), which authority the Secretary has delegated
                to NHTSA at 49 CFR 1.95(a).
                 \2515\ 49 U.S.C. 32902(b)(1) (2007).
                 \2516\ 49 U.S.C. 32902(a) (2007).
                 \2517\ Id.
                 \2518\ 49 U.S.C. 32902(f) (2007).
                 \2519\ Both of these additional considerations also can be
                considered part of economic practicability, but NHTSA also has the
                authority to consider them independently of that statutory factor.
                 \2520\ Center for Biological Diversity v. NHTSA, 538 F. 3d 1172,
                1197 (9th Cir. 2008) (``Whatever method it uses, NHTSA cannot set
                fuel economy standards that are contrary to Congress's purpose in
                enacting the EPCA--energy conservation.'').
                ---------------------------------------------------------------------------
                 Besides the requirement that the standards be maximum feasible for
                the fleet in question and the model year in question, EPCA/EISA also
                contain several other requirements, as explained below.
                (a) Lead Time
                 EPCA requires that NHTSA prescribe new CAFE standards at least 18
                months before the beginning of each model year.\2521\ Thus, if the
                first year for which NHTSA is proposing to set new standards in this
                NPRM is MY 2022, NHTSA interprets this provision as requiring the
                agency to issue a final rule covering MY 2022 standards no later than
                April 1, 2020.
                ---------------------------------------------------------------------------
                 \2521\ 49 U.S.C. 32902(a) (2007).
                ---------------------------------------------------------------------------
                 For amendments to existing standards, EPCA requires that if the
                amendments make an average fuel economy standard more stringent, at
                least 18 months of lead time must be provided.\2522\ EPCA contains no
                lead time requirement to amend standards if the amendments make an
                average fuel economy standard less stringent. NHTSA therefore
                interprets EPCA as allowing amendments to reduce a standard's
                stringency up until the beginning of the model year in question. In the
                NPRM, NHTSA proposed to amend the standards for model year 2021. NHTSA
                explained that since the agency was proposing to reduce these
                standards, the action was not subject to a lead time requirement.
                ---------------------------------------------------------------------------
                 \2522\ 49 U.S.C. 32902(g)(2) (2007).
                ---------------------------------------------------------------------------
                 The States and Cities commenters argued that NHTSA had counted 18
                months incorrectly, and that ``18 months prior to September 1, 2021 is
                in fact March 1, 2020.'' \2523\ NHTSA agrees that 18 months prior to
                September 1 would be March 1 of the year prior; the statement in the
                NPRM that ``NHTSA has consistently interpreted the ``beginning of the
                model year'' as September 1 of the CY prior'' was a typographical
                error. As prior Federal Register notices indicate, NHTSA has in fact
                long interpreted the beginning of the model year for CAFE compliance
                purposes as October 1 of the CY prior.\2524\ Thus, counting backwards,
                18 months prior to October 1 is properly identified as April 1, meaning
                that new standards for MY 2022 must be established by April 1, 2020.
                ---------------------------------------------------------------------------
                 \2523\ States and Cities, NHTSA-2018-0067-11735, Detailed
                Comments, at 78, fn. 211.
                 \2524\ See, e.g., 75 FR 25546 (May 7, 2010).
                ---------------------------------------------------------------------------
                 With regard to the amendments to the MY 2021 standards, a coalition
                of environmental groups commented that NHTSA's legal construction of
                EPCA's lead time requirement as not applying to MY 2021 was ``not . . .
                permissible,'' arguing that section 32902(g)(1) only permits amendments
                to existing CAFE standards that ``meet[ ] the requirement of subsection
                (a) or (d) as appropriate,'' and that section 32902(a) requires fuel
                economy standards to be prescribed 18 months before the beginning of
                the model year.\2525\ The environmental group coalition therefore
                argued that the two identified provisions must be read together to
                compel all amendments to standards to be prescribed at least 18 months
                before a model year, and concluded that because it was impossible to
                finish a final rule 18 months before the start of MY 2021, that MY 2021
                standards could not be amended.\2526\ The States and Cities group
                provided similar comments, arguing that NHTSA's interpretation of
                (g)(2) rendered the reference in (g)(1) to (a) ``a nullity,'' and that
                the ``as appropriate'' language in (g)(1) referred to the determination
                of whether providing 18 months of lead time was appropriate, rather
                than to whether (a) or (d) was the relevant provision governing the
                standards in question.\2527\ NCAT commented that ``Congress in Sec.
                32902 has indicated that at least 18 months of lead time are
                appropriate when setting standards,'' and stated that ``Manufacturers'
                need for adequate lead time when designing products and developing
                compliance strategies is the same regardless of whether the agency
                [[Page 25123]]
                is making standards more stringent, less stringent, or simply changing
                the structure or compliance options provided under the standards.''
                \2528\ NADA, in contrast, argued that NHTSA does ``have the authority
                and discretion to reopen the MY 2021 standards,'' and that the
                ``mandate for at least 18 months of lead time before new standards may
                take effect does not apply to instances, such as for MY 2021, where
                standards are being relaxed.'' \2529\ CEI also agreed with NHTSA's
                interpretation of lead time set forth in the NPRM.\2530\
                ---------------------------------------------------------------------------
                 \2525\ Center for Biological Diversity, Conservation Law
                Foundation, Earthjustice, Environmental Defense Fund, Environmental
                Law and Policy Center, Natural Resources Defense Council, Public
                Citizen, Sierra Club, Union of Concerned Scientists (hereafter,
                ``environmental group coalition''), Appendix A, NHTSA-2018-0067-
                12000, at 66.
                 \2526\ Id.
                 \2527\ States and Cities, NHTSA-2018-0067-11735, Detailed
                Comments, at 78-79.
                 \2528\ NCAT, NHTSA-2018-0067-11969, at 46.
                 \2529\ NADA, NHTSA-2018-0067-12064, at 9.
                 \2530\ CEI, NHTSA-2018-0067-12015, at 3-4.
                ---------------------------------------------------------------------------
                 NHTSA agrees that section 32902(g)(1) states that amendments must
                meet the requirements of subsection (a) or (d) as appropriate, and that
                32902(a) states that standards must be prescribed 18 months in advance
                of the model year. However, NHTSA cannot agree that the 18-month lead
                time requirement applies to amendments to existing standards that
                reduce stringency. Section 32902(g)(2) clearly states that ``[w]hen the
                Secretary of Transportation prescribes an amendment under this section
                that makes an average fuel economy standard more stringent (emphasis
                added), the Secretary shall prescribe the amendment . . . at least 18
                months before the beginning of the model year to which the amendment
                applies.'' Commenters' construction of the statute would render
                superfluous the words ``more stringent'' in 32902(g)(2), and there is a
                presumption against superfluity.\2531\ Congress purposely included the
                words ``more stringent'' in order to exclude the contrary situation--
                ``less stringent''--from the 18-month lead time requirement. A plain
                reading of (g)(1) simply provides that the Secretary (by delegation,
                NHTSA) should refer to the correct provision depending on whether the
                standard being amended is generally applicable (pointing to section
                (a)) or a standard applicable to low-volume manufacturer pursuant to an
                exemption (pointing to section (d)). Reading (g)(1) and (g)(2) together
                is the appropriate way to give effect to both provisions. This reading
                provides that NHTSA may amend the MY 2021 standard by following the
                requirements for generally-applicable standards; this reading also
                provides that 18 months' lead time is only required for amendments that
                increase stringency. NHTSA also does not agree that (g)(1) can be read
                to imply that the agency must provide 18 months of lead time ``if
                appropriate,'' as the States and Cities suggest, nor that there is any
                statutory basis to extend the lead time requirement to changes to the
                ``structure or compliance options provided under the standards'' as
                NCAT suggests. If new off-cycle technologies could not be recognized
                toward compliance without providing 18 months' lead time, manufacturer
                efforts to rely on that compliance flexibility to redress past
                shortfalls would be frustrated.
                ---------------------------------------------------------------------------
                 \2531\ See, e.g., Duncan v. Walker, 533 U.S. 167 (2001) (citing
                U.S. v. Menasche, 348 U.S. 528, 538-539 (1955)).
                ---------------------------------------------------------------------------
                 Moreover, automakers need more time to respond when NHTSA amends
                standards to be more stringent--doing so would likely require
                automakers to change their product and/or sales plans to ensure that
                they will meet more-stringent standards than those standards for which
                they may have already prepared. But such product or sales plans would
                not necessarily need to be changed if standards were amended to be less
                stringent--in fact an automaker would be rewarded by keeping existing
                plans to comply in place with additional bankable and tradable
                overcompliance credits. However, the environmental group coalition
                argued that ``[c]hanging the MY 2021 standard at this late date would
                penalize technologically advanced automakers and parts suppliers, who
                have already made significant investments in updating their
                technology.'' \2532\ The States and Cities group made similar
                comments,\2533\ as did NCAT.\2534\ The environmental group coalition
                further suggested that amending the MY 2021 standard would reduce the
                need for (and thus the value) of overcompliance credits, ``which would
                be disruptive to the manufacturers that have done the most to further
                EPCA's conservation goals.'' \2535\ NCAT made similar comments, arguing
                that ``The practical and financial impact of the change accordingly is
                not materially different from increasing the stringency of a standard
                this late in the product cycle.'' \2536\
                ---------------------------------------------------------------------------
                 \2532\ Environmental group coalition, NHTSA-2018-0067-12000,
                Appendix A, at 66.
                 \2533\ States and Cities, NHTSA-2018-0067-11735, Detailed
                Comments, at 78, fn. 213.
                 \2534\ NCAT, NHTSA-2018-0067-11969, at 46-47.
                 \2535\ Environmental group coalition, NHTSA-2018-0067-12000,
                Appendix A. at 66-67.
                 \2536\ NCAT, NHTSA-2018-0067-11969, at 47.
                ---------------------------------------------------------------------------
                 NHTSA believes that to the extent that some manufacturers have
                already invested in future fuel economy improvements, those
                manufacturers will continue to be well-positioned both to respond to
                increasing standards in the future, and to take advantage of any market
                demand for higher fuel economy/reduced tailpipe CO2
                emissions from consumers who put a premium on those aspects. NHTSA is
                also aware that several companies have self-imposed emissions-reduction
                goals which may drive their decisions on technology application
                regardless of regulatory obligations. NHTSA does not believe that
                companies which have already invested in higher levels of technology
                consider those investments to be bad ones. The agencies note that
                manufacturer commenters, despite the concerns expressed by others, did
                not comment about a lack of lead time associated with changing the MY
                2021 standards; rather, many manufacturer commenters expressly cited
                the need to revise MY 2021 standards, arguing that the previously-
                established values are beyond maximum feasible. Regarding the value of
                overcompliance credits under more or less stringent standards, NHTSA
                agrees that the need for credits may be less under less stringent
                standards, but this is true regardless of the lead time question.
                Further, NHTSA does not believe that this suggests only standards that
                compel reliance on overcompliance credits (especially those earned by
                competitors) can be maximum feasible; this topic will be addressed in
                further detail below, and regardless, NHTSA is prohibited from
                considering credit availability in determining maximum feasible CAFE
                standards.
                (b) Separate Standards for Cars and Trucks, and Minimum Standards for
                Domestic Passenger Cars
                 As discussed above, EPCA requires NHTSA to set separate CAFE
                standards for passenger cars and light trucks for each model
                year.\2537\ NHTSA interprets this requirement as preventing the agency
                from setting a single combined CAFE standard for cars and trucks
                together, based on the plain language of the statute. Congress
                originally required separate CAFE standards for cars and trucks to
                reflect the different fuel economy capabilities of those different
                types of vehicles,\2538\ and over the history of the CAFE program, has
                never revised this requirement. Even as many cars and trucks have come
                to resemble each other more closely over time--many crossover and
                sport-utility models, for example, come in versions today that may be
                subject to either the car standards or the truck standards depending on
                their characteristics--it is still accurate to say that vehicles with
                [[Page 25124]]
                truck-like characteristics such as 4 wheel drive, cargo-carrying
                capability, etc., consume more fuel per mile than vehicles without
                these characteristics. Thus, NHTSA believes that the different fuel
                economy capabilities of cars and trucks would generally make separate
                standards appropriate for these different types of vehicles, regardless
                of the plain language of the statute which requires such treatment.
                ---------------------------------------------------------------------------
                 \2537\ 49 U.S.C. 32902(b)(1) (2007).
                 \2538\ Indeed, EPCA initially only required NHTSA to establish
                CAFE standards for passenger cars; establishment of light truck
                standards was permissible.
                ---------------------------------------------------------------------------
                 EPCA, as amended by EISA, also requires another separate standard
                to be set for domestically-manufactured \2539\ passenger cars. Unlike
                standards for passenger cars and light trucks described above, the
                compliance burden of the minimum domestic passenger car standard is the
                same for all manufacturers: The statute clearly states that any
                manufacturer's domestically-manufactured passenger car fleet must meet
                the greater of either 27.5 mpg on average, or
                ---------------------------------------------------------------------------
                 \2539\ In the CAFE program, ``domestically-manufactured'' is
                defined by Congress in 49 U.S.C. 32904(b). The definition roughly
                provides that a passenger car is ``domestically manufactured'' as
                long as at least 75% of the cost to the manufacturer is attributable
                to value added in the United States, Canada, or Mexico, unless the
                assembly of the vehicle is completed in Canada or Mexico and the
                vehicle is imported into the United States more than 30 days after
                the end of the model year.
                92 percent of the average fuel economy projected by the Secretary
                for the combined domestic and non-domestic passenger automobile
                fleets manufactured for sale in the United States by all
                manufacturers in the model year, which projection shall be published
                in the Federal Register when the standard for that model year is
                promulgated in accordance with [49 U.S.C. 32902(b)].\2540\
                ---------------------------------------------------------------------------
                 \2540\ 49 U.S.C. 32902(b)(4) (2007).
                 Since that requirement was promulgated, the ``92 percent'' has
                always been greater than 27.5 mpg. NHTSA published the 92-percent
                minimum domestic passenger car standards for model years 2017-2025 at
                49 CFR 531.5(d) as part of the 2012 final rule. For MYs 2022-2025,
                531.5(e) states that these were to be applied if, when actually
                proposing MY 2022 and subsequent standards, the previously identified
                standards for those years are deemed maximum feasible, but if NHTSA
                determines that the previously identified standards are not maximum
                feasible, the 92-percent minimum domestic passenger car standards would
                also change. This is consistent with the statutory language that the
                92-percent standards must be determined at the time an overall
                passenger car standard is promulgated and published in the Federal
                Register. Thus, any time NHTSA establishes or changes a passenger car
                standard for a model year, the minimum domestic passenger car standard
                for that model year will also be evaluated or reevaluated and
                established accordingly. NHTSA explained this in the rulemaking to
                establish standards for MYs 2017 and beyond and received no
                comments.\2541\
                ---------------------------------------------------------------------------
                 \2541\ 77 FR 62624, 63028 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 The 2016 Alliance/Global petition for rulemaking asked NHTSA to
                revise the 92-percent minimum domestic passenger car standards
                retroactively for MYs 2012-2016 ``to reflect 92 percent of the required
                average passenger car standard taking into account the fleet mix as it
                actually occurred, rather than what was forecast.'' The petitioners
                stated that doing so would be ``fully consistent with the statute.''
                \2542\
                ---------------------------------------------------------------------------
                 \2542\ Automobile Alliance and Global Automakers Petition for
                Direct Final Rule with Regard to Various Aspects of the Corporate
                Average Fuel Economy Program and the Greenhouse Gas Program (June
                20, 2016) at 5, 17-18, available at https://www.epa.gov/sites/production/files/201609/documents/petition_to_epa_from_auto_alliance_and_global_automakers.pdf
                (hereinafter Alliance/Global Petition).
                ---------------------------------------------------------------------------
                 NHTSA explained in the NPRM that NHTSA understood that determining
                the 92 percent value ahead of the model year to which it applies, based
                on the information then available to the agency, would result in a
                different mpg number than if NHTSA determined the 92 percent value
                based on the information available at the end of the model year in
                question. NHTSA further explained that it understood that determining
                the 92 percent value ahead of time could make the minimum domestic
                passenger car standard more stringent than it could be if it were
                determined at the end of the model year, if manufacturers end up
                producing more larger-footprint passenger cars than what NHTSA had
                originally anticipated.
                 Accordingly, NHTSA sought comment on the request by Alliance/
                Global. Additionally, recognizing the uncertainty inherent in
                projecting specific values far into the future, NHTSA also sought
                comment on whether it is possible to define the 92 percent valueas a
                range, if NHTSA defined the values associated with a CAFE standard
                (i.e., the footprint curve) as a range rather than as a single number.
                NHTSA referred to the sensitivity analysis included in the proposal and
                in the accompanying PRIA as a basis for such an mpg range ``defining''
                the passenger car standard in any given model year. If NHTSA took that
                approach, 92 percent of that ``standard'' would also, necessarily, be a
                range. NHTSA broadly sought comment on that approach or other similar
                approaches.
                 The Alliance and FCA commented that they ``supported the NHTSA
                proposal'' to calculate 92 percent as a range rather than as a single
                value, with the ultimate minimum domestic passenger car standard to be
                determined at the end of the MY to which it applies.\2543\ Both
                organizations cited compliance difficulties when the 92 percent
                calculated at the time of the rulemaking turns out to be more stringent
                than 92 percent of the final MY compliance obligations for passenger
                cars, and argued that minimum domestic passenger car standards should
                be recalculated as part of this rulemaking for all model years, rather
                than only MYs 2021-2026, in order to ameliorate that compliance
                difficulty retroactively. The Alliance argued that the 18 month lead
                time requirement should not be interpreted to apply to the minimum
                domestic passenger car standards, because if the 92 percent value is a
                range like the overall passenger car curve, then that value cannot be
                determined until after the model year is completed.\2544\ Because
                manufacturers' individual compliance obligations are not subject to the
                18 month lead time requirement, the Alliance requested that the 92
                percent should similarly not be.\2545\ Separately, Kreucher commented
                that NHTSA should expand the credit transfer provision to allow
                transferred credits to be used to meet the minimum domestic passenger
                car standard.\2546\
                ---------------------------------------------------------------------------
                 \2543\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 41;
                FCA, NHTSA-2018-0067-11943, at 64.
                 \2544\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 42-
                43.
                 \2545\ Id.
                 \2546\ Kreucher, NHTSA-2018-0067-0444, at 11.
                ---------------------------------------------------------------------------
                 In contrast, the States and Cities and ACEEE opposed changes to the
                minimum domestic passenger car standard, with the States and Cities
                commenting that NHTSA ``is proposing to retroactively revise the 92
                percent based on actual fleet mix'' \2547\ and ACEEE simply noting that
                the Alliance/Global had requested that NHTSA do this.\2548\ ACEEE
                stated that NHTSA did not have discretion to alter the statutory
                requirement, and argued that calculating 92 percent at the end of the
                model year was ``entirely counter to the intent of the law--the so-
                called backstop is designed explicitly to protect against the market
                shifts for which the [industry is] asking the standard to be
                adjusted.'' \2549\
                [[Page 25125]]
                The States and Cities similarly argued that ``the 92 percent
                requirement is expressly intended to be a projection, not a
                retrospective recalculation,'' and ``the statute does not contemplate a
                `range,' but rather a `minimum' with a set value--92 percent. If
                Congress had intended the value to be a range, it would have included
                that language in the statute, and would not have determined the value
                with such specificity.'' \2550\
                ---------------------------------------------------------------------------
                 \2547\ States and Cities, NHTSA-2018-0067-11735, at 79.
                 \2548\ ACEEE, NHTSA-2018-0067-12122, Attachment (joint NGO
                comment to manufacturer petition for flexibilities), at 15.
                 \2549\ Id. ACEEE cited a NHTSA statement in the 2010 final rule
                establishing standards for MYs 2012-2016 in support of this
                argument, noting that NHTSA had said ``this minimum standard was
                intended to act as a `backstop,' ensurng that domestically-
                manufactured passenger cars reached a given mpg level even if the
                market shifted in ways likely to reduce overall fleet mpg.'' Id.
                (emphasis added).
                 \2550\ States and Cities, NHTSA-2018-0067-11735, at 79.
                ---------------------------------------------------------------------------
                 NHTSA considered comments about setting the MDPCS as a range. NHTSA
                recognizes that the approach discussed in the NPRM may not be within
                our statutory authority and therefore is setting the standards as
                specific values.
                 NHTSA agrees that setting the MDPCS after the model year is
                completed and the total passenger car fleet standard is known would
                provide standards that adapt with changes in consumer demand. However,
                such an approach would not establish the final numerical value until
                significantly after the model year completed, only after final
                compliance data has been submitted by all manufacturers and EPA and
                NHTSA have completed compliance work for the total passenger car fleet.
                In addition, the standard would be based on the production of all
                manufacturers of passenger cars, providing no means for an individual
                manufacturer to have certainty over its final standard. Individual
                manufacturers likewise would have no control over the value by
                controlling their production mix. For these reasons, NHTSA is denying
                the Alliance/Global petition that the 92 percent value for the MDPCS be
                determined based on the information available at the end of the model
                year in question.
                 That said, NHTSA agrees that the actual total passenger car fleet
                standards have differed significantly the 2012 projection, and examined
                the projections from past rulemakings in greater detail. NHTSA reviewed
                the total passenger car fleet (all domestic and import passenger cars)
                standard that was projected at the time of rulemakings for MYs 2011 to
                2018 and compared those projections to the actual total fleet passenger
                car standard for each of those model years from compliance data, based
                on the actual footprints and production volume of the models produced
                in those model years. Table VIII-1 shows the projected standards and
                the actual standards on a fuel economy basis, and Table VIII-2 shows
                the fuel economy values converted to fuel consumption values which was
                used as the basis for and analyzing the differences between the
                projected standards and actual standards.\2551\ Table VIII-2 also shows
                the percentage difference between the total passenger car fleet
                standard at the time of the rulemaking and the actual fleet standard
                based on compliance data.
                ---------------------------------------------------------------------------
                 \2551\ Consistent with EPCA/EISA and corresponding regulations,
                CAFE compliance calculations have been conducted on a mile per
                gallon basis. However, engineering computations have almost
                exclusively been conducted on a fuel consumption basis (i.e., in
                gallons per mile), because the underlying engineering relationships
                are more meaningfully defined on a fuel consumption basis.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 25126]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.729
                [GRAPHIC] [TIFF OMITTED] TR30AP20.730
                [[Page 25127]]
                 The data show that the standards projected in 2012 were
                consistently more stringent than the actual standards, by an average of
                1.9 percent. This difference indicates that in rulemakings conducted in
                2009 through 2012, the agencies' projections of passenger car vehicle
                footprints and production volumes consistently underestimated the
                consumer demand for larger passenger cars over the MYs 2011 to 2018
                period.
                BILLING CODE 4910-59-C
                 To establish minimum standards for domestic passenger cars in these
                past rulemakings, NHTSA computed the average of manufacturers'
                requirements given the attribute-based standards being issued, and
                given the projected distribution of passenger car footprints as
                indicated in the analysis fleet (aka market forecast) used to analyze
                impacts of the standards. The joint NHTSA-EPA rulemaking establishing
                standards for MYs 2012-2016 presented analysis that, in turn, used a
                ``2008-based'' market forecast that combined detailed information
                regarding the MY 2008 fleet with a commercial market forecast (by brand
                and segment) and a range of agency assumptions. Importantly, the
                commercial market forecast showed Chrysler's production falling
                dramatically, and never recovering; as well as Chrysler passenger cars
                being distributed more than most OEMs (other than Jaguar and Mercedes)
                toward larger footprints, and this forecast impacted the NHTSA's
                projection of overall average requirements for passenger cars under the
                footprint-based standards. For example, the 2008-based forecast showed
                production of Chrysler brands (Chrysler, Dodge, Jeep, and Ram) for the
                U.S. market totaling 0.8 million units by MY 2017, and today's analysis
                fleet uses a MY 2017 fleet showing 1.9 million Chrysler-branded units.
                Also, among the agencies' assumptions, was that some manufacturers
                (Chrysler, Ford, Subaru, Mazda, and Mitsubishi) would rapidly increase
                production of small footprint vehicles not observed in the MY 2008
                fleet.
                 The joint rulemaking establishing standards for MYs 2017-2025 also
                used this 2008-based fleet for the NPRM, showing more than 1.3 million
                units smaller than 41 square feet in MY 2017, far more than the 0.3m
                units shown in the model inputs for today's analysis. For the 2012
                final rule, the agencies conducted side-by-side analysis, one using the
                2008-based fleet, and one using a 2010-based fleet. The 2010-based
                fleet used a newer commercial forecast that was considerably more
                sanguine regarding, for example, FCA's prospects. Minimum standards for
                domestic passenger cars were based on an average of results for the
                2008-based and 2010-based total passenger car fleets.
                 The analysis fleet underlying today's reference case analysis is
                discussed above in Section VI.A.2 and available in full detail with the
                model inputs and outputs accompanying today's notice.\2552\ For the
                current rulemaking, NHTSA also considered that, unlike the passenger
                car standards and light truck standards which are vehicle attribute-
                based and automatically adjust with changes in consumer demand, that
                MDPCSs are not attribute-based, and therefore do not adjust with
                changes in consumer demand. They are fixed standards that are
                established at the time of the rulemaking. The MYs 2011-2018 MDPCS were
                more stringent and placed more burden on manufacturers of domestic
                passenger cars than was projected and expected at the time of the
                rulemakings. NHTSA agrees with the Alliance's concerns over the impact
                of changes in consumer demand on manufacturers' ability to comply with
                the MDPCS and in particular, manufacturers that produce larger
                passenger cars domestically.
                ---------------------------------------------------------------------------
                 \2552\ https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
                ---------------------------------------------------------------------------
                 Additionally, as discussed in more detail in Section VIII.B.4
                below, consumer demand may shift even more in the direction of larger
                passenger cars if fuel prices continue to remain low. The fuel prices
                used in the analysis for this final rule rely on EIA's future forecasts
                of fuel prices, which were made prior to the recent collapse of oil
                prices. If the former OPEC+ members continue to pursue market share,
                fuel prices will likely continue to drop. If, instead of pursuing
                market share, they try to control prices restricting supply, U.S. shale
                production could begin to ramp back up and exert downward pressure on
                price. If fuel prices end up even lower than our analysis assumes,
                benefits from saving additional fuel will be worth even less to
                consumers. Our analysis captures none of these effects. Sustained low
                oil prices can be expected to have real effects on consumer demand for
                additional fuel economy, and consumers may foreseeably be even less
                interested in smaller passenger cars than they are at present.
                 To help avoid similar outcomes in the rulemaking timeframe to what
                has happened with the MDPCS over the last several model years, NHTSA
                determined it is reasonable and appropriate to consider the recent
                projection errors as part of estimating the projected total passenger
                car fleet fuel economy for MYs 2021-2026. As stated above the average
                difference over MYs 2011-2018 was 1.9 percent. As explained above,
                those differences are largely attributable to aspects of the forecasts
                that turned out to be far different from reality. NHTSA is projecting
                the total passenger car fleet fuel economy using the central analysis
                value in each model year and applying an offset based on the historical
                1.9 percent difference identified for MYs 2011-2018. Table VIII-3 hows
                the calculation values used to determine the total passenger car fleet
                fuel economy value for each model year.
                 NHTSA will continue its practice of determining the MDPCS as
                specific values at the same time that it sets passenger car standards,
                at 92 percent of the projected passenger cars standard in each model
                year. Table VIII-3 also shows the computations for the MDPCS for each
                model year. The new MDPCS are prescribed in the regulatory text below.
                [[Page 25128]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.731
                 Table VIII-4 lists the minimum domestic passenger car standards
                reflecting the updated analysis discussed above, and comparing these to
                standards that would correspond to each of the other regulatory
                alternatives considered. NHTSA has updated these to reflect its overall
                analysis and resultant projection for the CAFE standards finalized
                today, highlighted below as ``Preferred (Alternative 3),'' and has
                calculated what those standards would be under the no action
                alternative (as issued in 2012, as updated for the NPRM, and as further
                updated by today's analysis) and under the other alternatives described
                and discussed further in Section V, above. As explained in a separate
                memorandum to the document, while the CAFE Model analysis underlying
                the FEIS, FRIA, and final rule does not reflect this change, separate
                analysis that does reflect the change demonstrates that doing so does
                not change estimated impacts of any of the regulatory alternatives
                under consideration.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.732
                [[Page 25129]]
                Attribute-Based and Defined by Mathematical Function
                 EISA requires NHTSA to set CAFE standards that are ``based on 1 or
                more attributes related to fuel economy and express[ed] . . . in the
                form of a mathematical function.'' \2553\ Historically, NHTSA has based
                standards on vehicle footprint and proposes to continue to do so for
                all the reasons described in previous rulemakings. As in previous
                rulemakings, NHTSA proposed to define the standards in the form of a
                constrained linear function that generally sets higher (more stringent)
                targets for smaller-footprint vehicles and lower (less stringent)
                targets for larger-footprint vehicles. These footprint curves are
                discussed in much greater detail in Section V above. NHTSA sought
                comment both on the choice of footprint as the relevant attribute and
                on the rationale for the constrained linear functions chosen to
                represent the standards; those comments and NHTSA's responses are
                discussed above in Section V.
                ---------------------------------------------------------------------------
                 \2553\ 49 U.S.C. 32902(b)(3)(A).
                ---------------------------------------------------------------------------
                d) Number of Model Years for Which Standards May Be Set at a Time
                 EISA also states that NHTSA shall ``issue regulations under this
                title prescribing average fuel economy standards for at least 1, but
                not more than 5, model years.'' \2554\ In the 2012 final rule, NHTSA
                interpreted this provision as preventing the agency from setting final
                standards for all of MYs 2017-2025 in a single rulemaking action, so
                the MYs 2022-2025 standards were termed ``augural,'' meaning ``that
                they represent[ed] the agency's current judgment, based on the
                information available to the agency [then], of what levels of
                stringency would be maximum feasible in those model years.'' \2555\
                That said, NHTSA also repeatedly clarified that the augural standards
                were in no way final standards and that a future de novo rulemaking
                would be necessary in order both to propose and to promulgate final
                standards for MYs 2022-2025.
                ---------------------------------------------------------------------------
                 \2554\ 49 U.S.C. 32902(b)(3)(B).
                 \2555\ 77 FR 62623, 62630 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 In the NPRM, NHTSA proposed to establish new standards for MYs
                2022-2026 and to revise the previously-established final standards for
                MY 2021. NHTSA explained that legislative history suggests that
                Congress included the five year maximum limitation so NHTSA would issue
                standards for a period of time where it would have reasonably realistic
                estimates of market conditions, technologies, and economic
                practicability (i.e., not set standards too far into the future).\2556\
                However, NHTSA suggested that the concerns Congress sought to address
                by imposing those limitations are not present for nearer model years
                where NHTSA already has existing standards, and noted that revisiting
                existing standards is contemplated by both 49 U.S.C. 32902(c) and
                32902(g). NHTSA stated that the agency therefore believed that it is
                reasonable to interpret section 32902(b)(3)(B) as applying only to the
                establishment of new standards rather than to the combined action of
                establishing new standards and amending existing standards.
                ---------------------------------------------------------------------------
                 \2556\ See 153 Cong. Rec. 2665 (Dec. 28, 2007).
                ---------------------------------------------------------------------------
                 Moreover, NHTSA argued, it would be an absurd result if the five
                year maximum limitation were interpreted to prevent NHTSA from revising
                a previously-established standard that the agency had determined to be
                beyond maximum feasible, while concurrently setting five years of
                standards not so distant from today. The concerns Congress sought to
                address are much starker when NHTSA is trying to determine what
                standards would be maximum feasible 10 years from now as compared to
                three years from now.
                 NADA commented that NHTSA has discretion and authority to set
                standards for MY 2026 and that the ``statutory five-year rule is not a
                barrier to doing so,'' \2557\ while the environmental group coalition
                argued that NHTSA ``is limited to prescribing fuel economy standards
                for only five model years at a time,'' but ``[h]ere, NHTSA is setting
                standards for six model years, 2021 through 2026. This exceeds NHTSA's
                statutory authority.'' \2558\ Consumers Union argued that ``[i]f
                Congress had intended the statute to only apply to the establishment of
                new standards, as the agencies contend, it certainly could have stated
                as such. But Congress did not include any language even hinting at this
                interpretation.'' \2559\
                ---------------------------------------------------------------------------
                 \2557\ NADA, NHTSA-2018-0067-12064, at 9.
                 \2558\ Environmental group coalition, NHTSA-2018-0067-12000, at
                66.
                 \2559\ Consumers Union, NHTSA-2018-0067-12068, Attachment A, at
                24.
                ---------------------------------------------------------------------------
                 NHTSA continues to believe, consistent with the legislative
                history, that the five year limitation was intended to prevent NHTSA
                from setting standards too far into the future, recognizing that
                predicting the future is difficult. Consumers Union is correct that
                nothing in the statute compels the interpretation that the five year
                limitation applies only to the setting of new standards rather than to
                the combined action of establishing new standards and amending existing
                standards, but NHTSA does not believe that the statute precludes this
                interpretation, either. The statute allows NHTSA to revisit existing
                standards; the statute separately allows NHTSA to prescribe new
                standards for at least 1, but not more than 5, model years when it
                ``issues regulations.'' It is not clear whether the statute precludes
                multiple concurrent or quickly-sequential rulemakings ``issuing
                regulations'' for different periods of time. If this approach were
                used, for example, to try to set ten years' worth of CAFE standards
                essentially at once, this would appear directly contrary to the
                statute. If this approach were used to revisit an existing standard and
                then (in a separate rulemaking) set five years' worth of standards for
                the immediately ensuing model years, this would seem consistent with
                Congressional intent, but an unnecessary use of tax dollars that could
                be saved by consolidating agency (and commenter) work into a single
                rulemaking action. NHTSA does not believe that Congress intended to
                force the agency to waste resources, and continues to believe that the
                current interpretation is reasonable and appropriate.
                (e) Maximum Feasible Standards
                 As discussed above, EPCA requires NHTSA to consider four factors in
                determining what levels of CAFE standards would be maximum feasible,
                and NHTSA presents in the sections below its understanding of the
                meaning of those four factors. All factors should be considered, in the
                manner appropriate, and then the maximum feasible standards should be
                determined.
                (1) Technological Feasibility
                 ``Technological feasibility'' refers to whether a particular method
                of improving fuel economy is available for deployment in commercial
                application in the model year for which a standard is being
                established. Thus, NHTSA is not limited in determining the level of new
                standards to technology that is already being commercially applied at
                the time of the rulemaking. For the proposal, NHTSA explained that it
                had considered a wide range of technologies that improve fuel economy,
                subject to the constraints of EPCA regarding how to treat alternative
                fueled vehicles, such as battery-electric vehicles, in determining
                maximum feasible standards, and considering the need to account for
                which technologies have already been applied to which vehicle model/
                configuration, and the need to realistically estimate the cost and fuel
                economy impacts of each technology.
                [[Page 25130]]
                NHTSA explained that it had not attempted to account for every
                technology that might conceivably be applied to improve fuel economy
                and considered it unnecessary to do so given that many technologies
                address fuel economy in similar ways.\2560\ NHTSA noted that
                technological feasibility and economic practicability are often
                conflated, trying to explain that the question of whether a fuel-
                economy-improving technology does or will exist (technological
                feasibility) is a different question from what economic consequences
                could ensue if NHTSA effectively requires that technology to become
                widespread in the fleet and the economic consequences of the absence of
                consumer demand for technology that are projected to be required
                (economic practicability). NHTSA explained that it is therefore
                possible for standards to be technologically feasible but still beyond
                the level that NHTSA determines to be maximum feasible due to
                consideration of the other relevant factors.
                ---------------------------------------------------------------------------
                 \2560\ For example, NHTSA has not considered high-speed
                flywheels as potential energy storage devices for hybrid vehicles;
                while such flywheels have been demonstrated in the laboratory and
                even tested in concept vehicles, commercially available hybrid
                vehicles currently known to NHTSA use chemical batteries as energy
                storage devices, and the agency has considered a range of hybrid
                vehicle technologies that do so.
                ---------------------------------------------------------------------------
                 The States and Cities commenters argued that NHTSA's interpretation
                of the technological feasibility factor was unreasonable, stating that
                ``. . . fuel economy standards under EPCA are 'intended to be
                technology forcing' because Congress recognized 'that 'market forces .
                . . may not be strong enough to bring about the necessary fuel
                conservation which a national energy policy demands.' '' \2561\ The
                States and Cities commenters thus argued that all alternatives less
                stringent than the baseline/augural standards alternative were
                unacceptable because they would not force technologies to be developed
                and applied, and NHTSA had ``conce[ded] that the technology already
                exists that could meet the more stringent augural standards.'' \2562\
                These commenters stated that ``NHTSA is therefore impermissibly and
                unreasonably (and even implicitly) re-interpreting this factor in a
                manner contrary to the plain meaning of 'feasibility' and ignoring
                EPCA's technology-forcing purpose. See Chevron, 467 U.S. at 843; Fox
                Television, 556 U.S. at 515 (`An agency may not . . . depart from a
                prior policy sub silentio.').'' CARB \2563\ and CBD et al.\2564\ also
                argued that EPCA was intended to be technology forcing.
                ---------------------------------------------------------------------------
                 \2561\ States and Cities, NHTSA-2018-0067-11735, Detailed
                Comments, at 66, citing CAS, 793 F.2d at 1339 (citing S. Rep. No.
                179, 94th Cong., 1st Sess. 2 (1975) at 9).
                 \2562\ Id. at 66.
                 \2563\ CARB, NHTSA-2018-0067-11873, Detailed Comments, at 84
                (``Since market inefficiencies may preclude sufficient improvement
                without regulatory incentives, EPCA requires standards that advance
                technology. (Citing CAS v. NHTSA, 793 F.2d 1322, 1339, citing S.
                Rep. No. 179, 94th Cong., 1st Sess. 2 (1975), U.S.C.C.A.N. 1975 at
                9)'').
                 \2564\ CBD et al., NHTSA-2018-0067-12057, at 2.
                ---------------------------------------------------------------------------
                 The States and Cities commenters also argued that NHTSA had
                previously stated in rulemakings that it considered ``all types of
                technologies that improve real-world fuel economy,'' but in the NPRM
                NHTSA stated instead that it had ``not attempted to account for every
                technology that might conceivably be applied to improve fuel economy
                and consider[ed] it unnecessary to do so given that many technologies
                address fuel economy in similar ways.'' \2565\ The States and Cities
                commenters stated that ``[t]his is an unexplained departure from the
                agency's past practice and prior interpretation of `technological
                feasibility,A' citing Fox Television, and argued that NHTSA had not
                explained ``1) what `similar ways' means, or 2) why the fact that a
                technology that might improve fuel economy `in similar ways' to another
                technology obviates NHTSA's obligation to consider its availability,
                particularly given the differences in costs between different
                technologies.'' \2566\ The States and Cities commenters pointed to the
                examples of HCR1 and HCR2 as technologies ``already widely available in
                the market'' that should have been considered, and claimed that NHTSA
                had ``failed to even consult with EPA regarding which technologies the
                agency considered,'' ``result[ing] in fundamentally flawed predictions
                of what technology can be applied in model years 2021-2026.'' \2567\
                ---------------------------------------------------------------------------
                 \2565\ Id. at 67, referring to 83 FR at 43208.
                 \2566\ Id.
                 \2567\ Id.
                ---------------------------------------------------------------------------
                 Mazda, in contrast, stated that it agreed that ``mere development
                and introduction of advanced fuel efficient technologies is not
                sufficient for manufacturers to comply with established GHG and fuel
                efficiency standards. The technologies must be widely adopted by
                consumers for them to provide the expected environmental benefit.''
                \2568\ Mr. Kreucher stated that manufacturers have been applying
                ``unprecedented levels of technology'' but are still falling short of
                their compliance obligations, pointing in particular to light truck
                compliance in MY 2016. Kreucher argued that ``[t]his indicates a
                serious overestimation of technological feasibility in the prior [2012]
                analysis that must be corrected.'' \2569\
                ---------------------------------------------------------------------------
                 \2568\ Mazda, NHTSA-2018-0067-11727, at 2.
                 \2569\ Kreucher, NHTSA-2018-0067-0444, at 7.
                ---------------------------------------------------------------------------
                 UCS stated that the NPRM analysis ``undermined'' an assessment of
                ``technical feasibility,'' by ``paint[ing] fuel-saving technologies as
                less effective and more costly than real-world data indicate,'' through
                several mechanisms.\2570\ First, UCS argued that the analysis had
                underestimated ICE efficiency possibilities, ``frequently ignoring
                technology that is already commercialized or is widely anticipated to
                be readily available within the timeframe of the standards.'' \2571\
                Second, UCS suggested that the NPRM analysis had ``overstate[d] the
                degree to which manufacturers have deployed some of the most cost-
                effective technologies, while errors in full vehicle simulation and
                rampant disregard for the current state of technology underestimates
                the potential for future improvement.'' \2572\ UCS claimed that
                ``[f]requently the agencies have departed from past precedence in
                specific ways in order to increase technology costs associated with
                technology deployment, sometimes failing to provide even a glimmer of
                reasonable justification for such decisions.'' \2573\ (emphasis added)
                Third, UCS argued that the model had been deliberately constructed to
                avoid choosing the most cost-effective technology pathways, showing
                higher costs and more future overcompliance than UCS analysis
                showed.\2574\ Finally, UCS argued that better modeling of credit
                trading and use would further reduce technology costs. UCS concluded
                that ``The mischaracterization of technology and unrealistic model
                construction lead to an inaccurate assessment of technological
                feasibility, effectively undermining this factor's weight in
                considering maximum feasible standards.'' \2575\
                ---------------------------------------------------------------------------
                 \2570\ UCS, NHTSA-2018-0067-12039, at 4.
                 \2571\ Id.
                 \2572\ Id.
                 \2573\ Id.
                 \2574\ Id.
                 \2575\ Id.
                ---------------------------------------------------------------------------
                 Contrary to the assertion by several commenters that NHTSA has
                historically claimed that it must set technology-forcing standards,
                NHTSA has previously described the technological feasibility factor as
                allowing the agency to set standards that force the development and
                application of new fuel-efficient technologies.\2576\ In the same
                preamble section in which that description was set forth, NHTSA stated
                [[Page 25131]]
                that ``[i]t is important to remember that technological feasibility
                must also be balanced with the other of the four statutory factors.
                Thus, while 'technological feasibility' can drive standards higher by
                assuming the use of technologies that are not yet commercial, 'maximum
                feasible' is also defined in terms of economic practicability, for
                example, which might caution the agency against basing standards (even
                fairly distant standards) entirely on such technologies.'' \2577\ NHTSA
                further stated that ``. . . as the `maximum feasible' balancing may
                vary depending on the circumstances at hand for the model year in which
                the standards are set, the extent to which technological feasibility is
                simply met or plays a more dynamic role may also shift.'' \2578\
                ---------------------------------------------------------------------------
                 \2576\ See, e.g., 77 FR at 63015 (Oct. 15, 2012).
                 \2577\ Id.
                 \2578\ Id.
                ---------------------------------------------------------------------------
                 NHTSA continues to believe that, for purposes of this rulemaking
                covering standards for MYs 2021-2026, the crucial question is not
                whether technologies exist to meet the standards--they do. The question
                is rather, given that the technology exists, how much of it should be
                required to be added to new cars and trucks in order to conserve more
                energy, and how to appropriately balance additional energy conserved
                and additional cost for new vehicles. Regardless of whether
                technological feasibility allows the agency to set technology-forcing
                standards, technological feasibility does not require, by itself, NHTSA
                to set technology-forcing standards if other statutory factors would
                point the agency in a different direction. NHTSA has expressed this
                interpretation of technological feasibility over the course of multiple
                rulemakings.\2579\ The States and Cities commenters appear, at the
                root, to be contesting the agency's determination of maximum feasible
                standards, by way of arguing that NHTSA must interpret the
                technological feasibility factor as necessarily driving greater energy
                conservation. The balancing of factors to determine maximum feasible
                standards is a separate issue, for which EPCA/EISA gives NHTSA
                considerable discretion.
                ---------------------------------------------------------------------------
                 \2579\ Id., see also 75 FR at 25605 (May 7, 2010).
                ---------------------------------------------------------------------------
                 The States and Cities commenters focus on previous rulemaking
                language when they suggest that the agency was arbitrary and capricious
                for not explaining more fully why it need not expressly evaluate every
                single technology that does or could exist in MYs 2021-2026. While
                NHTSA stated in 2012 that it had ``considered all types of technologies
                that improve real-world fuel economy, including air-conditioner
                efficiency and other off-cycle technology, PHEVs, EVs, and highly-
                advanced internal combustion engines not yet in production,'' \2580\
                that statement was only one in a larger discussion. The 2012 final rule
                also stated expressly that ``[t]here are a number of other potential
                technologies available to manufacturers in meeting the 2017-2025
                standards that the agencies have evaluated but have not considered in
                our final analyses. These include HCCI, 'multi-air', and camless valve
                actuation, and other advanced engines currently under development.''
                \2581\ (emphasis added) Thus, even under the prior analysis that some
                commenters appear to prefer, it is not entirely correct to say that
                NHTSA had considered all technologies in existence or that could exist,
                because some technologies were clearly and purposely left out of the
                prior rule's analysis. In response to commenters' apparent confusion
                regarding NHTSA's statement that it did not consider technologies that
                improved fuel economy in ``similar ways'' as other technologies
                discussed in the NPRM, the meaning behind that statement was discussed
                at greater length in the section of the NPRM that substantively covered
                those technologies. For example, in discussing the ``HCR2'' technology,
                the agencies explained that while the agencies were not modeling HCR2
                expressly due to concerns that it remained ``entirely speculative,''
                ``[t]he CAFE model allows for incremental improvement over existing
                HCR1 technologies with the addition of improved accessory devices
                (IACC), a technology that is available to be applied on many baseline
                MY 2016 vehicles with HCR1 engines and may be applied as part of a
                pathway of compliance to further improve the effectiveness of existing
                HCR1 engines.'' \2582\ In this and in other instances, technologies
                included in the analysis improved fuel economy in similar ways to other
                technologies not included. Here, HCR1, when combined with IACC, results
                in ``a step past'' HCR1, which is similar to the unproven HCR2. As in
                the 2012 rule, the agencies explained in the NPRM why certain
                technologies were not considered, and sought comment. In response to
                comments received, some technologies have been added to the analysis
                for the final rule. See Section VI for more information.
                ---------------------------------------------------------------------------
                 \2580\ 77 FR at 63037 (Oct. 15, 2012).
                 \2581\ 77 FR at 62706 (Oct. 15, 2012).
                 \2582\ 83 FR at 43038 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 While the agencies respond to many of UCS's analytical concerns in
                Sections IV and VI (which include extensive discussion of changes made
                in response to comments), NHTSA recognizes that some commenters believe
                that more technologies are ``available for deployment'' more widely,
                and sooner, than the final rule's analysis reflects. This question has
                long been a topic of debate in CAFE and CO2 rulemakings--the
                agencies consider which technologies can be applied to which vehicles
                in which model years in order to assess the costs and benefits of
                pushing the industry to reach different levels of standards, which in
                turn helps to inform stringency determinations. In response to
                comments, the agencies have expanded the number of technologies and the
                vehicles to which they may be applied for this final rule, but continue
                to disagree that certain technologies can be applied widely in the
                rulemaking timeframe. NHTSA does not believe, for example, that HCCI
                will be unavailable for widespread application in the rulemaking
                timeframe because it wishes to believe this prediction--NHTSA believes
                it based on the fact that HCCI has been in the research phase for
                several decades, and the only production applications to date use a
                highly-limited version that restricts HCCI combustion to a very narrow
                range of engine operating conditions. Section VI contains further
                discussion of these issues.
                (2) Economic Practicability
                 ``Economic practicability'' has traditionally referred to whether a
                standard is one ``within the financial capability of the industry, but
                not so stringent as to'' lead to ``adverse economic consequences, such
                as a significant loss of jobs or unreasonable elimination of consumer
                choice.'' \2583\ In evaluating economic practicability, NHTSA considers
                the uncertainty surrounding future market conditions and consumer
                demand for fuel economy alongside consumer demand for other vehicle
                attributes. NHTSA has explained in the past that this factor can be
                especially important during rulemakings in which the auto industry is
                facing significantly adverse economic conditions (with corresponding
                risks to jobs). Consumer acceptability is also a major component of
                economic practicability,\2584\ which can involve
                [[Page 25132]]
                consideration of anticipated consumer responses not just to increased
                vehicle cost, but also to the way manufacturers may change vehicle
                models and vehicle sales mix in response to CAFE standards. In
                attempting to determine the economic practicability of attribute-based
                standards, NHTSA considers a wide variety of elements, including the
                annual rate at which manufacturers can increase the percentage of their
                fleet that employs a particular type of fuel-saving technology,\2585\
                and manufacturer fleet mixes. NHTSA also considers the effects on
                consumer affordability resulting from costs to comply with the
                standards, and consumers' valuation of fuel economy, among other
                things.
                ---------------------------------------------------------------------------
                 \2583\ 67 FR 77015, 77021 (Dec. 16, 2002).
                 \2584\ See, e.g., Center for Auto Safety v. NHTSA (CAS), 793
                F.2d 1322 (DC Cir. 1986) (Administrator's consideration of market
                demand as component of economic practicability found to be
                reasonable); see also Public Citizen v. NHTSA, 848 F.2d 256
                (Congress established broad guidelines in the fuel economy statute;
                agency's decision to set lower standards was a reasonable
                accommodation of conflicting policies).
                 \2585\ For example, if standards effectively require
                manufacturers to make technologies widely available that consumers
                do not want, or to make technologies widely available before they
                are ready to be widespread, NHTSA believes that these standards
                could potentially be beyond economically practicable.
                ---------------------------------------------------------------------------
                 Prior to the MYs 2005-2007 rulemaking under the non-attribute-based
                (fixed value) CAFE standards, NHTSA generally sought to ensure the
                economic practicability of standards in part by setting them at or near
                the capability of the ``least capable manufacturer'' with a significant
                share of the market, i.e., typically the manufacturer whose fleet mix
                was, on average, the largest and heaviest, generally having the highest
                capacity and capability so as not to limit the availability of those
                types of vehicles to consumers. In the first several rulemakings
                establishing attribute-based standards, NHTSA applied marginal cost-
                benefit analysis, considering both overall societal impacts and overall
                consumer impacts. Whether the standards maximize net benefits has thus
                been a significant, but not dispositive, factor in the past for NHTSA's
                consideration of economic practicability. Executive Order 12866, as
                amended by Executive Order 13563, states that agencies should ``select,
                in choosing among alternative regulatory approaches, those approaches
                that maximize net benefits . . .'' In practice, however, agencies,
                including NHTSA, must consider that the modeling of net benefits does
                not capture all considerations relevant to economic practicability.
                Therefore, as in past rulemakings, NHTSA explained in the NPRM that it
                was considering net societal impacts, net consumer impacts, and other
                related elements in the consideration of economic practicability.
                 NHTSA's consideration of economic practicability depends on a
                number of elements. Expected availability of capital to make
                investments in new technologies matters; manufacturers' expected
                ability to sell vehicles with certain technologies matters; likely
                consumer choices matter; and so forth. NHTSA explained in the NPRM that
                NHTSA's analysis of the impacts of the proposal incorporated
                assumptions to capture aspects of consumer preferences, vehicle
                attributes, safety, and other elements relevant to an impacts estimate;
                but stated that it is difficult to capture every such constraint.
                Therefore, NHTSA explained, it is well within the agency's discretion
                to deviate from the level at which modeled net benefits are maximized
                if the agency concludes that that level would not represent the maximum
                feasible level for future CAFE standards. Economic practicability is
                complex, and like the other factors must also be considered in the
                context of the overall balancing and EPCA's overarching purpose of
                energy conservation. Depending on the conditions of the industry and
                the assumptions used in the agency's analysis of alternative standards,
                NHTSA stated that it could well find that standards that maximize net
                benefits, or that are higher or lower, could be at the limits of
                economic practicability, and thus potentially the maximum feasible
                level, depending on how the other factors are balanced.
                 NHTSA also stated in the NPRM that while the agency would discuss
                safety as a separate consideration, NHTSA also considered safety as
                closely related to, and in some circumstances a subcomponent of,
                economic practicability. On a broad level, manufacturers have finite
                resources to invest in research and development. Investment into the
                development and implementation of fuel saving technology necessarily
                comes at the expense of investing in other areas such as safety
                technology. On a more direct level, when making decisions on how to
                equip vehicles, manufacturers must balance cost considerations to avoid
                pricing further consumers out of the market. As manufacturers add
                technology to increase fuel efficiency, they may decide against
                installing additional safety equipment to reduce cost increases. And as
                the price of vehicles increase beyond the reach of more consumers, such
                consumers continue to drive or purchase older, less safe vehicles. In
                assessing practicability, NHTSA also considers the harm to the Nation's
                economy caused by highway fatalities and injuries.
                 CARB, the States and Cities commenters, and UCS all commented that
                the NPRM analysis, as the States and Cities put it, had ``inexplicably
                inflat[ed] technology costs and rel[ied] on flawed models to predict
                impacts on vehicle sales.'' \2586\ Both CBD et al. and UCS suggested
                that it was incorrect to assume that manufacturers would pass on 100
                percent of cost increases as price increases to consumers.\2587\ UCS
                further stated that ``The agencies have then strategically excluded
                well-established academic literature to limit the assumptions used to
                define a consumer's willingness to pay in ways that further increase
                costs to consumers and/or decrease the consumer benefits of fuel
                economy and greenhouse gas emissions.'' \2588\ UCS argued that assuming
                full pass-through of cost increases as price increases and assuming
                that consumers may not fully value improvements in fuel economy
                ``arbitrar[ily] . . . depress the sales of highly fuel-efficient
                vehicles in the model by systematically negating consumer benefits of
                these vehicles.'' \2589\ The States and Cities further argued that
                NHTSA had not ``substantiated its concern that an increase in new
                vehicle prices would place a particular burden on `low-income
                purchasers,' '' and stated that NHTSA had ``assume[d], without
                explanation, that'' less-stringent fuel economy standards resulted in
                greater net savings for consumers, which NHTSA ``acknowledge[d],
                without justification, `is a significantly different analytical result
                from the 2012 final rule.' '' \2590\ The States and Cities commenters
                implied that this different result and NHTSA's ``failure to acknowledge
                it'' was impermissible under the standard set forth in Fox
                Television.\2591\
                ---------------------------------------------------------------------------
                 \2586\ CARB, NHTSA-2018-0067-11873, at 79-80; States and Cities,
                NHTSA-2018-0067-11735, at 69-70; UCS, NHTSA-2018-0067-12039, at 4.
                 \2587\ CBD et al., NHTSA-2018-0067-12057, at 4; UCS, NHTSA-2018-
                0067-12039, at 4.
                 \2588\ UCS, NHTSA-2018-0067-12039, at 5.
                 \2589\ Id.
                 \2590\ States and Cities, NHTSA-2018-0067-11735, at 70.
                 \2591\ Id.
                ---------------------------------------------------------------------------
                 A number of commenters stated that the NPRM's estimates of job
                losses associated with the proposal conflicted with NHTSA's concerns
                about job losses if more stringent standards were promulgated. CBD et
                al. argued that NHTSA could not reasonably conclude that job losses
                make less-stringent standards more economically practicable than more-
                stringent
                [[Page 25133]]
                standards.\2592\ The States and Cities commenters stated that ``[b]y
                declining to address its own findings of significant job losses in the
                auto sector, NHTSA has ignored an important aspect of the problem and
                failed to propose a `rational connection between the facts found and
                the choice made.' '' \2593\ The States and Cities commenters also
                argued that ``the agency failed to acknowledge or explain its break
                with its own interpretation and practice of considering whether
                standards would cause a `significant loss of jobs.' '' \2594\ Some
                commenters argued that more-stringent standards would create more jobs
                (and conversely, that less-stringent standards would result in job
                losses), primarily for supplier companies,\2595\ and some noted that
                other studies had concluded that more-stringent standards would
                increase employment, citing, for example, the report by Synapse Energy
                Economics, Inc. on ``Cleaner Cars and Job Creation.'' \2596\ Some
                commenters further argued that less-stringent standards would hurt U.S.
                GDP,\2597\ and some argued that they would hurt U.S. industry's
                international competitiveness because other countries/regions have more
                stringent standards, and investment may shift to those countries if
                U.S. standards do not continue to compel it.\2598\ The States and
                Cities commenters stated that failing to address fully ``the negative
                employment and GDP impacts of the Proposed Rollback is an abdication of
                NHTSA's clear statutory duty to consider the economic practicability of
                its proposed standards, and an impermissible interpretation of the
                statutory text.'' \2599\
                ---------------------------------------------------------------------------
                 \2592\ CBD et al., NHTSA-2018-0067-12057, at 4.
                 \2593\ States and Cities, NHTSA-2018-0067-11735, at 68 (citing
                State Farm, 463 U.S. at 42).
                 \2594\ Id. (citing 83 FR at 43208; Fox Television, 556 U.S. at
                515).
                 \2595\ CBD et al., NHTSA-2018-0067-12057; Alliance for Vehicle
                Efficiency, NHTSA-2018-0067-11696, at 3-4; NESCAUM, NHTSA-2018-0067-
                11691, at 5.
                 \2596\ States and Cities, NHTSA-2018-0067-11735, at 68; UCS,
                NHTSA-2018-0067-12039, at 4.
                 \2597\ States and Cities, NHTSA-2018-0067-11735, at 68; UCS,
                NHTSA-2018-0067-12039, at 4.
                 \2598\ NESCAUM, NHTSA-2018-0067-11691, at 5; Alliance for
                Vehicle Efficiency, NHTSA-2018-0067-11696, at 4.
                 \2599\ States and Cities, NHTSA-2018-0067-11735, at 68 (citing
                49 U.S.C. 32902(f); Chevron, 467 U.S. at 843).
                ---------------------------------------------------------------------------
                 Commenters disagreed on whether and how NHTSA should consider
                consumer demand. Mr. Kreucher, the Texas Congressional
                Delegation,\2600\ and Senator Inhofe,\2601\ among others, all argued
                that considering consumer demand for fuel economy was important, while
                other commenters argued that while it may be permissible for NHTSA to
                consider consumer demand, NHTSA could not elevate that consideration
                above others. CARB and the States and Cities commenters both cited
                language from CAS v. NHTSA for the premise that ``Congress intended
                energy conservation to be a long-term effort that would continue
                through temporary improvements in energy availability. Thus, it would
                clearly be impermissible for NHTSA to rely on consumer demand to such
                an extent that it ignored the overarching goal of fuel conservation.''
                \2602\ The Minnesota agencies stated that ``making sweeping assumptions
                about consumer preferences should not trump the clear public benefit to
                reducing GHG emissions through these standards.'' \2603\ Mr. Kreucher
                commented, in contrast, that consumer preferences are driven entirely
                by ``[l]ong term fuel price expectations and fuel price alone,'' and
                disagreed with the historical ``implicit assumption that if you build
                it customers will come.'' \2604\
                ---------------------------------------------------------------------------
                 \2600\ Texas Congressional Delegation, NHTSA-2018-0067-1421, at
                1.
                 \2601\ Senator Inhofe, NHTSA-2018-0067-1422, at 1.
                 \2602\ CAS v. NHTSA, 793 F.2d 1322, 1340 (D.C. Cir. 1986), cited
                by CARB, NHTSA-2018-0067-11873, at 79, and by States and Cities,
                NHTSA-2018-0067-11735, at 69.
                 \2603\ Minnesota agencies, NHTSA-2018-0067-11706, at 4.
                 \2604\ Kreucher, NHTSA-2018-0067-0444, at 11-12.
                ---------------------------------------------------------------------------
                 The Minnesota agencies argued that focusing on consumer preferences
                represented an ``unreasonable and unprecedented shift in
                interpretation.'' \2605\ The States and Cities commenters stated
                similarly that NHTSA had ``redefined `economically practicable' to
                categorically exclude standards that, based on some unspecified metric,
                `widely apply technologies that consumers do not want,' '' and argued
                that ``NHTSA has offered no explanation for how it would define `wide
                application,' much less how it would supposedly determine what
                consumers do or do not want.'' \2606\ The States and Cities commenters
                argued that it was internally inconsistent (and therefore arbitrary and
                capricious) for NHTSA to rely in its justification on concerns about
                consumer acceptance of technologies, while concurrently ``acknowledging
                the `extensive debate over how much consumers do (and/or should) value
                fuel savings and fuel economy as an attribute in new vehicles.' ''
                \2607\ The States and Cities commenters stated that the NPRM's modeling
                ``assume[ed] that consumers assign no value to fuel savings
                whatsoever,'' and that ``This assumption is not only implausible but
                also flies in the face of the Agency's own statements that consumers
                likely value between half of and all future fuel savings.'' \2608\
                ---------------------------------------------------------------------------
                 \2605\ Minnesota agencies, NHTSA-2018-0067-11706, at 4.
                 \2606\ States and Cities, NHTSA-2018-0067-11735, at 69 (citing
                State Farm, 463 U.S. at 42-43).
                 \2607\ Id. (citing NPRM at 43216; Fox Television, 556 U.S. at
                515, and United States Sugar Corp., 830 F.3d at 650).
                 \2608\ Id. at 70 (citing NPRM at 43073).
                ---------------------------------------------------------------------------
                 With regard to whether consumers do want more fuel economy, NESCAUM
                stated that ``the most recent surveys indicate that consumers continue
                to place a high value on fuel efficient vehicles of all types,'' \2609\
                while Alliance for Vehicle Efficiency stated that ``Consumers have
                adopted incremental changes to new vehicles that increase fuel economy
                that don't compromise on power, size or safety.'' \2610\ The States and
                Cities commenters argued that ``consumer choice is, in fact, enhanced
                by providing consumers with the option of purchasing higher-efficiency
                vehicles.'' \2611\ CBD et al. and the States and Cities commenters
                stated that NHTSA had simply made assertions about consumer demands
                without supporting evidence,\2612\ with the States and Cities
                commenters also arguing that the fuel price assumptions in the NPRM
                were ``unsupported'' and ``contradicted by recent evidence,'' despite
                NHTSA's arguments that low fuel prices made ``fuel efficiency less
                attractive to consumers.'' \2613\ Somewhat in contrast, NESCAUM stated
                that ``[g]iven recent consumer preferences for larger vehicles,
                maximizing fuel efficiency and GHG emission reductions in larger
                footprint vehicles is even more important,'' noting that footprint
                based standards ``are intentionally flexible to accommodate industry
                and consumer preferences.'' \2614\ NESCAUM also stated that many HEV/
                PHEV/EV models are now available and that their sales ``reflect[ ]
                growing consumer acceptance of the technology, . . . despite the low
                availability of electric vehicle models in the Northeast Section 177
                States and the auto industry's continuing failure to actively market
                [them].'' \2615\
                ---------------------------------------------------------------------------
                 \2609\ NESCAUM, NHTSA-2018-0067-11691, at 2.
                 \2610\ Alliance for Vehicle Efficiency, NHTSA-2018-0067-11696,
                at 2.
                 \2611\ States and Cities, NHTSA-2018-0067-11735, at 70.
                 \2612\ CBD et al., NHTSA-2018-0067-12057, at 4; States and
                Cities, NHTSA-2018-0067-11735, at 70.
                 \2613\ Id.
                 \2614\ NESCAUM, NHTSA-2018-0067-11691, at 2.
                 \2615\ Id. at 3.
                ---------------------------------------------------------------------------
                [[Page 25134]]
                 Regarding the NPRM's statement that safety could be a subcomponent
                of economic practicability, the States and Cities commenters stated
                that this was ``an unreasonable interpretation of this factor, given
                that safety concerns are not discussed in EPCA and have no direct
                correlation to whether a standard is economically practicable.'' \2616\
                The States and Cities commenters further stated that ``NHTSA has never
                before analyzed safety considerations as falling under this factor, and
                fails to explain its reason for doing so now,'' \2617\ and said that it
                was ``unmoored from reality'' for NHTSA to state without support that
                ``[i]nvestment into the development and implementation of fuel saving
                technology necessarily comes at the expense of investing in other areas
                such as safety technology.'' \2618\ The States and Cities commenters
                argued that investment in fuel economy rather than safety ``does not
                explain why safety should be folded into a consideration of whether
                standards are economically practicable.'' \2619\ IPI argued that ``[i]t
                is arbitrary for NHTSA to count alleged safety costs as support for its
                propose [sic] rollback both under the economic practicability factor
                and as its own separate `bolster[ing] factor,' and yet never fully
                monetize climate- and pollution-related deaths and other welfare
                impacts under either the need to conserve energy factor nor under the
                economic practicability factor.'' \2620\
                ---------------------------------------------------------------------------
                 \2616\ States and Cities, NHTSA-2018-0067-11735, at 70
                (``arbitrary and capricious for agency to rely on factors `which
                Congress has not intended it to consider' '') (citing Chevron, 467
                U.S. at 843; State Farm, 463 U.S. at 43).
                 \2617\ Id. (citing Fox Television, 556 U.S. at 515).
                 \2618\ Id.
                 \2619\ Id.
                 \2620\ NYU IPI, NHTSA-2018-0067-12213, Appendix, at 6-7.
                ---------------------------------------------------------------------------
                 In response to these comments, NHTSA continues to believe that it
                is reasonable to interpret ``economic practicability'' as the agency
                has long interpreted it: As a question of whether a standard is one
                ``within the financial capability of the industry, but not so stringent
                as to'' lead to ``adverse economic consequences, such as a significant
                loss of jobs or the unreasonable elimination of consumer choice.''
                \2621\ NHTSA disagrees that this interpretation is new or divergent
                from past interpretations of economic practicability--this is, to the
                word, the same interpretation set forth in the 2010 and 2012 final
                rules, and in multiple earlier rules. Commenters disagreeing with the
                NPRM's assessment of economic practicability seem, fundamentally, to be
                disagreeing with how NHTSA applied this interpreted definition of
                economic practicability to the information then before the agency, and
                also with the agency's conclusion of how economic practicability
                weighed against the other statutory factors.
                ---------------------------------------------------------------------------
                 \2621\ 67 FR 77015, 77021 (Dec. 16, 2002).
                ---------------------------------------------------------------------------
                 The following text explains why NHTSA continues to believe that the
                pieces of the analysis it categorizes as relevant to economic
                practicability fit within the long-standing definition of that factor.
                Section VIII.B.4 below will explain how the agency has considered those
                pieces of the analysis in balancing economic practicability with the
                other statutory factors.
                 NHTSA has consistently described the manner in which it applies the
                ``economic practicability'' factor, and has given considerable weight
                to the phrasing of this description. Parsing the words of this
                description can be useful:
                 The core of the description is the phrase ``within the financial
                capability of the industry,'' but not so stringent as to lead to
                ``adverse economic consequences.'' The following clause ``such as a
                significant loss of jobs or the unreasonable elimination of consumer
                choice'' is set off by a comma from ``consequences,'' and use of the
                phrase ``such as'' indicates that it is a nonrestrictive clause.\2622\
                A nonrestrictive clause means that ``significant loss of jobs'' and
                ``unreasonable elimination of consumer choice'' are examples of
                ``adverse economic consequences,'' but are not an exclusive list of the
                possible adverse economic consequences that NHTSA may consider. Further
                evidence that this clause was intended simply to offer examples comes
                from the 1977 final rule establishing passenger car standards for MYs
                1981-1984, in which NHTSA examined the potential meaning of ``economic
                practicability'' at length and concluded that it should be interpreted
                as ``requiring the standards to be within the financial capability of
                the industry, but not so stringent as to threaten substantial economic
                hardship for the industry,'' i.e., lacking the final clause.\2623\
                ---------------------------------------------------------------------------
                 \2622\ See Strunk, William and E.B. White, The Elements of
                Style, Fourth Edition (2000), Rule 3, at 2-7.
                 \2623\ 42 FR 33534, 33537 (Jun. 30, 1977). It is worth noting
                that the agency considered and rejected an interpretation of
                economic practicability at that time based solely on cost-benefit
                analysis, stating ``A cost-benefit analysis would be useful in
                considering these factors [of economic practicability], but sole
                reliance on such an analysis would be contrary to the mandate of the
                act.'' Id.
                ---------------------------------------------------------------------------
                 A number of commenters took issue with NHTSA's consideration of
                consumer demand, citing the 1986 D.C. Circuit decision CAS v. NHTSA for
                the proposition that consumer demand cannot drive the balancing of
                factors in determining maximum feasible standards. In that case, the
                D.C. Circuit stated that ``[i]t is axiomatic that Congress intended
                energy conservation to be a long term effort that would continue
                through temporary improvements in energy availability. Thus, it would
                clearly be impermissible for NHTSA to rely on consumer demand to such
                an extent that it ignored the overarching goal of fuel conservation.''
                \2624\ NHTSA agrees that the CAS decision makes this point, and that
                the 9th Circuit decision in CBD v. NHTSA also underscored that the
                overarching purpose of EPCA is energy conservation. That said, the CAS
                decision also contains a number of other points that are relevant both
                to the facts at hand in this rulemaking and NHTSA's current use of
                consumer demand as an aspect of economic practicability and as a
                consideration in determining maximum feasible standards. NHTSA will
                discuss CAS more extensively below in Section VIII.B.4, but this
                section will cover it briefly, specifically with respect to NHTSA's
                interpretation of economic practicability.
                ---------------------------------------------------------------------------
                 \2624\ CAS, 793 F.2d 1322, 1340 (D.C Cir. 1986).
                ---------------------------------------------------------------------------
                 As noted in the NPRM and in the 2012 final rule, the CAS decision
                found NHTSA's consideration of market demand as a component of economic
                practicability reasonable.\2625\ In CAS, petitioners the Center for
                Auto Safety, Public Citizen, Union of Concerned Scientists, and
                Environmental Policy Institute sued NHTSA over CAFE standards for MY
                1986, arguing that NHTSA could not determine stringency on the basis of
                low expected consumer demand for fuel economy, and ``that technology
                permitted greater fuel savings and that the statutorily required
                `maximum feasible' level of fuel economy is higher than the standard''
                determined by NHTSA.\2626\ The court followed Chevron in evaluating
                whether NHTSA could consider consumer demand, and found that Congress
                had not directly spoken to the consideration of consumer demand. The
                court then assessed whether NHTSA's interpretation of the statute
                ``represents a reasonable accommodation of conflicting policies that
                were committed to the agency's care by statute,'' stating that ``The
                agency's interpretation of the statutory requirements is due
                considerable deference and must be
                [[Page 25135]]
                found adequate if it falls within the range of permissible
                constructions.'' \2627\
                ---------------------------------------------------------------------------
                 \2625\ 83 FR at 43208, fn. 402; 77 FR at 62668, fn. 111 (both
                citing CAS, 793 F.2d 1322, 1338 (D.C. Cir. 1986)).
                 \2626\ CAS, at 1328.
                 \2627\ CAS, at 1338.
                ---------------------------------------------------------------------------
                 In assessing NHTSA's interpretation, the court stated that
                ``Consumer demand is not specifically designated as a factor, but
                neither is it excluded from consideration; the factors of
                `technological feasibility' and `economic practicability' are each
                broad enough to encompass the concept. Thus, the unadorned language of
                the statute does not indicate a congressional intent concerning the
                precise objections raised by the petitioners.'' The court then examined
                EPCA's legislative history and concluded that ``this language neither
                precludes nor requires lower standards when consumer demand for heavy
                vehicles is strong. The agency is directed to weigh the `difficulties
                of individual automobile manufacturers;' there is no reason to conclude
                that difficulties due to consumer demand for a certain mix of vehicles
                should be excluded.'' \2628\ The court even noted that ``the
                petitioners [did] not challenge the consideration of consumer demand
                per se, but rather the weight the agency has given the factor in
                downgrading standards . . . .'' \2629\
                ---------------------------------------------------------------------------
                 \2628\ CAS, at 1338-1339.
                 \2629\ CAS, at 1340.
                ---------------------------------------------------------------------------
                 NHTSA continues to believe that it is reasonable to consider
                consumer demand as an element of economic practicability, as the CAS
                court recognized. Comments objecting to the consideration of consumer
                demand appear to focus more, like the petitioners in CAS, on the
                agency's focus on consumer demand in the overall balancing of factors
                to determine what CAFE standards would be maximum feasible, insofar as
                they are expressing concern about consumer demand undermining energy
                conservation. Again, this question will be addressed further in Section
                VIII.B.4 below. To the extent that commenters dispute any consideration
                of consumer demand, the D.C. Circuit put that question to rest decades
                ago.
                 Related to the agency's consideration of consumer demand, a number
                of commenters took issue with the agencies' estimates of the cost of
                meeting higher fuel economy standards, arguing essentially that the
                analysis was deliberately constructed to inflate costs and minimize
                consumer willingness to pay for fuel economy improvements in order to
                arrive at a policy conclusion that higher fuel economy standards would
                not be economically practicable. NHTSA does not believe that commenters
                mean to argue with the agency's legal interpretation (i.e., the
                consideration of cost as an aspect of economic practicability), but
                rather with the agencies' analytical findings which inform that
                consideration. Comments on those analytical findings, and the agencies'
                responses and changes to the analysis in response to those comments,
                are discussed in Sections VI and VII above. Consumer willingness to pay
                for additional fuel economy in their new vehicles, in particular, is
                represented throughout the final rule analysis as 2.5 years--that is,
                that consumers value, and manufacturers will voluntarily add, fuel
                economy-improving technology that pays for itself in fuel savings
                within 2.5 years.
                 More generally, NHTSA believes that the cost of meeting CAFE
                standards is inherently relevant to assessing whether those standards
                are ``within the financial capability of the industry but not so
                stringent as to lead to adverse economic consequences,'' for two
                primary reasons. First, vehicle manufacturers tend to have relatively
                fixed budgets for R&D and production, which are tied to overall
                revenues. If more of those budgets are spent on improving fuel economy,
                less of those budgets are available to spend on other vehicle
                characteristics (such as advanced safety features, or better
                performance or utility) that might improve sales. Offering less of
                those other vehicle characteristics in a market where many consumers
                are not particularly focused on fuel economy could lead to adverse
                economic consequences for those manufacturers. Manufacturers cannot
                simply increase budgets or turn limited resources toward supplying more
                of vehicle characteristics that do not motivate most sales. To the
                extent that more stringent standards drive manufacturing costs higher
                and those costs are passed forward to consumers in the form of price
                increases, those price increases can affect vehicle sales to some
                extent. NHTSA understands that some commenters disagree that higher
                manufacturing costs are necessarily passed forward to consumers in the
                way that the agencies have modeled them being passed forward, but the
                agencies do not have adequate information on which to base a different
                approach. Commenters disagreeing with this approach generally object on
                two fronts: First, because they believe that automakers cross-subsidize
                cost increases by raising the prices of certain models rather than all
                models, and second, because they believe that automakers could absorb
                regulatory costs and reduce profits. The agencies do not have enough
                information to model either of those issues in a meaningful way. Some
                amount of cross-subsidization no doubt occurs, but automakers closely
                hold pricing strategy information. The agencies do not attempt to model
                automakers voluntarily reducing profits in response to standards, again
                in part because the agencies do not have sufficient information, but
                also because these companies are publicly-traded and taking losses is
                not a long-term solution for companies whose success is measured by
                profitability. NHTSA believes that the analytical approach used today
                is reasonable given the information available to the agencies. While
                today's analysis does not show large sales effects due to price
                increases, and even accounting for fuel economy differences in this
                final rule still does not show large sales effects, it seems reasonable
                to call negative sales effects ``adverse economic consequences.''
                 Also related to consumer demand, NHTSA has previously considered
                manufacturer ``shortfalls'' as an aspect of economic
                practicability.\2630\ The CAFE standards are corporate average
                standards, by definition, giving manufacturers the flexibility to
                decide how to distribute fuel economy-improving technologies throughout
                their fleet. In other words, no given vehicle need, itself, meet a
                standard or even its ``target'' on the target curve, as long as the
                fleet as a whole meets the standard. However, CAFE compliance is
                measured on a sales-weighted basis, so if a manufacturer ultimately
                sells more vehicles that perform poorly relative to their targets than
                it sells vehicles that beat their targets, the manufacturer may fall
                short of its compliance obligation despite having applied fuel economy-
                improving technologies in amounts that the manufacturer originally
                anticipated would result in compliance. Recent compliance trends have
                illustrated this phenomenon, as discussed in Section IV above. When
                fuel is relatively inexpensive, Americans tend to be less interested in
                saving money on fuel, and thus less interested in fuel economy as
                compared to other vehicle attributes. Compliance shortfalls represent
                this consumer decision-making playing out in the market, and can thus
                be evidence of economic impracticability if sufficiently
                widespread.\2631\
                ---------------------------------------------------------------------------
                 \2630\ See 77 FR at 63040-43 (Oct. 15, 2012).
                 \2631\ See, e.g., Alliance comments (Full Comment Set) at 25-29,
                describing automaker shortfalls in terms of fleet fuel economy
                increases required by augural and prior standards.
                ---------------------------------------------------------------------------
                 As with the above-discussed aspects of economic practicability,
                commenters who objected to NHTSA's consideration
                [[Page 25136]]
                of employment impacts disagreed less with the principle of considering
                employment impacts, and more with how NHTSA discussed employment
                impacts in the proposal's justification given the NPRM's findings on
                employment. Namely, the NPRM included a simplistic analysis that
                converted reduced technology costs under the preferred alternative
                relative to the augural standards into ``job years'' metric and
                estimated U.S. auto sector labor would be slightly reduced under the
                proposal as compared to under the augural standards (reflecting those
                reduced technology costs). Although new vehicle sales increased
                slightly under the NPRM's preferred alternative, this was offset
                because ``manufacturing, integrating, and selling less technology means
                using less labor to do so.'' \2632\ However, NHTSA expressed concern in
                the proposal justification section that ``there could be potential for
                . . . loss of U.S. jobs . . . under nearly all if not all of the
                regulatory alternatives considered . . . .'' \2633\ A number of
                commenters argued that if more stringent standards led to higher
                employment, as the NPRM (and also outside analyses) appeared to show,
                there was no way that less stringent standards could be more
                economically practicable.
                ---------------------------------------------------------------------------
                 \2632\ 83 FR at 43436 (Aug. 24, 2018).
                 \2633\ Id. at 43216.
                ---------------------------------------------------------------------------
                 As in the NPRM, NHTSA recognizes that the employment analysis for
                this final rule does not capture certain potential effects that may be
                important. NHTSA explained in the NPRM that the NPRM's employment
                analysis did not account for the risks that vehicle sales may be facing
                a bubble situation, or that manufacturers facing higher production
                costs might choose to move production overseas.\2634\ This topic is
                discussed at greater length in Section VIII.B.4 below.
                ---------------------------------------------------------------------------
                 \2634\ Id. at 43224-25.
                ---------------------------------------------------------------------------
                 Commenters addressing NHTSA's consideration of safety as an aspect
                of economic practicability argued generally that EPCA did not call for
                discussion of safety concerns, and that it was unreasonable to assume
                that requiring higher levels of fuel economy might preclude investment
                in further vehicle safety improvements. NHTSA has already explained
                above that the long-standing definition of ``economic practicability''
                lists example ``adverse economic consequences'' in a nonrestrictive
                clause format, meaning that other things besides employment and
                consumer choice impacts could cause economic consequences and be
                relevant to economic practicability. NHTSA believes that it is
                reasonable and appropriate to consider some aspects of safety as part
                of its consideration of economic practicability, because NHTSA
                continues to believe that vehicle manufacturers have finite budgets for
                R&D and production that may be spent on fuel economy improvements when
                they may otherwise be spent on safety improvements, among other things
                that consumer value. Some commenters said that that was not a
                reasonable assumption, but it is supported by statements from vehicle
                manufacturers,\2635\ and NHTSA does not have a reason to disbelieve
                that companies have limited budgets. Moreover, case law does not object
                to consideration of safety as an aspect of economic
                practicability.\2636\ With regard to IPI's comment about monetization
                of climate and pollution-related deaths and other welfare impacts, the
                social cost of carbon and criteria pollutant damages estimates are
                intended to account for these impacts, and are considered both as part
                of the cost-benefit analysis and under the environmental implications
                aspect of the need of the U.S. to conserve energy. Given that the
                decision about what standards are ``maximum feasible'' is made by
                considering all of the factors, it is therefore less relevant under
                which factor a given issue is considered, so long as it is
                appropriately considered. To the extent that IPI disagrees with those
                estimated valuations, Section VI discusses comments on those topics and
                the agencies' responses.
                ---------------------------------------------------------------------------
                 \2635\ See, e.g., Toyota comments at 6, NHTSA-2018-0067-12098
                (``There are now more realistic limits placed on the number of
                engines and transmissions in a powertrain portfolio which better
                recognizes manufacturers must manage limited engineering resources
                and control supplier, production, and service costs.'').
                 \2636\ Competitive Enterprise Institute v. NHTSA, 901 F.2d 107,
                120, n. 11 (``Petitioners have never clearly identified the precise
                statutory basis on which safety concerns should be factored into the
                CAFE scheme, although they alluded to occupant safety as part of the
                `economic practicability' criterion in their MY 1989 petition to
                NHTSA and at oral argument. We do not find this failure fatal,
                however, because NHTSA has always examined the safety consequences
                of the CAFE standards in its overall consideration of relevant
                factors since its earliest rulemaking under the CAFE program,
                (citations omitted). Moreover, NHTSA itself believes Congress was
                cognizant of safety issues when it enacted the CAFE program. As
                evidence, NHTSA discusses a congressional report that dealt with the
                safety consequences of a downsized fleet of cars which had been
                considered by Congress during its enactment of the CAFE program.'').
                ---------------------------------------------------------------------------
                 Based on the above, NHTSA continues to believe that its
                interpretation of economic practicability is reasonable. Section
                VIII.B.4 will discuss how NHTSA has considered and balanced economic
                practicability for this final rule, and also respond to comments that
                addressed the NPRM's application of economic practicability to the
                information before the agency at that time.
                (3) The Effect of Other Motor Vehicle Standards of the Government on
                Fuel Economy
                 ``The effect of other motor vehicle standards of the Government on
                fuel economy'' involves analysis of the effects of compliance with
                emission, safety, noise, or damageability standards on fuel economy
                capability and thus on average fuel economy. In many past CAFE
                rulemakings, NHTSA has said that it considers the adverse effects of
                other motor vehicle standards on fuel economy. It said so because, from
                the CAFE program's earliest years \2637\ until recently, the effects of
                such compliance on fuel economy capability over the history of the CAFE
                program have been negative ones. For example, safety standards that
                have the effect of increasing vehicle weight thereby lower fuel economy
                capability, thus decreasing the level of average fuel economy that
                NHTSA can determine to be feasible. In the analyses for both the NPRM
                and this final rule, NHTSA has considered the additional weight that it
                estimates would be added in response to new safety standards during the
                rulemaking timeframe.\2638\ NHTSA has also accounted for EPA's ``Tier
                3'' standards for criteria pollutants in its estimates of technology
                effectiveness in both the NPRM and final rule analyses.\2639\
                ---------------------------------------------------------------------------
                 \2637\ 42 FR 63184, 63188 (Dec. 15, 1977). See also 42 FR 33534,
                33537 (Jun. 30, 1977).
                 \2638\ PRIA, Chapter 5; FRIA, Section 5.
                 \2639\ PRIA, Chapter 6; FRIA, Section 6.
                ---------------------------------------------------------------------------
                 NHTSA discussed in the NPRM whether to consider EPA's
                CO2 standards as an ``other motor vehicle standard of the
                Government'' among the other regulations typically considered, and if
                so, how. NHTSA explained that in the 2012 final rule establishing CAFE
                standards for MYs 2017-2021, NHTSA recognized that ``To the extent the
                GHG standards result in increases in fuel economy, they would do so
                almost exclusively as a result of inducing manufacturers to install the
                same types of technologies used by manufacturers in complying with the
                CAFE standards.'' \2640\ NHTSA concluded in 2012 that ``no further
                action was needed'' because ``the agency had already considered EPA's
                [action] and the harmonization benefits of the National Program in
                developing its own [action].'' \2641\
                ---------------------------------------------------------------------------
                 \2640\ 77 FR 62624, 62669 (Oct. 15, 2012).
                 \2641\ Id.
                ---------------------------------------------------------------------------
                [[Page 25137]]
                 In the NPRM, NHTSA considered the issue afresh, and determined that
                it was clear based on a purely textual analysis of the statutory
                language that EPA's CO2 standards applicable to light-duty
                vehicles are literally ``other motor vehicle standards of the
                Government,'' in that they are standards set by a Federal agency that
                apply to motor vehicles. Basic chemistry makes fuel economy and
                tailpipe CO2 emissions two sides of the same coin, as
                discussed at length above, and when two agencies functionally regulate
                both (because when regulating fuel economy, CO2 emissions
                are necessarily also regulated, and vice versa), it would be absurd not
                to link the standards.\2642\ The global warming potential of
                N2O, CH4, and HFC emissions are not closely
                linked with fuel economy, but neither do they affect fuel economy
                capabilities. Simply concluding that EPA's CO2 standards
                were ``other motor vehicle standards of the Government,'' however, did
                not answer how should NHTSA should consider them.
                ---------------------------------------------------------------------------
                 \2642\ In fact, EPA includes tailpipe CH4, CO, and
                CO2 in the measurement of tailpipe CO2 for
                CO2 compliance using a carbon balance equation so that
                the measurement of tailpipe CO2 exactly aligns with the
                measurement of fuel economy for the CAFE compliance.
                ---------------------------------------------------------------------------
                 NHTSA acknowledged in the NPRM that some stakeholders had
                previously suggested that NHTSA should implement this statutory factor
                by letting EPA decide what CO2 standards are appropriate and
                reasonable under the CAA and then simply setting CAFE standards with
                reference to CO2 stringency. NHTSA disagreed that such an
                approach would be a reasonable interpretation of EPCA, explaining that
                while EPA and NHTSA consider some similar factors under the CAA and
                EPCA/EISA, respectively, they are not identical, and standards that are
                appropriate under the CAA may not be ``maximum feasible'' under EPCA/
                EISA, and vice versa. Moreover, NHTSA explained, considering EPCA's
                language in the context in which it was written, it seemed unreasonable
                to conclude that Congress intended EPA to dictate CAFE stringency. In
                fact, Congress clearly separated NHTSA's and EPA's responsibilities for
                CAFE under EPCA by giving NHTSA authority to set standards and EPA
                authority to measure and calculate fuel economy. If Congress had wanted
                EPA to set CAFE standards, it could have given that authority to EPA in
                EPCA or at any point since Congress amended EPCA.\2643\
                ---------------------------------------------------------------------------
                 \2643\ The NPRM noted, for instance, that EISA was passed after
                the Massachusetts v. EPA decision by the Supreme Court. If Congress
                had wanted to amend EPCA in light of that decision, it would have
                done so at that time, but did not.
                ---------------------------------------------------------------------------
                 NHTSA explained that NHTSA and EPA are obligated by Congress to
                exercise their own independent judgment in fulfilling their statutory
                missions, even though both agencies' regulations affect both fuel
                economy and CO2 emissions. Because of this relationship, it
                is incumbent on both agencies to coordinate and look to one another's
                actions to avoid unreasonably burdening industry through inconsistent
                regulations,\2644\ but both agencies' programs must stand on their own
                merits. As with other recent CAFE and CO2 rulemakings, NHTSA
                explained that the agencies were continuing do all of these things in
                the proposal.
                ---------------------------------------------------------------------------
                 \2644\ Massachusetts v. EPA, 549 U.S. 497, 532 (2007) (``[T]here
                is no reason to think the two agencies cannot both administer their
                obligations and yet avoid inconsistency.'').
                ---------------------------------------------------------------------------
                 With regard to standards issued by the State of California, the
                NPRM explained that State tailpipe standards (whether for
                CO2 or for other pollutants) do not qualify as ``other motor
                vehicle standards of the Government'' under 49 U.S.C. 32902(f), and
                that therefore, NHTSA would not consider them as such in proposing
                maximum feasible average fuel economy standards. NHTSA explained that
                States may not adopt or enforce standards related to fuel economy
                standards, which are preempted under EPCA, regardless of whether EPA
                granted any waivers under the Clean Air Act (CAA).
                 NHTSA and EPA agreed in the NPRM that State tailpipe CO2
                emissions standards do not become Federal standards and qualify as
                ``other motor vehicle standards of the Government,'' when subject to a
                CAA preemption waiver. NHTSA stated that EPCA's legislative history
                supports that position, as follows:
                 EPCA, as initially passed in 1975, mandated average fuel economy
                standards for passenger cars beginning with model year 1978. The law
                required the Secretary of Transportation to establish, through
                regulation, maximum feasible fuel economy standards \2645\ for model
                years 1981 through 1984 with the intent to provide steady increases to
                achieve the standard established for 1985 and thereafter authorized the
                Secretary to adjust that standard.
                ---------------------------------------------------------------------------
                 \2645\ As is the case today, EPCA required the Secretary to
                determine ``maximum feasible average fuel economy'' after
                considering technological feasibility, economic practicability, the
                effect of other Federal motor vehicle standards on fuel economy, and
                the need of the Nation to conserve energy. 15 U.S.C. 2002(e)
                (recodified July 5, 1994).
                ---------------------------------------------------------------------------
                 For the statutorily-established standards for model years 1978-
                1980, EPCA provided each manufacturer with the right to petition for
                changes in the standards applicable to that manufacturer. A petitioning
                manufacturer had the burden of demonstrating a ``Federal fuel economy
                standards reduction'' was likely to exist for that manufacturer in one
                or more of those model years and that it had made reasonable technology
                choices. ``Federal standards,'' for that limited purpose, included not
                only safety standards, noise emission standards, property loss
                reduction standards, and emission standards issued under various
                Federal statutes, but also ``emissions standards applicable by reason
                of section 209(b) of [the CAA].'' \2646\ (Emphasis added). Critically,
                all definitions, processes, and required findings regarding a Federal
                fuel economy standards reduction were located within a single self-
                contained subsection of 15 U.S.C. 2002 that applied only to model years
                1978-1980.\2647\
                ---------------------------------------------------------------------------
                 \2646\ Section 202 of the CAA (42 U.S.C. 7521) requires EPA to
                prescribe air pollutant emission standards for new vehicles; Section
                209 of the CAA (42 U.S.C. 7543) preempts state emissions standards
                but allows California to apply for a waiver of such preemption.
                 \2647\ As originally enacted as part of Public Law 94-163, that
                subsection was designated as section 502(d) of the Motor Vehicle
                Information and Cost Savings Act.
                ---------------------------------------------------------------------------
                 In 1994, Congress recodified EPCA. As part of this recodification,
                the CAFE provisions were moved to Title 49 of the United States Code.
                In doing so, unnecessary provisions were deleted. Specifically, the
                recodification eliminated subsection (d). The House report on the
                recodification declared that the subdivision was ``executed,'' and
                described its purpose as ``[p]rovid[ing] for modification of average
                fuel economy standards for model years 1978, 1979, and 1980.'' \2648\
                It is generally presumed, when Congress includes text in one section
                and not in another, that Congress knew what it was doing and made the
                decision deliberately.
                ---------------------------------------------------------------------------
                 \2648\ H.R. Rep. No. 103-180, at 583-584, tbl. 2A.
                ---------------------------------------------------------------------------
                 NHTSA stated in the NPRM that it had previously considered the
                impact of California's Low Emission Vehicle standards in establishing
                fuel economy standards and occasionally has done so under the ``other
                standards'' sections.\2649\ During the 2012 rulemaking, NHTSA sought
                comment on the appropriateness of considering California's tailpipe
                CO2 emission standards in this section and concluded that
                doing so was unnecessary.\2650\ In light of the legislative history
                discussed above, however, NHTSA stated in the NPRM that such
                consideration would be inappropriate, and confirms that consideration
                of California's LEV
                [[Page 25138]]
                standards as among the ``other standards of the Government'' was
                inappropriate.
                ---------------------------------------------------------------------------
                 \2649\ See, e.g., 68 FR 16896, 71 FR 17643.
                 \2650\ See 77 FR 62669.
                ---------------------------------------------------------------------------
                 Commenters addressing criteria pollutant standards generally
                supported NHTSA's approach in the NPRM. AFPM commented that NHTSA
                ``must consider the effect on fuel economy of EPA's Title II standards,
                including the use of catalytic converters, PM traps and other
                technologies that address emissions and have a fuel economy impact.''
                \2651\ Ford also stated that previous analyses ``did not assess the
                impact of the criteria pollutant emission standards that were adopted
                subsequent to the [2012 final rule],'' which Ford said ``increased the
                challenge of meeting the fuel economy and GHG targets and should be
                taken into consideration.'' \2652\ Ford stated that the NPRM
                appropriately included ``updat[ed] core engine maps using correct,
                regular-grade octane test fuel,'' and that it accounts for ``ultra-low
                2025 MY Tier 3 and LEVIII emissions standards [which] will require
                aggressive cold start strategies [that] consume additional fuel at
                start-up in order to rapidly heat the catalyst to an effective
                operating temperature, which degrades CO2 and fuel economy
                performance on the FTP test [and] was not considered previously. . .
                .'' \2653\
                ---------------------------------------------------------------------------
                 \2651\ AFPM, NHTSA-2018-0067-12078, at 52.
                 \2652\ Ford, NHTSA-2018-0067-11928, at 7.
                 \2653\ Id.
                ---------------------------------------------------------------------------
                 Regarding how NHTSA should consider EPA's CO2 standards
                as ``other motor vehicle standards of the Government,'' ACEEE suggested
                amongst its comments that, in considering EPA's CO2
                standards, ``NHTSA should not weaken its program . . . to compensate
                for . . . inevitable, modest differences'' between EPA's and NHTSA's
                programs.\2654\ ``Indeed, to the extent that differences in the
                requirements of the two programs remain, it is clear that the more
                stringent requirement in any given respect should govern the
                obligations of the manufacturer.'' \2655\ AFPM commented similarly that
                ``Although NHTSA must consider the effect of other governmental
                regulations, Congress intended that NHTSA would have exclusive
                authority over a single set of national fuel economy standards.''
                \2656\ Mr. Dotson expressed his belief that ``Congress was cognizant of
                the relationship between EPCA and the Clean Air Act when crafting
                EISA'' and cited and discussed various types of legislative history for
                the proposition that EISA had not limited EPA's CAA authority, and that
                various legislative efforts to do so had been put forth in some fashion
                and had failed.\2657\
                ---------------------------------------------------------------------------
                 \2654\ ACEEE, NHTSA-2018-0067-12122, joint NGO comment to
                Alliance/Global petition for flexibilities, at 3.
                 \2655\ Id.
                 \2656\ AFPM, NHTSA-2018-0067-12078, at 52.
                 \2657\ Dotson, EPA-HQ-OAR-2018-0283-4132, Appendix A, at A2-A23.
                NHTSA disagrees with the persuasiveness of the legislative history
                cited by Mr. Dotson, which includes floor debates, colloquies, and
                other similar information that does not reflect the agreement of the
                Congress as a whole. NHTSA looks to the language Congress actually
                passed and the President signed into law.
                ---------------------------------------------------------------------------
                 NHTSA agrees that while it is appropriate for NHTSA to coordinate
                with and look to EPA's actions to avoid unreasonably burdening industry
                through inconsistent regulations, it would not be appropriate for NHTSA
                to reduce stringency below levels it believes to be maximum feasible
                solely for purposes of accommodating differences between programmatic
                flexibilities. The 2012 final rule clearly stated that while the
                agencies had made efforts to align their standards, programmatic
                differences existed, and how manufacturers chose to rely on compliance
                flexibilities could affect the relative stringency of NHTSA's and EPA's
                standards:
                 We note, however, that the alignment is based on the assumption
                that manufacturers implement the same level of direct A/C system
                improvements as EPA currently forecasts for those model years, and
                on the assumption of PHEV, EV, and FCV penetration at specific
                levels. If a manufacturer implements a higher level of direct A/C
                improvement technology (although EPA predicts 100% of manufacturers
                will use substitute refrigerants by MY 2021, and the GHG standards
                assume this rate of substitution) and/or a higher penetration of
                PHEVs, EVs and FCVs, then NHTSA's standards would effectively be
                more stringent than EPA's. Conversely, if a manufacturer implements
                a lower level of direct A/C improvement technology and/or a lower
                penetration of PHEVs, EVs and FCVs, then EPA's standards would
                effectively be more stringent than NHTSA's. Several manufacturers
                commented on this point and suggested that this meant the standards
                were not aligned, because NHTSA's standards might be more stringent
                in some years than EPA's. This reflects a misunderstanding of the
                agencies' purpose. The agencies have sought to craft harmonized
                standards such that manufacturers may build a single fleet of
                vehicles to meet both agencies' requirements. That is the case for
                these final standards. Manufacturers will have to plan their
                compliance strategies considering both the NHTSA standards and the
                EPA standards and assure that they are in compliance with both, but
                they can still build a single fleet of vehicles to accomplish that
                goal.\2658\
                ---------------------------------------------------------------------------
                 \2658\ 77 FR at 63054-55 (Oct. 15, 2012) (emphasis added).
                Thus, NHTSA has been consistent in its position that CO2
                stringency does not and should not, by itself, dictate CAFE stringency.
                That said, consideration of EPA's standards was inherent in development
                of this final rule, given that the same technologies improve fuel
                economy and reduce CO2 emissions, and given that
                CO2 emissions represent the majority of GHGs produced by
                light-duty vehicles, and given that the agencies have conducted the
                analysis for this rulemaking jointly. NHTSA believes that EPA's
                standards have been fully and appropriately considered as part of its
                decision on these final standards. To be clear, NHTSA did not assert in
                the NPRM that EISA constrained EPA's authorities under the CAA and do
                ---------------------------------------------------------------------------
                not disagree with that aspect of Mr. Dotson's comment.
                 Chemours argued that, contrary to the NPRM's statements about
                having considered EPA's GHG standards in developing the proposal, NHTSA
                had not adequately considered EPA's GHG standards because only the no-
                action alternative reflected EPA regulation of the non-CO2
                GHGs, and the analysis did not otherwise account for the non-
                CO2 GHG standards.\2659\ Chemours stated that those
                standards were ``required, pursuant to CAA section 202(a), to address
                `air pollution' from mobile sources,'' and that ``No assessment was
                done as to whether such standards could be made less stringent in order
                to avoid the various issues identified (e.g., changes in technology
                since the 2012 final rule, costs to consumers, the effect of
                `diminishing returns,' a changed petroleum market and other factors.''
                \2660\
                ---------------------------------------------------------------------------
                 \2659\ Chemours, NHTSA-2018-0067-12018, at 25.
                 \2660\ Id. at 25-26.
                ---------------------------------------------------------------------------
                 NHTSA disagrees that it was necessary for NHTSA to consider EPA's
                standards for non-CO2 GHG emissions any further than as
                discussed above. Regulation of CH4, N2O, and HFCs
                affects fuel economy only indirectly, if at all. As explained above and
                in the 2012 final rule, while NHTSA recognizes that some manufacturers
                may choose paths to compliance with EPA's GHG standards that make their
                compliance with CAFE standards more challenging, the agencies previewed
                this possibility and stated their expectation that manufacturers could
                make these decisions for themselves. To the extent that Chemours is
                asking NHTSA to examine regulatory alternatives reflecting less
                stringent CAFE standards in light of changed conditions since the 2012
                final rule, that is exactly what the NPRM and final rule analyses have
                done.
                 A number of commenters disagreed with NHTSA's explanation of how
                State standards need not be considered under this factor. The States
                and Cities
                [[Page 25139]]
                commenters stated that NHTSA was required to consider State tailpipe
                standards because 49 U.S.C. 32902(f) does not specify that
                ``Government'' refers only to ``Federal'' government; because NHTSA had
                not offered compelling evidence or arguments that Congress did not
                intend NHTSA to consider State tailpipe standards; and because ``case
                law . . . states unequivocally that California's standards must be
                considered by NHTSA under this factor [citing Green Mountain Chrysler's
                ``federalizing'' language].'' \2661\ The States and Cities commenters
                further argued that NHTSA was trying to argue simultaneously that it
                could not consider State standards under the ``other standards'' factor
                but could consider State standards ``under other EPCA factors, if and
                when it sees fit'' (citing NPRM language that technological feasibility
                and economic practicability are broad factors allowing NHTSA to
                consider elements not specifically designated by Congress).\2662\ The
                States and Cities commenters further argued, citing Fox Television,
                that NHTSA was deviating from past practice without a reasoned
                explanation by not specifically requesting comment in the NPRM on the
                fact that it was not considering California's standards as ``other
                motor vehicle standards of the Government.'' \2663\
                ---------------------------------------------------------------------------
                 \2661\ States and Cities, NHTSA-2018-0067-12018, at 71.
                 \2662\ Id. at 71-72.
                 \2663\ Id. at 72. Fox Television did not involve a rulemaking,
                and does not require agencies to specifically seek public comment
                when they deviate from past practice. In any event, by articulating
                in the NPRM that NHTSA was not considering California's standards as
                ``other motor vehicle standards of the Government'' the public had
                ample opportunity to provide comment on this issue, and commenters
                in fact did so as discussed above.
                ---------------------------------------------------------------------------
                 With regard to NHTSA's analysis of EPCA's original language for MYs
                1978-80 and the 1994 positive law recodification, the States and Cities
                commenters stated that ``NHTSA's statutory and legislative history
                arguments related to standards for model years 1978-1980 lack merit, as
                NHTSA has provided no reasonable argument that Congress meant NHTSA to
                consider a wider range of standards for those years than for others,''
                and stated that the section in question ``was removed from the statute
                because it expired, not because Congress took issue with NHTSA's
                consideration of California's waiver standards.'' \2664\ Mr. Dotson
                commented similarly that NHTSA could not rely on the 1994 positive law
                codification as basis to conclude that State tailpipe standards
                (whether for GHGs or other emissions) do not qualify as ``other motor
                vehicle standards of the Government,'' because it said ``without
                substantive change. . . .'' \2665\
                ---------------------------------------------------------------------------
                 \2664\ Id. at 71.
                 \2665\ Dotson, EPA-HQ-OAR-2018-0283-4132, Appendix A, at A23-
                A24.
                ---------------------------------------------------------------------------
                 Additionally, the States and Cities commenters stated that NHTSA
                could not argue that California's emissions standards are not ``other
                motor vehicle standards of the Government'' because they are preempted,
                because NHTSA ``has no authority to decide whether or not California's
                standards are preempted,'' and ``one of the reasons California's
                Advanced Clean Cars program is not preempted by EPCA is because those
                standards are `other motor vehicle standards of the Government' within
                the meaning of EPCA.'' \2666\ Besides this comment, a number of
                comments were submitted regarding NHTSA's statements in the NPRM about
                EPCA's preemption provision and how it applied to California's
                standards. Those comments have been addressed \2667\ as part of the
                separate final rule published on September 27, 2019,\2668\ and will not
                be discussed further as part of this action.
                ---------------------------------------------------------------------------
                 \2666\ States and Cities, NHTSA-2018-0067-12018, at 71.
                 \2667\ To the extent that any individual comment was not
                specifically addressed, NHTSA believes that the substance and themes
                of all substantive comments on EPCA preemption were addressed as
                part of that final rule.
                 \2668\ 84 FR 51310.
                ---------------------------------------------------------------------------
                 NHTSA affirms that its interpretation set forth in the NPRM that
                ``other motor vehicle standards of the Government'' does not apply to
                State emissions standards that relate to fuel economy. NHTSA does not
                understand how 49 U.S.C. 32919 could be given effect if the purpose of
                the ``other motor vehicle standards of the Government'' provision is to
                compel their inclusion in NHTSA's decision-making. NHTSA continues to
                disagree with the two district court cases suggesting that the ``other
                motor vehicle standards of the Government'' provision obviates 49
                U.S.C. 32919, as explained at some length in the ``One National
                Program'' final rule preceding this regulatory action.\2669\ NHTSA
                refers readers to that document for more detail on this topic.
                ---------------------------------------------------------------------------
                 \2669\ See, e.g., 84 FR at 51323 (Sep. 27, 2019).
                ---------------------------------------------------------------------------
                 With regard to State tailpipe standards that do not directly relate
                to fuel economy, NHTSA continues to believe that Congress's original
                direction to consider ``emissions standards applicable by reason of
                section 209(b) of [the CAA]'' applied only to CAFE standards for MYs
                1978-1980, as discussed in the NPRM. NHTSA agrees that the 1994
                positive law recodification was not intended to make substantive
                changes to EPCA; the NPRM explained that, in dropping Section 502(d),
                Congress made clear that that provision was executed, and that
                provision expressly directed NHTSA to consider State standards that had
                been granted preemption waivers under CAA 209(b). In order for States
                even to have their own emissions standards for motor vehicles,
                California must be granted a waiver of preemption under CAA section
                209(b). If Congress had intended for NHTSA to continue to consider
                State tailpipe standards post-MY 1980, the direction to consider
                emissions standards that had been granted Section 209 waivers could
                have been placed elsewhere in the statute. Congress did not do
                so.\2670\ While NHTSA may have considered State tailpipe standards in
                the past, it is not bound to do so, and NHTSA does not believe that it
                is unreasonable to consider those standards under technological
                feasibility or economic practicability if they are to be considered.
                ---------------------------------------------------------------------------
                 \2670\ The negative inference canon is logically and reasonably
                employed here, particularly given that, as a factual matter and as
                discussed further below, considering EPA's Tier 3 standards (which
                are clearly ``other motor vehicle standards of the Government'')
                effectively accounts for the technological implications of
                California's LEVIII standards.
                ---------------------------------------------------------------------------
                 State tailpipe standards primarily affect fuel economy by requiring
                gasoline ICE vehicles to burn additional fuel when the engine first
                starts. For most gasoline engines on the road today, the majority of
                tailpipe NOX, NMOG, and CO emissions occur during ``cold
                start,'' before the three-way catalyst has reached the very high
                temperature (e.g., 900-1000 [deg]F), at which point it is able to
                convert (through oxidation and reduction reactions) those emissions
                into less harmful derivatives. By strictly limiting the amount of those
                emissions, tailpipe smog standards require the catalyst to be brought
                to temperature extremely quickly, so modern vehicles employ cold start
                strategies that intentionally release fuel energy into the engine
                exhaust to heat the catalyst to the relevant temperature as quickly as
                possible. The additional fuel that must be used to heat the catalyst is
                typically referred to as a ``cold-start penalty,'' meaning that
                vehicle's fuel economy (over a test cycle) is reduced because the fuel
                consumed to heat the catalyst did not go toward the goal of moving the
                vehicle forward.\2671\ The Autonomie
                [[Page 25140]]
                work employed to develop technology effectiveness estimates for this
                final rule does, in fact, account for cold-start penalties.\2672\ The
                Autonomie model documentation discusses the fact that cold-start
                penalties were derived from an EPA database of MY 2016 vehicles, which
                would have met both EPA and California smog standards. Moreover, EPA
                regulations allow manufacturers to employ LEVIII data for Tier 3
                compliance. Based on all of these factors, NHTSA believes that the
                negative fuel economy effects of California's tailpipe standards for
                smog-related emissions are reasonably represented in the analysis for
                the final rule, regardless of whether NHTSA was obligated by law to
                consider them expressly.
                ---------------------------------------------------------------------------
                 \2671\ For more information on this, see, e.g., Pihl, Josh A.,
                et al., ``Development of a Cold Start Fuel Penalty Metric for
                Evaluating the Impact of Fuel Composition Changes on SI Engine
                Emissions Control,'' Oak Ridge National Laboratory, 2018. Available
                at https://www.osti.gov/biblio/1462896-development-cold-start-fuel-penalty-metric-evaluating-impact-fuel-composition-changes-si-engine-emissions-control.
                 \2672\ See ANL Model Documentation, Section 6.1.5, available in
                Docket No. NHTSA-2018-0067.
                ---------------------------------------------------------------------------
                 Ultimately, it would be illogical for NHTSA to consider legally
                unenforceable standards to be ``other motor vehicle standards of the
                Government.'' That is the case for State standards preempted by EPCA.
                While NHTSA understands that certain commenters disagree with a
                separate final rule that NHTSA issued concerning EPCA preemption, and
                the particular State standards that NHTSA considers preempted by EPCA,
                those issues are outside the scope of this final rule.
                (4) The Need of the United States To Conserve Energy
                 NHTSA has historically interpreted ``the need of the United States
                to conserve energy'' to mean ``the consumer cost, national balance of
                payments, environmental, and foreign policy implications of our need
                for large quantities of petroleum, especially imported petroleum.''
                \2673\
                ---------------------------------------------------------------------------
                 \2673\ 42 FR 63184, 63188 (Dec. 15, 1977).
                ---------------------------------------------------------------------------
                (a) Consumer Costs and Fuel Prices:
                 NHTSA explained in the NPRM that fuel for vehicles costs money for
                vehicle owners and operators. All else equal--a critical caveat--
                consumers benefit from vehicles that need less fuel to perform the same
                amount of work. Future fuel prices are a critical input into the
                economic analysis of potential CAFE standards because they determine
                the value of fuel savings both to new vehicle buyers and to society,
                the amount of fuel economy that the new vehicle market is likely to
                demand in the absence of new standards, and they inform NHTSA about the
                ``consumer cost . . . of our need for large quantities of petroleum.''
                In the proposal, NHTSA's analysis relied on fuel price projections from
                the U.S. Energy Information Administration's (EIA) Annual Energy
                Outlook (AEO) for 2017; in the final rule, on fuel price projections
                derived from the version of NEMS used to produce AEO 2019. Federal
                government agencies generally use EIA's price projections in their
                assessment of future energy-related policies.
                 Several commenters stated that consumer costs for fuel were an
                important consideration. ACEEE stated that ``The average U.S. household
                still spent nearly $2,000 on gasoline and motor oil (directly) in 2017,
                making oil savings very relevant for consumers,'' and argued that ``Oil
                price volatility remains a threat to U.S. consumers and businesses--the
                price of crude oil has more than doubled since 2016, belying the
                theoretical suggestion in the notice that conditions for oil price
                shocks no longer exist,'' suggesting that further fuel efficiency
                improvements were necessary to protect consumers.\2674\ NESCAUM
                commented that prior analyses had suggested that consumers would save
                $6,000 on net, after paying more for their vehicles upfront, and that
                the proposal would cost consumers more in fuel.\2675\ Both NESCAUM and
                the States and Cities commenters stated that higher fuel costs would
                disproportionately affect low-income consumers, who spend a higher
                share of their income on fuel costs.\2676\ The Congressional Tri-Caucus
                commented that ``As we see oil prices rising again, it makes no sense
                for DOT to roll back these standards.'' \2677\ The States and Cities
                commenters argued that increased gas expenditures would result ``in
                negative economy-wide effects'' for many years ``given that cars sold
                in the model years for which NHTSA proposes to freeze standards will,
                according to the Agencies, be on the road for decades,'' and stated
                that ``NHTSA's analysis is arbitrary and capricious because it entirely
                fails to consider how the Proposed Rollback would impact consumers and
                the economy as a whole due to increased gasoline expenditures.'' \2678\
                The States and Cities commenters further argued that NHTSA was
                incorrect in the NPRM when it interpreted ``the relevant question for
                the need of the U.S. to conserve energy is not whether there will be
                any movement in prices but whether that movement will be sudden and
                large,'' \2679\ and cited State Farm to say that NHTSA had ``failed to
                consider an important aspect of the problem'' by ``failing to analyze
                the likely impact of even moderate future increases and volatility in
                fuel prices.'' \2680\
                ---------------------------------------------------------------------------
                 \2674\ ACEEE, NHTSA-2018-0067-12122, at 2.
                 \2675\ NESCAUM, NHTSA-2018-0067-11691, at 4.
                 \2676\ NESCAUM, NHTSA-2018-0067-11691, at 5; States and Cities,
                NHTSA-2018-0067-11735, at 75, citing Synapse Report.
                 \2677\ Congressional Tri-Caucus, NHTSA-2018-0067-1424, at 2.
                 \2678\ States and Cities, NHTSA-2018-0067-11735, at 75.
                 \2679\ 83 FR at 43214, n. 444.
                 \2680\ States and Cities, NHTSA-2018-0067-11735, at 75.
                ---------------------------------------------------------------------------
                 A number of commenters addressed consumer willingness to pay more
                money upfront in order to save money on fuel costs. Many of these
                comments are addressed in Section VI.C as part of the discussion of how
                sales are modeled. More specifically in the context of how NHTSA
                interprets the need of the U.S. to conserve energy, IPI commented that
                NHTSA was incorrect that ``consumers' need to save money is now `less
                urgent' and no longer supports a strong overall need to conserve
                energy. The agencies assert that past rulemakings were overly and
                paternalistically focused on `myopia.' This statement ignores all the
                other pathways through which the 2012 standards benefit consumers' need
                to save money, including by correcting informational asymmetries,
                attention costs, and other informational failures; positional
                externalities; and various other supply-side and demand-side
                explanations for consumers' inability to achieve in an unregulated
                market the level of fuel economy that they desire. These components of
                the national need to conserve energy are discussed at length throughout
                these comments, and were specifically considered by the agencies in the
                2012 rule.'' \2681\
                ---------------------------------------------------------------------------
                 \2681\ IPI, NHTSA-2018-0067-12213, Appendix, at 5-6.
                ---------------------------------------------------------------------------
                 Several commenters disagreed with NHTSA's suggestion in the NPRM
                that increasing U.S. production and exports reduced volatility in the
                oil market. Securing America's Energy Future stated that ``. . . recent
                events are an important validation of public policies that support
                long-term goals like efficiency and fuel diversity. Indeed, in the
                absence of fuel-efficiency standards, global oil price volatility would
                likely render the country even more exposed to oil price shocks than it
                is currently.'' \2682\ Mr. Bordoff, IPI, the States and Cities
                commenters, and UCS all commented that the oil market is global, so
                increasing U.S. production does not prevent price shocks that occur
                [[Page 25141]]
                due to non-U.S. events or circumstances. Mr. Bordoff stated that ``In a
                globalized oil market, the consequence of a supply disruption anywhere
                is a price increase everywhere--regardless of how much oil the U.S.
                imports.'' \2683\ UCS made similar comments.\2684\ Mr. Bordoff further
                commented that U.S. gasoline prices still follow the fluctuations in
                global crude oil prices regardless of the U.S. oil import/export
                balance,\2685\ and stated that ``Gasoline prices at the pump are
                especially sensitive to changes in the global crude oil price due to
                the relatively low level of fuel taxation [in the U.S.] compared to
                other OECD countries.'' \2686\ Mr. Bordoff stated that gas price spikes
                are still possible due to ongoing geopolitical challenges in major oil
                producing areas, and concluded that ``Continuing with planned fuel
                economy increases through CAFE standards is one effective way to reduce
                the oil intensity of the economy and mitigate the adverse impact of
                future oil price increases on American drivers.'' \2687\ The States and
                Cities commenters cited to and echoed Mr. Bordoff's comments on this
                point.\2688\ CARB commented that the proposal had relied on AEO 2017,
                which reflected fuel prices that still assumed the augural standards
                remained in place, but that AEO 2018 assumes ``no new fuel efficiency
                standard'' and held fuel economy flat after 2021, and showed fuel
                prices would be higher.\2689\
                ---------------------------------------------------------------------------
                 \2682\ Securing America's Energy Future, NHTSA-2018-0067-12172,
                at 7.
                 \2683\ Bordoff, EPA-HQ-OAR-2018-0283-3906, at 6.
                 \2684\ UCS, NHTSA-2018-0067-12039, at 7.
                 \2685\ IPI cited and echoed these comments. IPI, NHTSA_2018-
                0067-12213, Appendix, at 3.
                 \2686\ Bordoff, EPA-HQ-OAR-2018-0283-3906, at 7.
                 \2687\ Id. at 10-12.
                 \2688\ States and Cities, NHTSA-2018-0067-11735, at 74-75.
                 \2689\ CARB, NHTSA-2018-0067-11783, at 318.
                ---------------------------------------------------------------------------
                 Mr. Bordoff also commented that the future of shale oil in the U.S.
                was uncertain, and therefore increased U.S. oil production was not a
                basis on which to assume future global price stability.\2690\ Mr.
                Bordoff argued that ``Although shale oil is more responsive to price
                changes than conventional supply, it cannot serve as a swing supplier
                to stabilize oil markets in the way true spare capacity (held by Saudi
                Arabia) can. It takes at least 6-12 months for U.S. shale to respond to
                price changes.'' \2691\ Bordoff continued, stating that ``For example,
                although shale oil is more responsive to oil prices, oil prices still
                plunged below $30 per barrel at the start of 2016 and soared to $80 per
                barrel earlier this year. Shale oil could not swing quickly enough to
                stabilize markets. This role fell to OPEC instead in both cases, first
                to put a floor under prices by cutting supply and, more recently, to
                provide relief by ramping up production.'' \2692\ Bordoff further
                commented that political or popular pressures due to environmental
                concerns may significantly increase the cost and/or difficulty of
                expanding shale infrastructure,\2693\ and that even disregarding
                uncertainty in supply, ongoing uncertainty in demand (both U.S. and
                abroad) also contributed to global price uncertainty.\2694\
                ---------------------------------------------------------------------------
                 \2690\ Bordoff, EPA-HQ-OAR-2018-0283-3906, at 3.
                 \2691\ Id., at 7.
                 \2692\ Id., at 7-8.
                 \2693\ Id., at 9-10.
                 \2694\ Id., at 3.
                ---------------------------------------------------------------------------
                 NHTSA agrees with commenters that consumer costs for fuel are
                relevant to the need of the U.S. to conserve energy. NHTSA also agrees
                that future fuel prices are uncertain, and that shale oil development
                in the U.S. is (1) still proceeding and subject to uncertainty, (2)
                very different from traditional sources like Saudi Arabia, and (3) not
                enough, by itself, to preclude any possibility of major swings in
                future global oil prices. That said, NHTSA continues to believe that
                U.S. shale development may reduce the negative price effects of global
                price swings due to events and situations outside of our borders. Shale
                represents a large, new, relatively-geopolitically-stable oil supply
                source, and traditional oil producers appear to understand that
                stabilizing prices below the price at which shale production starts to
                ramp up faster helps those traditional producers take market advantage
                of their lower cost of production.\2695\ The net effect of this, for
                American drivers, should be greater fuel price stability, at least at
                the upper end of fuel prices. NHTSA also continues to believe that, for
                purposes of considering consumer cost of fuel as part of the need of
                the U.S. to conserve energy, the fact that Americans' gasoline costs
                might be minutely lower under more stringent CAFE standards and
                minutely higher under comparatively less stringent CAFE standards is
                not dispositive by itself. There is some tolerance in the market for
                some amount of fluctuation in fuel prices, as evidenced by the
                discussion in Section VI. Slow increases in fuel prices are relatively
                easy for households to absorb; sharp increases are more difficult.
                ---------------------------------------------------------------------------
                 \2695\ Since 1995, EIA data indicates that OPEC production
                roughly stabilized in late 2016 and has either remained steady or
                fallen since then. See https://www.eia.gov/opendata/qb.php?category=1039874&sdid=STEO.PAPR_OPEC.M. See also Ilya
                Arkhipov, Will Kennedy, Olga Tanas, and Grant Smith, ``Putin Dumps
                MBS to Start a War on America's Shale Oil Industry,'' March 7, 2020,
                Bloomberg News, available at https://www.bloomberg.com/news/articles/2020-03-07/putin-dumps-mbs-to-start-a-war-on-america-s-shale-oil-industry (describing the collapse of the OPEC+ coalition);
                EIA, ``This Week in Petroleum--OPEC shift to maintain market share
                will result in global inventory increases and lower prices,'' March
                11, 2020, https://www.eia.gov/petroleum/weekly/; DOE, ``DOE Responds
                to Recent Oil Market Activity,'' March 9, 2020, https://www.energy.gov/articles/doe-responds-recent-oil-market-activity.
                ---------------------------------------------------------------------------
                 Increases in CAFE stringency reduce the effects of all types of
                increases in fuel prices, at least to the extent that people can buy
                new cars and trucks, but as discussed below in Section VIII.B.4, fuel
                costs and per-vehicle costs balance against one another for many
                buyers. With respect to relatively low U.S. gasoline taxes creating
                more pass-through effects of global oil price fluctuations, that would
                be true regardless of stringency. Broadly speaking, while consumer fuel
                costs are an important consideration of the need of the U.S. to
                conserve energy, at this time NHTSA believes, as discussed in Section
                VI, that American consumers generally understand fuel costs and their
                tolerance for fluctuations, and tend to purchase vehicles accordingly.
                Requiring consumers to save more fuel over the longer term by spending
                more money upfront on new vehicle purchases may involve more tradeoffs
                than suggested in prior rulemakings, and this rulemaking seeks to keep
                these possible tradeoffs in mind.
                (b) National Balance of Payments:
                 As the NPRM explained, the need of the United States to conserve
                energy has historically included consideration of the ``national
                balance of payments'' because of concerns that importing large amounts
                of oil created a significant wealth transfer to oil-exporting countries
                and left the U.S. economically vulnerable.\2696\ As recently as 2009,
                nearly half the U.S. trade deficit was driven by petroleum,\2697\ yet
                this concern has largely laid fallow in more recent CAFE actions,
                arguably in part because other factors besides petroleum consumption
                have since played a bigger role in the U.S. trade deficit. Given
                [[Page 25142]]
                recent significant increases in U.S. oil production and corresponding
                decreases in oil imports, this concern seems likely to remain fallow
                for the foreseeable future.\2698\ Increasingly, changes in the price of
                fuel have come to represent transfers between domestic consumers of
                fuel and domestic producers of petroleum rather than gains or losses to
                foreign entities. NHTSA explained in the NPRM that some commenters have
                lately raised concerns about potential economic consequences for
                automaker and supplier operations in the U.S. due to disparities
                between CAFE standards at home and their counterpart fuel economy/
                efficiency and CO2 standards abroad. NHTSA finds these
                concerns more relevant to technological feasibility and economic
                practicability than to the national balance of payments. Moreover, to
                the extent that an automaker decides to globalize a vehicle platform to
                meet more stringent standards in other countries, that automaker would
                comply with United States' standards and additionally generate
                overcompliance credits that it can save for future years if facing
                compliance concerns, or sell to other automakers. While CAFE standards
                are set at maximum feasible rates, efforts of manufacturers to exceed
                those standards are rewarded not only with additional credits but a
                market advantage in that those consumers who place a large weight on
                fuel savings will find such vehicles that much more attractive.
                ---------------------------------------------------------------------------
                 \2696\ See 42 FR 63184, 63192 (Dec. 15, 1977) (``A major reason
                for this need [to reduce petroleum consumption] is that the
                importation of large quantities of petroleum creates serious balance
                of payments and foreign policy problems. The United States currently
                spends approximately $45 billion annually for imported petroleum.
                But for this large expenditure, the current large U.S. trade deficit
                would be a surplus.'').
                 \2697\ See Today in Energy: Recent improvements in petroleum
                trade balance mitigate U.S. trade deficit, U.S. Energy Information
                Administration (July 21, 2014), https://www.eia.gov/todayinenergy/detail.php?id=17191.
                 \2698\ For an illustration of recent increases in U.S.
                production, see, e.g., U.S. crude oil and liquid fuels production,
                Short-Term Energy Outlook, U.S. Energy Information Administration
                (June 2018), https://www.eia.gov/outlooks/steo/images/fig13.png.
                While it could be argued that reducing oil consumption frees up more
                domestically-produced oil for exports, and thereby raises U.S. GDP,
                that is neither the focus of the CAFE program nor consistent with
                Congress' original intent in EPCA. EIA's Annual Energy Outlook (AEO)
                series provides midterm forecasts of production, exports, and
                imports of petroleum products, and is available at https://www.eia.gov/outlooks/aeo/.
                ---------------------------------------------------------------------------
                 Several commenters addressed how much oil the U.S. imports, and the
                assumptions about imports in the NPRM analysis. Securing America's
                Energy Future commented that ``Because there are no readily available
                substitutes to oil in the U.S. transportation sector, volatile crude
                oil and petroleum product prices represent an enduring threat to the
                U.S. economy.'' \2699\ ACEEE commented that overall U.S. oil imports
                are higher now than they were in 1975, and nearly as high as they were
                in 2012, and also stated that compared to a small overall trade surplus
                in 1975, ``the U.S. now runs a large overall trade deficit.'' \2700\
                The States and Cities commenters made a similar point, arguing that the
                U.S. still imports large amounts of petroleum; that imports made up
                about 25 percent of total U.S. oil consumption in 2017; and that EIA
                indicates that ``imports as a share of oil consumption in the United
                States are only about 10% lower today as compared to 1975, and we are
                producing the same amount of crude oil domestically today as we were in
                1970.'' \2701\ IPI stated that EIA analysis shows that the ``U.S. will
                continue to import crude oil through 2050 and `remains a net importer
                of petroleum and other liquids on an energy basis.' '' \2702\ CARB
                disagreed that the U.S. was projected to become a net petroleum
                exporter, and stated that even if it were, the rollback would have
                negative effects on the U.S., because (1) it ignores short-run damages
                caused by increased oil consumption and imports; (2) relies on
                projections of net imports of oil which also do not take account of the
                effects of the proposed rule; and (3) is not supported by the
                evidence.\2703\
                ---------------------------------------------------------------------------
                 \2699\ Securing America's Energy Future, NHTSA-2018-0067-12172,
                at 6.
                 \2700\ ACEEE, NHTSA-2018-0067-12122, at 2.
                 \2701\ States and Cities, NHTSA-2018-0067-11735, at 76.
                 \2702\ IPI, NHTSA-2018-0067-12213, Appendix, at 3.
                 \2703\ CARB, NHTSA-2018-0067-11873, at 317.
                ---------------------------------------------------------------------------
                 Regarding assumptions about oil imports in the NPRM analysis, the
                States and Cities commented that in 2016 the agencies had assumed that
                ``90% of fuel savings from existing standards would lead directly to a
                reduction in imported oil,'' and argued that the NPRM analysis had
                ignored that previous assumption and ``la[id] great emphasis on the
                fact that `oil imports have declined while exports have increased'
                since 2005.'' \2704\ IPI argued that the NPRM analysis was internally
                inconsistent, assuming in NHTSA's need of the nation discussion that
                ``additional gasoline consumption will be entirely domestic,'' while
                ``upstream emissions calculations assume that 95% of increased
                consumption will either be from foreign refining or from foreign crude
                imports,'' and suggested that this inconsistency was purposeful to make
                the NPRM analysis look more favorable to the proposal.\2705\ ACEEE
                commented that ``The EIA AEO side cases suggest that reduced oil demand
                will primarily reduce oil imports, thus improving the overall balance
                of trade regardless of the narrow balance of trade in petroleum.''
                \2706\
                ---------------------------------------------------------------------------
                 \2704\ States and Cities, NHTSA-2018-0067-11735, at 75.
                 \2705\ IPI, NHTSA-2018-0067-12213, Appendix, at 3-4.
                 \2706\ ACEEE, NHTSA-2018-0067-12122, at 2.
                ---------------------------------------------------------------------------
                 Regarding the effects on the U.S. economy of increasing U.S. oil
                production, Mr. Morris agreed with the NPRM's suggestion that U.S.
                self-sufficiency in petroleum supply meant that higher consumer
                payments for fuel under less-stringent CAFE standards would be
                transfers within the U.S. economy, and stated that ``[a]t that point,
                the initial purpose of EPCA is entirely obviated.'' \2707\ The States
                and Cities commenters, in contrast, argued that focusing on this effect
                meant that NHTSA essentially claims that increasing revenues of oil
                companies--which report annual profits in the billions--is an even
                trade-off for adding cost pressures and oil-price shock exposure to
                American households.'' \2708\ The States and Cities commenters stated
                that ``. . .this assertion ignores the negative economic impacts that
                would result from increasing the cost burden on oil consumers,'' and
                was ``. . .so implausible that it could not be ascribed to a difference
                of view or the product of agency expertise,' citing State Farm, 463
                U.S. at 43.\2709\
                ---------------------------------------------------------------------------
                 \2707\ Morris (GWU RSC), EPA-HQ-OAR-2018-0283-4028, at 15.
                 \2708\ States and Cities, NHTSA-2018-0067-11735, at 76.
                 \2709\ Id.
                ---------------------------------------------------------------------------
                 As discussed above, NHTSA agrees that oil is a global commodity.
                Living in a globalized economy necessarily means that supply
                disruptions (and thus, price effects) can come from a great variety of
                sources--this was why the CAFE program was created, in recognition of
                this risk. Increasing U.S. energy independence reduces this risk. There
                are two ways to increase petroleum independence: To use less petroleum,
                and to produce more of our own petroleum and use less petroleum
                purchased from abroad. Both approaches work, and both are being
                followed today.
                 NHTSA also agrees that the Draft TAR text describes the analytical
                assumption that for every gallon of fuel not consumed as a result of
                more stringent standards, imported crude would be reduced by 0.9
                gallons. The Draft TAR stated that this assumption was based on
                ``changes in U.S. crude oil imports and net petroleum products in the
                AEO 2015 Reference Case in comparison [sic] the Low (i.e., Economic
                Growth) Demand Case,'' and also on a 2013 paper by Paul Leiby which
                ``suggests that `Given a particular reduction in oil demand stemming
                from a policy or significant technology change, the fraction of oil use
                savings that shows up as reduced U.S. imports, rather than reduced
                U.S., supply, is actually quite
                [[Page 25143]]
                close to 90 percent, and probably close to 95 percent.' '' \2710\
                ---------------------------------------------------------------------------
                 \2710\ Draft TAR, 2016, Chapter 10, Endnote 39, p. 10-59.
                ---------------------------------------------------------------------------
                 EIA data clearly states that while the U.S. still relies on oil
                imports, it is producing an increasingly large share of the petroleum
                it consumes.\2711\ In 2018, domestic petroleum production made up 86
                percent of domestic consumption, while imports made up 11 percent. EIA
                data also clearly states that U.S. reliance on petroleum imports peaked
                in 2005 and has declined since then, and that the import-percentage-of-
                consumption in 2018 was the lowest it has been since 1957--this despite
                the fact that overall U.S. petroleum consumption has increased
                significantly over that time period as the on-road fleet has grown and
                VMT (both individual and collective) has increased. Of the 11 percent
                of oil consumed that was imported, 43 percent came from Canada, and 16
                percent came from Persian Gulf countries. AEO 2019 states that under
                its Reference case assumptions, which it describes as a ``best
                assessment'' and ``a reasonable baseline case,'' \2712\ the U.S.
                remains projected to become a net exporter of petroleum liquids by
                2020.\2713\ During several weeks in 2019, the U.S. also exported more
                oil than it imported.\2714\
                ---------------------------------------------------------------------------
                 \2711\ EIA, ``Oil: Crude and Petroleum Products Explained, Oil
                Imports and Exports,'' updated May 29, 2019, available at https://www.eia.gov/energyexplained/oil-and-petroleum-products/imports-and-exports.php.
                 \2712\ AEO 2019, at 5.
                 \2713\ AEO 2019, at 14.
                 \2714\ See https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=pet&s=wttntus2&f=4.
                ---------------------------------------------------------------------------
                 U.S. Census data indicate that the U.S. balance of trade has
                generally grown over time, although it has fluctuated since peaking in
                2006.\2715\ U.S. Census data further indicate that the U.S. petroleum
                balance of trade, in particular, has fluctuated over time, peaking in
                2008 at roughly -$386 million and decreasing to -$50 million in 2018.
                2019 trends demonstrate further decreases. In percentage terms,
                petroleum trade as a percentage of total trade went from roughly 52
                percent in 1992 (the earliest year for which Census appears to have
                data online), to 47 percent in 2008, to less than 6 percent in 2018. In
                terms of national balance of payments, this is fairly clear evidence
                that petroleum has decreased rapidly as part of the problem. Part of
                this is due to improvements in fleet fuel economy over time, and part
                is due to increases in U.S. production, particularly in the last
                several years.
                ---------------------------------------------------------------------------
                 \2715\ ``U.S. Trade in Goods and Services--Balance of Payments
                (BOP) Basis,'' June 6, 2019, available at https://www.census.gov/foreign-trade/statistics/historical/gands.pdf.
                ---------------------------------------------------------------------------
                 NHTSA notes also that the Draft TAR previewed the possibility of
                this outcome, discussing the ``Shale Oil Revolution'' and the fact that
                ``[t]he recent economics literature on whether oil shocks are the
                threat to economic stability that they once were is mixed.'' \2716\ The
                Draft TAR stated that because of increased U.S. shale oil production,
                ``The resulting decrease in foreign imports . . . effectively permits
                U.S. supply to act as a buffer against artificial or other supply
                restrictions (the latter due to conflict or a natural disaster, for
                example).'' \2717\
                ---------------------------------------------------------------------------
                 \2716\ See Draft TAR at 10-30--10-33.
                 \2717\ Draft TAR at 10-31.
                ---------------------------------------------------------------------------
                 Since the Draft TAR was issued, U.S. shale production has developed
                even further, and U.S. petroleum imports have continued to fall. If
                more oil is being produced in the U.S., and more of domestic
                consumption comes from domestic production, then even though oil is a
                global commodity and thus subject to price changes resulting from non-
                U.S. events, the U.S. economy is inherently better off. When money
                moves around within the U.S. instead of having to leave the U.S., and
                everyone's needs are being met, U.S. citizens are better off when
                things outside the U.S. go wrong--this is what NHTSA means when it
                refers to within-U.S. transfers not being a bad thing as compared to
                greater reliance on imports for consumption needs. To the extent that
                some commenters find within-U.S. transfers problematic because they
                increase U.S. oil company revenues without reducing fuel cost burdens
                on consumers, NHTSA notes that, as discussed above, consumers seem
                willing and able to tolerate some amount of fuel price increases and
                fluctuation risk, as evidenced by their purchasing decisions. Prices
                may still fluctuate, but shortages may foreseeably be reduced.
                 The Draft TAR stated that ``despite continuing uncertainty about
                oil market behavior and outcomes and the sensitivity of the U.S.
                economy to oil shocks, it is generally agreed that it is beneficial to
                reduce petroleum fuel consumption from an energy security standpoint.
                It is not just imports alone, but both imports and consumption of
                petroleum from all sources and their role in economic activity, that
                may expose the U.S. to risk from price shocks in the world oil price.
                Reducing fuel consumption reduces the amount of domestic economic
                activity associated with a commodity whose price depends on volatile
                international markets.'' NHTSA continues to agree with these
                statements, but cannot ignore the fact that increased U.S. petroleum
                production represents the other side of the coin. Again, both national
                balance of payments and energy security can be improved on both the
                supply side and the demand side. While today's final rule continues to
                improve on the demand side by setting standards that continue to push
                CAFE levels upward, it also recognizes that supply side improvements
                are playing a role.
                (c) Environmental Implications
                 The NPRM explained that higher fleet fuel economy can reduce U.S.
                emissions of CO2 as well as various other pollutants by
                reducing the amount of oil that is produced and refined for the U.S.
                vehicle fleet, but can also increase emissions by reducing the cost of
                driving, which can result in increased vehicle miles traveled (i.e.,
                the rebound effect). Thus, the net effect of more stringent CAFE
                standards on emissions of each pollutant depends on the relative
                magnitudes of its reduced emissions in fuel refining and distribution
                and increases in its emissions from vehicle use. Fuel savings from CAFE
                standards also necessarily result in lower emissions of CO2,
                the main gas emitted as a result of refining, distribution, and use of
                transportation fuels. Reducing fuel consumption directly reduces
                CO2 emissions because the primary source of transportation-
                related CO2 emissions is fuel combustion in internal
                combustion engines.
                 NHTSA has considered environmental issues, both within the context
                of EPCA and the context of the National Environmental Policy Act
                (NEPA), in making decisions about the setting of standards since the
                earliest days of the CAFE program. As courts of appeal have noted in
                three decisions stretching over the last 20 years,\2718\ NHTSA defined
                ``the need of the United States to conserve energy'' in the late 1970s
                as including, among other things, environmental implications. In 1988,
                NHTSA included climate change concepts in its CAFE notices and prepared
                its first environmental assessment addressing that subject.\2719\ It
                cited concerns about climate change as one of its reasons for limiting
                the extent of its reduction of the CAFE standard for MY 1989 passenger
                [[Page 25144]]
                cars.\2720\ Since then, NHTSA has considered the effects of reducing
                tailpipe emissions of CO2 in its fuel economy rulemakings
                pursuant to the need of the United States to conserve energy by
                reducing petroleum consumption.
                ---------------------------------------------------------------------------
                 \2718\ CAS, 793 F.2d 1322, 1325 n. 12 (D.C. Cir. 1986); Public
                Citizen, 848 F.2d 256, 262-63 n. 27 (D.C. Cir. 1988) (noting that
                ``NHTSA itself has interpreted the factors it must consider in
                setting CAFE standards as including environmental effects''); CBD,
                538 F.3d 1172 (9th Cir. 2007).
                 \2719\ 53 FR 33080, 33096 (Aug. 29, 1988).
                 \2720\ 53 FR 39275, 39302 (Oct. 6, 1988).
                ---------------------------------------------------------------------------
                 Many commenters addressed the environmental implications of CAFE
                standards and the proposal. ACEEE stated that ``The environmental need
                to save energy is much greater than we realized in 1975,'' and that
                ``The notice argues that since improved standards will not by
                themselves solve global warming, they are not necessary. That logic
                would equally suggest that since no one soldier would win a war, we
                should never deploy any troops. No one measure will solve global
                warming. . . . vehicle standards have been the most important.'' \2721\
                The Harvard environmental law clinic commenters similarly stated that
                ``It is illogical to argue against taking a single step on the basis
                that a single step is insufficient to reach one's goal,'' and commented
                that it was unreasonable for the DEIS to state that ``[t]he emission
                reductions necessary to keep global emissions within this carbon budget
                could not be achieved solely with drastic reductions in emissions from
                the U.S. passenger car and light truck fleet.'' \2722\ UCS also argued
                that with respect to the environmental implications of the standards,
                NHTSA's ``argument that the augural standards would only limit global
                warming by 0.02 degrees C in 2100 actually supports the need to
                maintain the standards. That a single U.S. policy could make that much
                difference in limiting global warming is, in fact, quite significant.''
                \2723\
                ---------------------------------------------------------------------------
                 \2721\ ACEEE, NHTSA-2018-0067-12122, main comments, at 2.
                 \2722\ Harvard environmental law clinic, EPA-HQ-OAR-2018-0283-
                5486, at 13.
                 \2723\ UCS, NHTSA-2018-0067-12039, at 7.
                ---------------------------------------------------------------------------
                 The States and Cities commenters objected to NHTSA's consideration
                in the NPRM of ``whether rapid ongoing increases in CAFE stringency . .
                . can sufficiently address climate change to merit their costs,''
                arguing that NHTSA had ``completely disregard[ed] environmental costs''
                contrary to NHTSA's own long-standing approach to CAFE standards.\2724\
                The States and Cities commenters then framed the CO2 impacts
                of the proposal in tons (specifically, 7,400 million metric tons
                additional CO2 emitted by 2100 as compared to the augural
                standards) and argued that ``the agency effectively ignores its own
                findings, in a sharp and unexplained break with the agency's past
                practice of considering climate impacts,'' citing Fox Television, 556
                U.S. at 515 and the 2010 and 2012 final CAFE rules which discussed
                reduced economic damages from lower climate impacts for those standards
                compared to their baselines.\2725\ IPI also argued that if NHTSA had
                focused on economic damages rather than fractions of degrees Celsius,
                ``Once climate damages are fully monetized (as the agencies are
                required to do), it will become apparent that the proposed rollback
                will cause billions of dollars in climate damages. Billions of dollars
                lost to avoidable climate damages is not a small effect, and it very
                clearly is a `destructive and wasteful' effect.'' \2726\ CARB also
                argued that the NPRM had ``wholly fail[ed] to analyze the economic
                effects of the climate change and public health implications of the
                rollback,'' stating that [t]he Agencies assert these are insignificant,
                but that is only because the Agencies' projections of climate change
                are so extreme. An appropriate analysis of a proposal that speeds
                progress toward such a calamitous condition must acknowledge and
                analyze the expected effects.'' \2727\
                ---------------------------------------------------------------------------
                 \2724\ States and Cities, NHTSA-2018-0067-11735, at 73.
                 \2725\ Id.
                 \2726\ IPI, NHTSA-2018-0067-12213, Appendix, at 4-5.
                 \2727\ CARB, NHTSA-2018-0067-11873, Detailed Comments, at 84.
                ---------------------------------------------------------------------------
                 The States and Cities commenters also argued that NHTSA had not
                explained what the NPRM's definition of ``conservation'' as meaning
                ``avoid[ing] wasteful or destructive use'' ``actually means and how it
                changes the agency's past practice of considering environmental
                impacts,'' citing State Farm, 463 U.S. at 43, and Fox Television, 556
                U.S. at 515.\2728\
                ---------------------------------------------------------------------------
                 \2728\ States and Cities, NHTSA-2018-0067-11735, at 73.
                ---------------------------------------------------------------------------
                 Regarding non-climate impacts, IPI commented that the NPRM ``only
                briefly mention[ed] the possible effects on other emissions without
                detailing any of the myriad non-climate public health and welfare
                consequences from pollution associated with petroleum production and
                combustion for motor vehicles.'' \2729\ The States and Cities
                commenters similarly stated that ``NHTSA's evaluation of this factor
                fails to include any analysis of environmental costs related to air
                quality,'' and that the NPRM/DEIS analysis substantially understates
                the actual impacts of the Proposed Rollback on criteria air pollutants
                (such as NOX and PM) and air toxics (such as benzene),
                making it inappropriate to rely upon.'' \2730\
                ---------------------------------------------------------------------------
                 \2729\ IPI, NHTSA-2018-0067-12213, Appendix, at 5.
                 \2730\ States and Cities, NHTSA-2018-0067-11735, at 73-74.
                ---------------------------------------------------------------------------
                 NHTSA agrees that the NPRM considered environmental implications of
                the standards somewhat differently from past rulemaking discussions.
                The 2012 final rule, for example, stated that ``[t]he need of the
                nation to conserve energy has long operated to push the balancing
                toward more stringent standards,'' and asked ``[i]n this final rule,
                then, the question raised by this factor, combined with technological
                feasibility, becomes `how stringent can NHTSA set standards before
                economic practicability considerations intercede?' '' \2731\ The NPRM
                discussed the dictionary definition of ``to conserve,'' tentatively
                concluded that thousandths of a degree centigrade in 2100 did not rise
                to the level of being ``wasteful,'' and suggested that ultimately ``we
                no longer view the need of the U.S. to conserve energy as nearly
                infinite.'' \2732\ This is an evolution in interpretation that was
                expressly acknowledged in the NPRM--the words ``we no longer view''
                clearly indicate acknowledgement of a change in view, i.e.,
                interpretation. The NPRM's climate findings were not ignored, they were
                directly examined and discussed at 83 FR 43215-16 in the context of
                NHTSA's interpretation of their significance. The NPRM also discussed
                overall costs and benefits and net benefits in the context of the
                proposed maximum feasible determination, and the cost of carbon
                emissions was included in those values. This final rule similarly
                directly examines and discusses the analytical findings below.
                ---------------------------------------------------------------------------
                 \2731\ 77 FR at 63038-39.
                 \2732\ 83 FR at 43215-16.
                ---------------------------------------------------------------------------
                 Moreover, contrary to commenters' statements that NHTSA did not
                acknowledge that its interpretation of the effect of the ``need of the
                U.S. to conserve energy'' factor was changing, or that the balancing of
                factors was different, the NPRM directly stated that:
                 NHTSA well recognizes that the decision it proposes to make in
                today's NPRM is different from the one made in the 2012 final rule
                that established standards for MY 2021 and identified `augural'
                standard levels for MYs 2022-2025. Not only do we believe that the
                facts before us have changed, but we believe that those facts have
                changed sufficiently that the balancing of the EPCA factors and the
                other considerations must also change.
                The standards that we are proposing today reflect that
                balancing.\2733\
                ---------------------------------------------------------------------------
                 \2733\ 83 FR at 43213. See also 83 FR at 43226 (``In the 2012
                final rule . . . , NHTSA stated that `maximum feasible standards
                would be represented by the mpg levels that we could require of the
                industry before we reach a tipping point that presents risk of
                seriously adverse economic consequences.' [citation omitted]
                However, the context of that rulemaking was meaningfully different
                from the current context. At that time, NHTSA understood the need of
                the U.S. to conserve energy as necessarily pushing the agency toward
                setting stricter and stricter standards. Combining a then-paramount
                need of the U.S. to conserve energy with the perception that
                technological feasibility should no longer be seen as a limiting
                factor, NHTSA then concluded that only significant economic harm
                would be the basis for controlling the pace at which CAFE stringency
                increased over time. Today, the relative importance of the need of
                the U.S. to conserve energy has changed . . . a great deal even
                since the 2012 rulemaking. [T]he need of the U.S. to conserve energy
                may no longer disproportionately outweigh other statutorily-mandated
                considerations such as economic practicability--even when
                considering fuel savings from potentially more-stringent
                standards.'').
                [[Page 25145]]
                ---------------------------------------------------------------------------
                NHTSA believes that this is clear acknowledgement of the differences in
                interpretation and the effect of those differences on policy decisions.
                [[Page 25146]]
                 That said, NHTSA agrees (indeed, has always agreed) with commenters
                that environmental implications exist as a result of changes in CAFE
                stringency. While CO2 emissions will be higher under this
                final rule than if NHTSA had determined that the augural standards were
                maximum feasible, they will be lower than they would have been under
                the proposal--for the ``standard setting'' runs, which are what NHTSA
                looks at for assistance in determining maximum feasible standards,
                NHTSA estimates that, accounting for both tailpile and upstream
                emissions, CO2 emissions in 2050 under the final standards
                will total 1,134 mmt, as compared to 1,149 mmt under the proposed
                standards, or 1,020 mmt under the augural standards. According to the
                Final EIS, which uses a ``real-world'' analysis that incorporates
                models and modeling approaches that permit the agency to take a hard
                look at the potential environmental impacts of the rule,\2734\ NHTSA
                estimates that these amounts of CO2 emissions would lead to
                the following global temperature, sea level, and ocean acidification
                effects: \2735\
                ---------------------------------------------------------------------------
                 \2734\ See Kleppe v. Sierra Club, 427 U.S. 390, 410, n. 21
                (1976).
                 \2735\ As discussed in Section 5.3.1 of the FEIS, NHTSA used the
                Global Change Assessment Model (GCAM) Reference scenario to
                represent the No Action Alterantive (Alternative 0) in the modeling
                runs used to create Table I-1. The GCAM Reference Scenario is based
                on a set of assumptions about drivers such as population,
                technology, and socioeconomic changes, in the absence of global
                action to mitigate climate change. It can be described as a
                ``business-as-usual'' scenario. NHTSA also conducted an analysis in
                Chapter 8 of the FEIS using the GCAM6.0 scenario, which assumes a
                moderate level of global GHG reductions and corresponds to
                stabilization, by 2100, of total radiative forcing and associated
                CO2 concentrations at roughly 678 ppm. Several commenters
                argued that NHTSA presented climate results in the NPRM/DEIS in the
                context of a ``doomsday scenario,'' in which no actions at all are
                taken to mitigate carbon emissions, but NHTSA emphasizes that this
                is simply the GCAM Reference Scenario, which is a reasonable
                scenario to run given that GCAM is a widely accepted climate model.
                Running the analysis using the GCAM Reference Scenario and GCAM6.0
                Scenario results in different absolute values for the climate
                variables presented in this table and Table 8.6.4-1 of the FEIS, but
                again, this is because of the underlying scenarios, which reflect
                very different levels of global action. When the differences in
                levels of global action are accounted for, the relative impact of
                each action alternative as compared to the No Action Alternative is
                very similar. Thus, regardless of what GCAM scenario the agencies
                consider regarding global action to mitigate climate change, it is
                still meaningful to draw conclusions about the relative impacts of
                the alternatives, because the alternatives are what is within the
                agencies' authority to affect.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 25147]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.733
                [[Page 25148]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.734
                BILLING CODE 4910-59-C
                 NHTSA understands that some commenters view climate change as an
                imminent existential threat. NHTSA does not agree, however, that
                Congress
                [[Page 25149]]
                intended for NHTSA to set aside other statutory factors in determining
                what CAFE standards would be maximum feasible. Even the maximum
                feasible discussion for the 2012 final rule stated that
                 We recognize that higher standards would help the need of the
                nation to conserve more energy . . ., but based on our analysis and
                the evidence presented by the industry, we conclude that higher
                standards would not represent the proper balancing for MYs 2017-2025
                cars and trucks. [footnote omitted] We conclude that the correct
                balancing recognizes economic practicability concerns as discussed
                above, and sets standards at the levels that the agency is
                promulgating in this final rule for MYs 2017-2021 and presenting for
                MYs 2022-2025.\2736\
                ---------------------------------------------------------------------------
                 \2736\ 77 FR at 63055.
                The footnote following the last sentence quoted above further stated
                that ``We underscore that the agency's decision regarding what
                standards would be maximum feasible for MYs 2017-2025 is made with
                reference to the rulemaking time frame and the circumstances of this
                final rule. Each CAFE rulemaking (indeed, each stage of any given CAFE
                rulemaking) presents the agency with new information that may affect
                how the agencies we balance the relevant factors.'' \2737\ NHTSA has
                been consistent over time, despite commenters' suggestions to the
                contrary, that maximum feasible is a balancing of factors; that all
                factors must be considered; and that information before the agency may
                change how the agency both understands and balances the statutory
                factors.
                ---------------------------------------------------------------------------
                 \2737\ Id at fn. 1275.
                ---------------------------------------------------------------------------
                 With regard to criteria and toxic air pollutant emissions, NHTSA
                agrees with commenters that the NPRM discussion of environmental
                implications did not specifically identify these emissions, but notes
                that air quality issues were discussed in a variety of places in the
                NPRM, DEIS, and PRIA, and that the monetized effects of air quality
                impacts were included in the overall cost-benefit analysis which
                informed NHTSA's balancing of factors, as discussed above. To the
                extent that commenters disagreed with the values or the agency's air
                quality analyses, those topics will be addressed in Section VII and
                VIII and in the FEIS. NHTSA has considered all of these findings along
                with other factors, as discussed below.
                (d) Foreign Policy Implications
                 In the NPRM, NHTSA explained that U.S. consumption and imports of
                petroleum products impose costs on the domestic economy that are not
                reflected in the market price for crude petroleum or in the prices paid
                by consumers for petroleum products such as gasoline. These costs
                include (1) higher prices for petroleum products resulting from the
                effect of U.S. oil demand on world oil prices, (2) the risk of
                disruptions to the U.S. economy caused by sudden increases in the
                global price of oil and its resulting impact of fuel prices faced by
                U.S. consumers, and (3) expenses for maintaining the strategic
                petroleum reserve (SPR) to provide a response option should a
                disruption in commercial oil supplies threaten the U.S. economy, to
                allow the U.S. to meet part of its International Energy Agency
                obligation to maintain emergency oil stocks, and to provide a national
                defense fuel reserve.\2738\ Higher U.S. consumption of crude oil or
                refined petroleum products increases the magnitude of these external
                economic costs, thus increasing the true economic cost of supplying
                transportation fuels above the resource costs of producing them.
                Conversely, reducing U.S. consumption of crude oil or refined petroleum
                products (by reducing motor fuel use) can reduce these external costs.
                ---------------------------------------------------------------------------
                 \2738\ While the U.S. maintains a military presence in certain
                parts of the world to help secure global access to petroleum
                supplies, that is neither the primary nor the sole mission of U.S.
                forces overseas. Moreover, the scale of oil consumption reductions
                associated with CAFE standards would be insufficient to alter any
                existing military missions focused on ensuring the safe and
                expedient production and transportation of oil around the globe.
                Chapter 7 of the PRIA discussed this topic in more detail.
                ---------------------------------------------------------------------------
                 The NPRM stated that while these costs are considerations, the
                United States has significantly increased oil production capabilities
                in recent years to the extent that the U.S. is currently producing
                enough oil to satisfy nearly all of its energy needs and is projected
                to continue to do so or become a net energy exporter. This has added
                new stable supply to the global oil market and reduced the urgency of
                the U.S. to conserve energy. The NPRM referred readers to the balancing
                discussion for more detail on this issue.
                 Securing America's Energy Future commented that continuing to raise
                stringency would be good for energy security, spur innovation, and
                ``advance the administration's energy dominance agenda.'' \2739\ CARB
                argued that the proposal would ``significantly diminish U.S. energy
                security,'' ``. . . contrary to the President's recent executive order
                to promote national security, and contrary to the intent of Congress in
                EPCA.'' \2740\
                ---------------------------------------------------------------------------
                 \2739\ Securing America's Energy Future, NHTSA-2018-0067-12172,
                at 6.
                 \2740\ CARB, NHTSA-2018-0067-11783, at 316.
                ---------------------------------------------------------------------------
                 Several commenters disagreed with the NPRM's suggestion that
                increases in U.S. oil production reduced the foreign policy
                implications relevant to the need of the U.S. to conserve energy. ACEEE
                commented that because the market for oil is global, ``. . . regardless
                of actual imports, the nation is still affected by what happens to oil
                worldwide, and oil remains a foreign policy concern . . . .'' \2741\
                Securing America's Energy Future commented that increased U.S.
                production ``. . . has reduced some of the negative consequences of oil
                dependence, energy security is primarily a function of consumption, not
                production.'' \2742\ IPI argued that ``. . . the agencies falsely and
                inconsistently argue that the need to conserve energy has diminished
                because U.S. reliance on foreign oil has decreased,'' disagreeing with
                the NPRM's assumption that monopsony and military security costs
                resulting from the proposal would be zero.\2743\ The States and Cities
                commenters raised similar points, stating that ``U.S. military and
                foreign policy institutes'' place emphasis on ``global oil market
                stability and the stability of major oil-exporting nations,'' which the
                States and Cities argued had not changed as
                [[Page 25150]]
                U.S. exports have risen.\2744\ The States and Cities commenters further
                argued that if a quarter of U.S. oil consumed is still imported, then
                increases in consumption would necessarily raise imports, and thus also
                monopsony and military security costs associated with those
                imports.\2745\
                ---------------------------------------------------------------------------
                 \2741\ ACEEE, NHTSA-2018-0067-12122, main comments, at 2.
                 \2742\ Securing America's Energy Future, NHTSA-2018-0067-12172,
                at 6.
                 \2743\ IPI, NHTSA-2018-0067-12213, Appendix, at 2-3.
                 \2744\ States and Cities, NHTSA-2018-0067-11735, at 76-77.
                 \2745\ Id.
                ---------------------------------------------------------------------------
                 CARB questioned whether it was accurate to assume that the U.S.
                would ever reach net exporter status, and commented that even if
                becoming a net exporter helped to insulate the Nation from the effects
                of reducing CAFE stringency, it would not lead to greater energy
                security until at least 2029, the first year for which AEO 2018
                forecasts that the U.S. will stop being a net importer.\2746\ CARB
                further argued that increased domestic oil production did not insulate
                the U.S. from risk, and that in fact ``. . . current conditions are
                more prone to risk due to lower available spare oil production capacity
                in major oil producing countries, meaning that a supply disruption is
                more likely to have a more pronounced effect on oil prices and U.S.
                energy security.'' \2747\
                ---------------------------------------------------------------------------
                 \2746\ CARB, NHTSA-2018-0067-11783, at 317.
                 \2747\ Id., at 319.
                ---------------------------------------------------------------------------
                 Mr. Bordoff commented that geopolitical risk can still affect
                global oil prices, citing U.S. withdrawal from the Iran nuclear
                agreement and the reimposition of sanctions on Iranian oil sales; the
                collapse of Libyan oil production following conflict there; ongoing
                problems in Venezuela; a variety of short-term production outages in
                other producing areas; and even situations where geopolitics can result
                in lower prices rather than higher prices.\2748\
                ---------------------------------------------------------------------------
                 \2748\ Bordoff, EPA-HQ-OAR-2018-0283-3906, at 3-4.
                ---------------------------------------------------------------------------
                 IPI stated that ``. . . the protective value that the SPR offers
                given its size does automatically change as total U.S. petroleum
                consumption changes,'' and argued that it was not sufficient to
                consider only ``the budgetary costs for maintaining [the size of] the
                SPR.'' IPI thus argued that ``The agencies have failed to assess how
                much the relative protective value of the SPR will change as total U.S.
                consumption rises following the proposed rollback, and therefore have
                failed entirely to consider one important element of the national need
                to conserve energy.'' \2749\
                ---------------------------------------------------------------------------
                 \2749\ IPI, NHTSA-2018-0067-12213, Appendix, at 4.
                ---------------------------------------------------------------------------
                 Total energy independence for any country is only possible if it
                does not participate in the global energy markets, either because it
                consumes no energy (which is unrealistic) or because it produces enough
                energy to meet all of its energy needs and uses only energy that is
                produced domestically. As discussed above, NHTSA agrees with commenters
                that the oil market is global, and that events and situations abroad
                can affect oil prices even as U.S. oil production increases. The fact
                that the U.S. became a net oil exporter, at least on a weekly basis, in
                November 2019, and the evidence indicates that it will become a net oil
                exporter on a longer-term basis in MY 2020 does not change geopolitics
                in many parts of the world. Striving for energy independence in a
                global market necessarily means reducing risks, because even if the
                U.S. consumed only domestically-produced petroleum and continued to
                export, the U.S. economy would still be subject to oil price
                fluctuations due to external events and situations. The NPRM was clear
                on all of these points.\2750\ The NPRM and PRIA repeatedly emphasized
                that changes in the oil market meant that the risk of damage to the
                U.S. economy and of additional pain for U.S. drivers is lower than it
                was at the beginning of the CAFE program, not that it was eliminated
                entirely. NHTSA agrees with commenters that risk still exists, and that
                both production and consumption of oil are relevant to how big that
                risk might be. NHTSA simply believes, as explained in the NPRM and as
                explained again below, that the risk is lower than it would have been
                in the absence of the rapid growth in U.S. oil production, and that the
                lower risk means that the need of the U.S. to conserve energy, from
                this perspective, is less dire than it was at earlier points in the
                program.
                ---------------------------------------------------------------------------
                 \2750\ See 83 FR at 43213-15.
                ---------------------------------------------------------------------------
                 The analyses for both the NPRM and the final rule account for the
                ongoing economic risk of participating in the global oil market by
                placing a value on energy security. The energy security value is made
                of several components. While commenters are correct that neither the
                NPRM nor the final rule analyses attributed a positive cost to the
                monopsony or military security components, the agencies do employ a
                cost for macroeconomic shock risk as part of energy security. Section
                VI discusses these estimates in more detail; for purposes of this
                discussion, NHTSA only notes that these issues are accounted for in the
                agencies' cost-benefit analysis, and to the extent that zero values are
                used for some elements, the reason for that is explained at length in
                those sections and public comments received on these issues did not
                present new information to change the agencies' minds on those values.
                 With regard to the comment that NHTSA should be accounting for the
                ``protective value'' of the SPR along with the literal cost of
                maintaining it, NHTSA is not in a position at this time to attempt to
                estimate such a value, and notes that the commenter provided no
                suggestions as to how to do so. The Department of Energy's website
                states that the maximum number of days of import protection provided by
                the SPR is 143 days, and that it takes 13 days from Presidential
                decision for SPR fuel to enter the market.\2751\ The 1973 OPEC oil
                embargo lasted from October 1973 to March 1974, roughly 150 days. As
                explained, NHTSA continues to believe that the effect of increased U.S.
                oil production is to stabilize, broadly, global oil markets. The longer
                a sustained spike in prices due to geopolitical events continues, the
                greater incentive U.S. shale production has to respond. NHTSA believes
                that it is foreseeable that the SPR could be utilized to help mitigate
                a price shock in the interim, for the majority of foreseeable shock
                situations.
                ---------------------------------------------------------------------------
                 \2751\ See https://www.energy.gov/fe/services/petroleum-reserves/strategic-petroleum-reserve/spr-quick-facts-and-faqs.
                ---------------------------------------------------------------------------
                (5) Factors That NHTSA Is Prohibited From Considering
                 The NPRM explained that EPCA also provides that in determining the
                level at which it should set CAFE standards for a particular model
                year, NHTSA may not consider the ability of manufacturers to take
                advantage of several EPCA provisions that facilitate compliance with
                CAFE standards and thereby reduce the costs of compliance.\2752\ As
                discussed further in Section IX below, NHTSA cannot consider compliance
                credits that manufacturers earn by exceeding the CAFE standards and
                then use to achieve compliance in years in which their measured average
                fuel economy falls below the standards. NHTSA also cannot consider the
                use of alternative fuels by dual fuel vehicles nor the availability of
                dedicated alternative fuel vehicles--including battery-electric
                vehicles--in any model year. EPCA encourages the production of
                alternative fuel vehicles by specifying that their fuel economy is to
                be determined using a special calculation procedure that results in
                those vehicles being assigned a higher equivalent fuel economy level
                than they actually achieve.
                ---------------------------------------------------------------------------
                 \2752\ 49 U.S.C. 32902(h).
                ---------------------------------------------------------------------------
                 The NPRM further explained that the effect of the prohibitions
                against
                [[Page 25151]]
                considering these statutory flexibilities in setting the CAFE standards
                is that the flexibilities remain voluntarily-employed measures. If
                NHTSA were instead to assume manufacturer use of those flexibilities in
                setting new standards, higher standards would appear less costly and
                therefore more feasible, which would thus effectively require
                manufacturers to use those flexibilities in order to meet higher
                standards. By keeping NHTSA from including them in our stringency
                determination, the provision ensures that these statutory credits
                remain true compliance flexibilities.
                 Additionally, for the non-statutory fuel economy improvement value
                program that NHTSA developed by regulation, the NPRM stated that NHTSA
                does not consider these subject to the EPCA prohibition on considering
                flexibilities. EPCA is very clear as to which flexibilities are not to
                be considered. When the agency has introduced additional flexibilities
                such as A/C efficiency and ``off-cycle'' technology fuel economy
                improvement values, NHTSA has considered those technologies as
                available in the analysis. Thus, today's analysis includes assumptions
                about manufacturers' use of those technologies, as detailed in Section
                VI.
                 Michalek and Whitefoot commented that ``[w]e find [the statutory
                prohibition on considering certain flexibilities in determining maximum
                feasible CAFE standards] problematic because the automakers use these
                flexibilities as a common means of complying with the regulation, and
                ignoring them will bias the cost-benefit analysis to overestimate
                costs.'' \2753\ IPI commented that ``it is not clear that the statutory
                prohibition on considering credit availability was intended to apply to
                banked credits,'' because 49 U.S.C. 32902(h)(3) was
                ---------------------------------------------------------------------------
                 \2753\ Michalek and Whitefoot, NHTSA-2018-0067-11903, at 10-11.
                added . . . as a `conforming amendment' to EISA, which was the
                statute that gave NHTSA authority to allow credit trading and
                transferring; meanwhile, banking and borrowing have been part of
                NHTSA's authority since EPCA in 1975. In 1989, e.g., NHTSA
                explicitly relied on the availability of `credit banks' to justify
                maintaining the MY 1990 standard at 27.5 mpg instead of lowering its
                stringency. NHTSA has not explained why it now believes it may not
                more fully consider banking.\2754\
                ---------------------------------------------------------------------------
                 \2754\ IPI, NHTSA-2018-0067-12213, Appendix, at 19.
                 NHTSA agrees, as explained in the NPRM, that if the agency was able
                to consider the compliance flexibilities in determining maximum
                feasible standards, more-stringent standards would appear less costly
                and therefore more feasible. NHTSA is nevertheless bound by the
                statutory prohibition on considering the above-mentioned flexibilities.
                As for IPI's disagreement that 32902(h)(3) should apply to banked
                credits because it was labeled a ``conforming amendment,'' NHTSA looks
                to the specific statutory language provided, which prohibits
                ``[consideration], when prescribing a fuel economy standard, [of] the
                trading, transferring or availability of credits . . . .'' (Emphasis
                added.) IPI's suggested interpretation would render ``availability'' as
                surplusage. If Congress had meant the prohibition to apply only to
                traded and transferred credits, it would have said so. Instead,
                Congress also prohibited consideration of the ``availability of
                credits,'' which must be read reasonably to refer to ``what credits are
                available,'' i.e., banked credits. The fact that NHTSA considered the
                availability of banked credits in 1989, prior to establishment of this
                statutory prohibition, has no bearing in a post-EISA world.
                 Nonetheless, NHTSA notes that it is informed by the ``real-world''
                analysis presented in the FRIA, which accounts for credit availability
                and usage, and manufacturers' ability to employ alternative fueled
                vehicles--for purpose of conformance with E.O. 12866. Under the real-
                world analysis, compliance does, in fact, appear less costly. For
                example, today's ``real world'' analysis shows manufacturers' costs
                averaging about $1,420 in MY 2029 under the final standards, as
                compared to the $1,640 shown by the ``standard setting'' analysis.
                However, for purposes of determining maximum feasible CAFE levels,
                NHTSA considers only the ``standard-setting'' analysis shown in the
                NPRM, consistent with Congress's direction.
                (f) EPCA/EISA Requirements That No Longer Apply Post-2020
                 The NPRM explained that Congress amended EPCA through EISA to add
                two requirements not yet discussed in this section relevant to
                determination of CAFE standards during the years between MY 2011 and MY
                2020 but not beyond. First, Congress stated that, regardless of NHTSA's
                determination of what levels of standards would be maximum feasible,
                standards must be set at levels high enough to ensure that the combined
                U.S. passenger car and light truck fleet achieves an average fuel
                economy level of not less than 35 mpg no later than MY 2020.\2755\ And
                second, between MYs 2011 and 2020, the standards must ``increase
                ratably'' in each model year.\2756\ Neither of these requirements apply
                after MY 2020, so given that this rulemaking concerns the standards for
                MY 2021 and after, the NPRM stated that they are not relevant to this
                rulemaking.
                ---------------------------------------------------------------------------
                 \2755\ 49 U.S.C. 32902(b)(2)(A).
                 \2756\ 49 U.S.C. 32902(b)(2)(C).
                ---------------------------------------------------------------------------
                 CARB commented that because the proposal did not ``provide for
                improved efficiency of motor vehicles'' over the long term,
                ``Stagnating the standards violates Congressional direction to ratably
                increase fuel economy when the technology for doing so has been
                demonstrated to exist (which it does . . .) or could be developed in
                the necessary time.'' \2757\
                ---------------------------------------------------------------------------
                 \2757\ CARB, NHTSA-2018-0067-11873, Detailed Comments, at 84.
                ---------------------------------------------------------------------------
                 NHTSA notes, again, that the statutory language is clear that
                Congress only directed ratable increases in stringency through MY 2020.
                After MY 2020, the statutory language is clear that standards simply
                need be ``maximum feasible, as determined by the Secretary.'' Some
                commenters may have disagreed that the proposal represented maximum
                feasible levels, but there is no statutory basis for arguing that the
                ``ratable increase'' requirement extends beyond MY 2020.
                (g) Other Considerations in Determining Maximum Feasible Standards
                 The NPRM explained that NHTSA has historically considered the
                potential for adverse safety consequences in setting CAFE standards.
                This practice has been consistently approved in case law. As courts
                have recognized, ``NHTSA has always examined the safety consequences of
                the CAFE standards in its overall consideration of relevant factors
                since its earliest rulemaking under the CAFE program.'' Competitive
                Enterprise Institute v. NHTSA, 901 F.2d 107, 120 n. 11 (D.C. Cir. 1990)
                (``CEI-I'') (citing 42 FR 33534, 33551 (June 30, 1977)). The courts
                have consistently upheld NHTSA's implementation of EPCA in this manner.
                See, e.g., Competitive Enterprise Institute v. NHTSA, 956 F.2d 321, 322
                (D.C. Cir. 1992) (``CEI-II'') (in determining the maximum feasible fuel
                economy standard, ``NHTSA has always taken passenger safety into
                account'') (citing CEI-I, 901 F.2d at 120 n. 11); Competitive
                Enterprise Institute v. NHTSA, 45 F.3d 481, 482-83 (D.C. Cir. 1995)
                (``CEI-III'') (same); Center for Biological Diversity v. NHTSA, 538
                F.3d 1172, 1203-04 (9th Cir. 2008) (upholding NHTSA's analysis of
                vehicle safety issues associated with weight in connection with the MYs
                2008-2011
                [[Page 25152]]
                light truck CAFE rulemaking). Thus, NHTSA explained that in evaluating
                what levels of stringency would result in maximum feasible standards,
                NHTSA assesses the potential safety impacts and considers them in
                balancing the statutory considerations and to determine the maximum
                feasible level of the standards.
                 The attribute-based standards that Congress requires NHTSA to set
                help to mitigate the negative safety effects of the historical single
                number standards originally required in EPCA, and in past rulemakings,
                NHTSA constrained its modeling so as not to consider possible mass
                reduction in lower weight vehicles in its analysis, which affected the
                resulting assessment of potential adverse safety impacts. That
                analytical approach did not reflect, however, the likelihood that
                automakers may pursue the most cost-effective means of improving fuel
                efficiency to comply with CAFE requirements. For the NPRM, as for the
                final rule, the modeling did not limit the amount of mass reduction
                that is applied to any segment, but rather considered that automakers
                may apply mass reduction based upon cost-effectiveness, similar to most
                other technologies. NHTSA does not, of course, mandate the use of any
                particular technology by manufacturers in meeting the standards. The
                NPRM and today's final rule, like the Draft TAR, also considered the
                safety effect associated with the additional vehicle miles traveled due
                to the rebound effect.
                 NHTSA explained that the NPRM considered the safety effects of
                vehicle scrappage rates on the fleet as a whole. The NPRM also
                explained NHTSA's consideration of the effect of additional expenses in
                fuel savings technology on the affordability of vehicles--the
                likelihood that increased standards will result in consumers being
                priced out of the new vehicle market and choosing to keep their
                existing vehicle or purchase a used vehicle. Since new vehicles are
                significantly safer than used vehicles, slowing fleet turnover to newer
                vehicles results in older and less safe vehicles remaining on the roads
                longer. NHTSA stated that this significantly affects the safety of the
                United States light duty fleet, as described more fully in in the
                safety section of the NPRM and in Chapter 11 of the PRIA. Furthermore,
                as fuel economy standards become more stringent, and more fuel
                efficient vehicles are introduced into the fleet, fueling costs are
                reduced. This results in consumers driving more miles, which results in
                more crashes and increased highway fatalities.
                 A number of commenters disagreed with a variety of aspects of the
                NPRM's analysis of safety, and several also disagreed with how NHTSA
                considered safety along with the other factors in the proposal. The
                States and Cities commenters, for example, agreed that ``NHTSA has
                historically considered safety impacts when setting maximum feasible
                standards,'' but argued that:
                in the Proposed Rollback, NHTSA departs from its past practice by
                relying on completely novel and unsupported theories regarding the
                linkages between fuel economy and safety that do not reflect
                reality. In the past, NHTSA has considered the safety of the
                technologies that improve fuel economy. [citations omitted] In the
                Proposed Rollback, however, NHTSA has linked safety concerns with
                rebound and scrappage effects of more stringent fuel economy
                standards. [citations omitted] As discussed [elsewhere], these
                theories are unsupported, implausible, and contradicted by numerous
                experts--rendering them arbitrary and capricious. The agency has
                also failed to acknowledge or adequately justify its break with past
                analyses of safety. See Fox Television, 556 U.S. at 515.'' \2758\
                ---------------------------------------------------------------------------
                 \2758\ States and Cities, NHTSA-2018-0067-11735, at 77.
                 EDF commented that NHTSA cannot ``. . . lawfully rely upon the
                repercussions of increased driving as a justification. . . . The fact
                that the standards do not `compel' this driving prevents such reliance,
                and . . . [EPCA/EISA] nowhere indicate that [NHTSA] can refuse to
                comply with [its] statutory obligations by pointing to a projection
                that individuals might drive more and in doing so, some of them will
                get into traffic accidents.\2759\ EDF further argued that:
                ---------------------------------------------------------------------------
                 \2759\ EDF, NHTSA-2018-0067-12137, Supplemental Safety Comments,
                at 3.
                 It is especially unlikely that Congress intended for NHTSA to
                consider potential increases in driving (or . . . `VMT'). Under
                basic economic theory and under the Agency's traditional analysis
                (including their analysis of this proposal), an improvement in fuel
                economy--which makes driving cheaper--would be expected to lead to
                some increase in driving for households that are sensitive to and
                conscious of that effect on their budgets. Thus, consideration of
                VMT impacts could be used to undermine any fuel economy standard.
                Because VMT is `a factor [that] is both so indirectly related to
                [fuel economy] and so full of potential for canceling the
                conclusions drawn from [a fuel economy analysis] . . . it would
                surely have been expressly mentioned in [the statute] had Congress
                meant it to be considered.' Whitman v. Am. Trucking Associations,
                531 U.S. 457, 469 (2001).'' \2760\
                ---------------------------------------------------------------------------
                 \2760\ Id.
                 Other comments on safety as part of the legal justification varied.
                NESCAUM claimed that NHTSA's safety justification ``is disputed by
                EPA's technical staff based on their identification of flaws in NHTSA's
                analysis,'' suggesting that it was therefore invalid and not a basis
                for decision-making.\2761\ Global commented that there was no policy
                reason for freezing the level of standards due to mass reduction
                concerns (i.e., safety), given footprint standards.\2762\ IPI argued
                that it was inappropriate to account for vehicle safety-related deaths
                and injuries ``without an adequate discussion of the health and safety
                impacts of the Proposed Rule's increased emissions or without an
                accurate estimate of the actual safety impact of the rollback versus
                the 2012 standards.'' \2763\
                ---------------------------------------------------------------------------
                 \2761\ NESCAUM, NHTSA-2018-0067-11691, at 3.
                 \2762\ Global, NHTSA-2018-0067-12032, Attachment A, at A-32.
                 \2763\ IPI, NHTSA-2018-0067-12213, Appendix, at 11.
                ---------------------------------------------------------------------------
                 NHTSA agrees with commenters that the safety analysis conducted to
                inform this rulemaking (both NPRM and final rule) is different from--
                broader than--past safety analyses conducted to inform CAFE and
                CO2 rulemakings. NHTSA disagrees, however, that the agency
                failed to acknowledge or explain this fact. The NPRM directly
                acknowledges and explains the evolution of the safety analysis over
                time and why, specifically, the NPRM included the safety effects of
                rebound and scrappage phenomena.\2764\ The NPRM also expressly sought
                comment on these elements of the safety analysis and the safety
                analysis generally, before explaining how they worked and describing
                their tentative findings in considerable detail. It is inaccurate for
                commenters to claim that the agency did not acknowledge or explain
                these changes. Commenters' disagreement with the substance of the
                safety analysis does not create a valid process complaint here. Section
                VI discusses in detail the comments received on the substance of the
                safety analysis, including a number of comments citing deliberative
                feedback provided by some members of EPA staff during NPRM development,
                and contains the agencies' responses. With regard to the comment from
                EDF, as explained above, the premise that vehicles may be driven more
                or less in response to more or less stringent CAFE (or CO2)
                standards is called the rebound effect, and it is discussed at length
                in Section VI above. The rebound effect has been factored into
                rulemaking cost-benefit analyses and reduced CAFE and CO2
                standard benefits in such analyses for well over a decade,\2765\ and
                EPA and NHTSA have
                [[Page 25153]]
                written repeatedly about and considered the magnitude of this effect.
                NHTSA is aware that some commenters disagree that a rebound effect even
                exists for fuel economy, and understands how such commenters would
                correspondingly disagree that VMT-related safety effects could arise
                from differences in CAFE standards. But NHTSA does not agree that the
                rebound effect is zero, and correspondingly believes that safety
                effects from additional driving (due to exposure to crashes) exist and
                are capable of quantification for analytical purposes.
                ---------------------------------------------------------------------------
                 \2764\ See 83 FR at 43106-07.
                 \2765\ See, e.g., 68 FR 16868, 16878 (Apr. 7, 2003).
                ---------------------------------------------------------------------------
                 Moreover, if EDF were correct that agencies may consider only the
                behavior that regulations directly ``compel,'' then CAFE analysis would
                be challenged to consider even fuel savings--the purpose of CAFE
                standards--because the standards do not compel Americans to drive, or
                to buy new vehicles, or to buy any vehicles at all. Reasonable
                assumptions about how much Americans drive (depending on how much it
                costs to drive, among other things), and what vehicles Americans buy
                and how often they buy them (depending on how much those vehicles cost,
                among other things), are useful and important for including in analyses
                that help decision-makers distinguish between different levels of
                potential CAFE standards. Circular A-4 additionally directs agencies to
                consider ancillary effects of rulemakings.\2766\ NHTSA believes that it
                is reasonable to consider these effects as part of the safety analysis,
                and to consider safety effects as part of its determination of maximum
                feasible standards.
                ---------------------------------------------------------------------------
                 \2766\ See OIRA, ``Regulatory Impact Analysis: A Primer,'' at 7,
                https://www.reginfo.gov/public/jsp/Utilities/circular-a-4_regulatory-impact-analysis-a-primer.pdf (``In addition to the
                direct benefits and costs of each alternative, the list should
                include any important ancillary benefits and countervailing risks.
                An ancillary benefit is a favorable impact of the alternative under
                consideration that is typically unrelated or secondary to the
                purpose of the action (e.g., reduced refinery emissions due to more
                stringent fuel economy standards for light trucks). A countervailing
                risk is an adverse economic, health, safety, or environmental
                consequence that results from a regulatory action and is not already
                accounted for in the direct cost of the action (e.g., adverse safety
                impacts from more stringent fuel-economy standards for light
                trucks). As with other benefits and costs, an effort should be made
                to quantify and monetize both ancillary benefits and countervailing
                risks.'')
                ---------------------------------------------------------------------------
                (2) Administrative Procedure Act
                 To be upheld under the ``arbitrary and capricious'' standard of
                judicial review in the APA, an agency rule must be rational, based on
                consideration of the relevant factors, and within the scope of the
                authority delegated to the agency by the statute. The agency must
                examine the relevant data and articulate a satisfactory explanation for
                its action including a ``rational connection between the facts found
                and the choice made.'' \2767\
                ---------------------------------------------------------------------------
                 \2767\ Burlington Truck Lines, Inc., v. United States, 371 U.S.
                156, 168 (1962).
                ---------------------------------------------------------------------------
                 Statutory interpretations included in an agency's rule are subject
                to the two-step analysis of Chevron, U.S.A. v. Natural Resources
                Defense Council.\2768\ Under step one, where a statute ``has directly
                spoken to the precise question at issue,'' id. at 842, the court and
                the agency ``must give effect to the unambiguously expressed intent of
                Congress.'' \2769\ If the statute is silent or ambiguous regarding the
                specific question, the court proceeds to step two and asks ``whether
                the agency's answer is based on a permissible construction of the
                statute.'' \2770\
                ---------------------------------------------------------------------------
                 \2768\ 467 U.S. 837 (1984).
                 \2769\ Id. at 843.
                 \2770\ Id.
                ---------------------------------------------------------------------------
                 If an agency's interpretation differs from the one that it has
                previously adopted, the agency need not demonstrate that the prior
                position was wrong or even less desirable. Rather, the agency would
                need only to demonstrate that its new position is consistent with the
                statute and supported by the record and acknowledge that this is a
                departure from past positions. The Supreme Court emphasized this in FCC
                v. Fox Television.\2771\ When an agency changes course from earlier
                regulations, ``the requirement that an agency provide a reasoned
                explanation for its action would ordinarily demand that it display
                awareness that it is changing position,'' but ``need not demonstrate to
                a court's satisfaction that the reasons for the new policy are better
                than the reasons for the old one; it suffices that the new policy is
                permissible under the statute, that there are good reasons for it, and
                that the agency believes it to be better, which the conscious change of
                course adequately indicates.'' \2772\ The APA also requires that
                agencies provide notice and comment to the public when proposing
                regulations,\2773\ as the agencies did when publishing the NPRM for
                this rulemaking.
                ---------------------------------------------------------------------------
                 \2771\ 556 U.S. 502 (2009).
                 \2772\ Id., at 1181.
                 \2773\ 5 U.S.C. 553.
                ---------------------------------------------------------------------------
                a) Requests To Extend the Comment Period
                 On August 2, 2018, the agencies published the NPRM on the agencies'
                respective websites, soliciting public comments.\2774\ On August 24,
                2018, the Federal Register published the NPRM, which began a 60-day
                public comment period.\2775\ The public comment period would have ended
                on October 23, 2018, but the agencies extended the comment period until
                October 26, 2018.\2776\ In the Federal Register notice extending the
                comment period, the agencies explained that they were denying requests
                for an extension of the comment period by at least 60 days, explaining
                that ``[a]utomakers will need maximum lead time to respond to the final
                rule[.]'' \2777\ Although the comment period ultimately closed on
                October 26, 2018, the agencies' dockets remained open, and the agencies
                continued to accept and consider comments, to the extent possible, for
                more than one year after the comment period began.\2778\
                ---------------------------------------------------------------------------
                 \2774\ https://www.nhtsa.gov/corporate-average-fuel-economy/safe; https://www.epa.gov/newsreleases/us-epa-and-dot-propose-fuel-economy-standards-my-2021-2026-vehicles.
                 \2775\ 83 FR 42986 (Aug. 24, 2018).
                 \2776\ See 83 FR 48578 (Sept. 26, 2018) (extending comment
                period).
                 \2777\ Id.
                 \2778\ The agencies notified the public of this possibility in
                the NPRM, stating that: ``To the extent practicable, we will also
                consider comments received after'' the close of the comment period.
                83 FR 42986, 43471 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 After publishing the NPRM, the agencies received a number of
                requests to extend the comment period, generally for an additional 60
                days.\2779\ For example, seventeen States and the District of Columbia
                jointly requested a 60-day extension of the comment period.\2780\ That
                request cited the voluminous record, the complexity of the material,
                and the profound potential impact on human health and the environment,
                among other things.\2781\ The City of Los Angeles and New York State
                Department of Environmental Conservation also requested a 60-day
                extension, for similar reasons.\2782\ In addition, 32 United States
                Senators jointly requested a 60-day extension of the comment
                period.\2783\ The Senators argued that an extension was appropriate to
                ensure adequate public participation with such an important rule.\2784\
                Several non-government organizations similarly requested a 60-day
                extension of the comment period due to the complexity of the issues and
                [[Page 25154]]
                the importance of the proposed rule.\2785\ Other organizations also
                requested a 60-day extension, stressing the complexity of the issues
                and the significance of the proposed rule's impact on the
                environment.\2786\ The American Lung Association also requested a 60-
                day extension of the comment period, asserting that it needed more time
                to analyze the impact of the proposed rule on human health.\2787\ The
                California Air Resources Board (CARB) likewise requested a 60-day
                extension, in part, based on information that it asserted should have
                been included in the NPRM.\2788\ New York University School of Law's
                Institute for Policy Integrity similarly requested a 60-day extension
                based on information that it contended should have been included in the
                NPRM's ``sensitivity analysis table for the `Cumulative Changes in
                Fleet Size, Travel (VMT), Fatalities, Fuel Consumption and C02
                Emissions through MY2029.' '' \2789\
                ---------------------------------------------------------------------------
                 \2779\ See 83 FR 48578 (Sept. 26, 2018).
                 \2780\ See comments from the State of California et al., Request
                for an extension, Docket No. NHTSA-2018-0067-3458.
                 \2781\ See id.
                 \2782\ Also for similar reasons, the Minnesota Pollution Control
                Agency and the Minnesota Department of Transportation submitted a
                joint request for a 120-day extension of the comment period. See
                comments from the Minnesota Pollution Control Agency and Minnesota
                Department of Transportation, Docket No. NHTSA-2018-0067-3580.
                 \2783\ See comments from 32 U.S. Senators (Kamala D. Harris et
                al.), Docket No. NHTSA-2018-0067-5643.
                 \2784\ See id.
                 \2785\ See, e.g., comments from the Alliance of Automobile
                Manufacturers, Docket No. NHTSA-2018-0067-3619; Communities for a
                Better Environment, Docket No. EPA-HQ-OAR-2018-0283-1095; Consumer
                Federation of America, NHTSA-2018-0067-3400; Edison Electric
                Institute, received by mail; and South Coast Air Quality Management
                District, Docket No. EPA-HQ-OAR-2018-0283-0885.
                 \2786\ See, e.g., comments from the Environmental Law and Policy
                Center, NHTSA-2018-0067-2728; Georgetown Climate Center, Docket No.
                NHTSA-2018-0067-3610; Center for Biological Diversity, Conservation
                Law Foundation, Earthjustice, Environmental Defense Fund, Natural
                Resources Defense Council, Public Citizen,
                 Sierra Club, and Union of Concerned Scientists, Docket No.
                NHTSA-2018-0067-3278; and National Governors Association, Docket No.
                EPA-HQ-OAR-2018-0283-0871.
                 \2787\ See comments from American Lung Association, Docket No.
                NHTSA-2018-0067-3615.
                 \2788\ See comments from California Air Resources Board, Docket
                No. NHTSA-2018-0067-4166.
                 \2789\ See comments from New York University School of Law's
                Institute for Policy Integrity, NHTSA-2018-0067-5641.
                ---------------------------------------------------------------------------
                 The agencies do not believe a further extension of the comment
                period was warranted under the circumstances.\2790\ The APA does not
                specify a minimum number of days for a comment period.\2791\ Two
                Executive Orders also provide direction to Federal agencies with
                respect to the length of a comment period for a proposed rule.\2792\
                Executive Order 12,866 states that ``[e]ach agency shall (consistent
                with its own rules, regulations, or procedures) provide the public with
                meaningful participation in the regulatory process . . . . In addition,
                each agency should afford the public a meaningful opportunity to
                comment on any proposed regulation, which in most cases should include
                a comment period of not less than 60 days.'' \2793\ Additionally,
                Executive Order 13,563 reaffirmed Executive Order 12,866's directive
                that comment periods should generally not be less than 60 days,
                stating: ``To the extent feasible and permitted by law, each agency
                shall afford the public a meaningful opportunity to comment through the
                internet on any proposed regulation, with a comment period that should
                generally be at least 60 days.'' \2794\ More recently, in December of
                2018, the Department of Transportation implemented DOT Order 2100.6,
                which provides its operating administrations, including NHTSA, with
                direction on appropriate rulemaking processes and procedures.\2795\
                While not yet effective at the time the proposal was published, the
                Order provides that ``the comment period for significant DOT rules
                should be at least 45 days.'' \2796\ The 63 day comment period for the
                proposal far exceeded this amount.
                ---------------------------------------------------------------------------
                 \2790\ See 83 FR 48578 (Sept. 26, 2018) (extending comment
                period until October 26, 2018 and denying requests for longer
                extensions).
                 \2791\ See 5 U.S.C. 553(c).
                 \2792\ The Executive Orders do not create any enforceable right
                or benefit by a party against any federal agency. E.O. 12,866 Sec.
                10; E.O. 13,563 Sec. 7(d).
                 \2793\ Executive Order 12,866 Sec. 6(a)(1).
                 \2794\ Executive Order 13,563 Sec. 2(b).
                 \2795\ DOT Order 2100.6, ``Policies and Procedures for
                Rulemakings,'' available at: https://www.transportation.gov/sites/dot.gov/files/docs/regulations/328561/dot-order-21006-rulemaking-process-signed-122018.pdf.
                 \2796\ Id., at (11)(i)(3).
                ---------------------------------------------------------------------------
                 Consistent with these principles, courts give broad discretion to
                agencies in determining the reasonableness of a comment period. Courts
                have frequently upheld comment periods that were significantly less
                than the 63-day comment period here. See Connecticut Light & Power Co.
                v. Nuclear Regulatory Comm'n, 673 F.2d 525, 534 (D.C. Cir. 1982)
                (upholding a 30-day comment period and stating that ``neither statute
                nor regulation mandates that the agency do more''); see also North
                American Van Lines v. ICC, 666 F.2d 1087, 1092 (7th Cir. 1981)
                (upholding a 45-day comment period).\2797\ In addition to the length of
                a comment period, courts consider the number of comments received and
                whether comments had an effect on an agency's final rule, in assessing
                whether the public had a meaningful opportunity to comment.\2798\
                ---------------------------------------------------------------------------
                 \2797\ In certain circumstances, particularly urgent ones,
                courts have even upheld comment periods of less than 30 days. See
                Omnipoint Corp. v. FCC, 78 F.3d 620, 629-30 (D.C. Cir. 1996)
                (holding that a 14-day comment period was sufficient given the
                ``urgent necessity for rapid administrative action under the
                circumstances''); see also Fla. Power & Light Co. v. United States,
                846 F.2d 765, 772 (D.C. Cir. 1988) (upholding a 15-day comment
                period given a deadline that Congress imposed on the Nuclear
                Regulatory Commission to finalize its rule).
                 \2798\ See Florida Power & Light, Co. v. United States, 846 F.2d
                765, 772 (D.C. Cir. 1988); see also Conference of State Bank Sup'rs
                v. Office of Thrift Supervision, 792 F. Supp. 837, 844 (D.D.C.
                1992).
                ---------------------------------------------------------------------------
                 These principles are easily satisfied here. Here, the agencies
                initially provided a 60-day comment period and then further extended it
                to ensure compliance with the Clean Air Act. The Clean Air Act requires
                that the record of proceedings allowing oral presentation of data,
                views, and arguments on a proposed rule be kept open for 30 days after
                completion of a proceeding to provide an opportunity for submission of
                rebuttal and supplementary information.\2799\ Because the final
                ``proceeding allowing oral presentation of data, views, and arguments''
                was expected to be on September 26, 2018, the comment period for the
                proposed rule was extended by three days to meet that
                requirement.\2800\
                ---------------------------------------------------------------------------
                 \2799\ 42 U.S.C. 7607(d)(5).
                 \2800\ See 83 FR 48578, 48581 (Sept. 26, 2018).
                ---------------------------------------------------------------------------
                 The 63-day comment period was consistent with what the law
                requires.\2801\ While the agencies understand and agree with commenters
                about the importance and complexity of the issues here, the public
                docket demonstrates that the public had a meaningful opportunity to
                comment on the proposed rule.\2802\ The agencies received a total of
                more than 750,000 public comments, many of which commented on detailed,
                technical portions of the proposed rule. For instance, the California
                Air Resources Board provided 415 pages of detailed comments involving
                very specific aspects of the proposal,\2803\ and the Auto Alliance
                filed 202 pages of detailed comments, and commissioned a separate
                econometric study analyzing the effects of multiple alternatives.\2804\
                This is clear evidence that the public had not only the opportunity to
                review and comment on the proposal, but to do so with an extraordinary
                level of detail.
                ---------------------------------------------------------------------------
                 \2801\ In any event, the two Executive Orders explicitly state
                that they do not create any enforceable right or benefit by a party
                against any federal agency. See Executive Order 12,866 Sec. 10; see
                also Executive Order 13,563 Sec. 7(d).
                 \2802\ See Rural Cellular Ass'n v. FCC, 588 F.3d 1095, 1101
                (D.C. Cir. 2009).
                 \2803\ NHTSA-2018-0067-11873.
                 \2804\ NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 Finally, notwithstanding the sufficiency of the agencies' 63-day
                comment period, the agencies published their NPRM on their websites on
                August 2, 2018, more than three weeks before the comment period
                formally opened on August 24, and this effectively provided the public
                with 22 additional days in
                [[Page 25155]]
                which to review the proposal and draft comments.\2805\
                ---------------------------------------------------------------------------
                 \2805\ The agencies' public dockets also remained open for more
                than one year after the start of the comment period, and the
                agencies considered some late comments received, to the extent
                practicable, although many late comments were simply too untimely to
                be considered.
                ---------------------------------------------------------------------------
                b) Other Comments on Public Participation
                 Several commenters objected to NHTSA's 15-page limit on primary
                comments, asserting that it impacted the public's ability to
                meaningfully participate in the rulemaking process.\2806\ However, as
                certain of the commenters acknowledged, the NPRM also explicitly stated
                that commenters could also submit attachments--without any page
                limit.\2807\ Thus, the page limit on primary comments did not prevent
                commenters from presenting any information they deemed relevant to the
                agencies. Both primary comments and their attachments are available in
                the agencies' public dockets, and were considered by the agencies in
                this rulemaking as demonstrated by the responses to comments discussed
                throughout this final rule.
                ---------------------------------------------------------------------------
                 \2806\ See States of California et al., Attachment1_States and
                Cities Detailed Comments, Docket No. NHTSA-2018-0067-11735, at 46;
                Center for Biological Diversity, et al., NHTSA-2018-0067-12088;
                CARB, NHTSA-2018-0067-1187; Environmental Defense Fund, NHTSA-2018-
                0067-12108; BlueGreen Alliance, NHTSA-2018-0067-12440; Connecticut
                Department of Energy and Environmental Protection (DEEP), EPA-HQ-
                OAR-2018-0283-4202.
                 \2807\ 83 FR 43470 (Aug. 24, 2018) (citing 49 CFR 553.21).
                ---------------------------------------------------------------------------
                 NHTSA's 15-page limit simply prescribed the form that comments
                should take: A concise summary comment of up to 15 pages, with optional
                attachments with no page limit. Many commenters submitted extensive
                attachments to their comments, including commenters that objected to
                the 15-page limit for primary comments. For example, several States and
                cities that jointly commented submitted a 13-page primary comment,
                accompanied by 145 pages of ``detailed comments'' and three appendices
                totaling 101 additional pages.\2808\ The 15-page limit had the effect
                of creating executive summaries of otherwise voluminous comments, which
                increased efficiency during the rulemaking process. This was NHTSA's
                stated purpose for the 15-page limit. As explained in the NPRM: ``NHTSA
                established this limit to encourage you to write your primary comments
                in a concise fashion.'' \2809\ In any event, no commenter was prevented
                from submitting information to the agencies based on NHTSA's page
                limitation for primary comments. The agencies strongly disagree that
                public participation was impeded by NHTSA's specification that primary
                comments were limited to 15 pages.
                ---------------------------------------------------------------------------
                 \2808\ States of California et al., NHTSA-2018-0067-11735.
                 \2809\ 83 FR 43470 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 On August 2, 2018, the agencies published a joint Notice of
                Proposed Rulemaking (NPRM) on the agencies' respective websites, which
                solicited public comments on ``The Safer Affordable Fuel-Efficient
                (SAFE) Vehicles Rule for Model Years 2021-2026 Passenger Cars and Light
                Trucks.'' \2810\ The NPRM indicated that the public may submit written
                comments by any of the following methods: Online through the Federal
                eRulemaking Portal at www.regulations.gov, by fax, by mail, or by hand
                delivery. The NPRM also notified the public that the agencies planned
                to hold three joint public hearings, and would accept oral and written
                comments at the hearings. The NPRM indicated that the agencies planned
                to hold the hearings in Washington, DC; the Detroit, Michigan area; and
                the Los Angeles, California area, but indicated that the specific
                addresses and dates for the hearings would be announced in a
                supplemental Federal Register notice.\2811\ On August 24, 2018, the
                agencies published a notice in the Federal Register, which provided new
                locations for two of the three hearings and added dates for each
                hearing.\2812\ That notice informed the public that the agencies
                planned to hold three joint public hearings during the comment period:
                (1) On September 24, 2018 in Fresno, California; (2) on September 25,
                2018 in Dearborn, Michigan; and (3) on September 26, 2018 in
                Pittsburgh, Pennsylvania.\2813\
                ---------------------------------------------------------------------------
                 \2810\ https://www.nhtsa.gov/corporate-average-fuel-economy/safe; https://www.epa.gov/newsreleases/us-epa-and-dot-propose-fuel-economy-standards-my-2021-2026-vehicles. The Agencies subsequently
                published the NPRM in the Federal Register on August 24, 2018. 83 FR
                42986 (August 24, 2018).
                 \2811\ 83 FR 42986 (August 24, 2018).
                 \2812\ 83 FR 42817 (August 24, 2018).
                 \2813\ Id.
                ---------------------------------------------------------------------------
                 The agencies also received several comments with respect to the
                sufficiency of the agencies' public hearings during the comment period.
                For example, the South Coast Air Quality Management District asserted
                that EPA failed to meet its obligation to hold public hearings under
                the Clean Air Act, claiming that an EPA ``political appointee'' did not
                have the legal authority to change hearing locations.\2814\ The comment
                also claimed that holding certain of the hearings in smaller
                metropolitan areas than originally announced resulted in 15 million
                fewer potential participants in the hearings.\2815\ Additionally, the
                comment noted that the NPRM and the notice that set the new locations
                of two of the public hearings were both published in the Federal
                Register on the same day, yet those documents contained conflicting
                hearing locations (the NPRM listed the originally planned hearing
                locations).\2816\
                ---------------------------------------------------------------------------
                 \2814\ See comments from the South Coast Air Quality Management
                District, Attachment 1--SCAQMD Combined NHTSA Waiver Comment (Oct.
                25, 2018), Docket No. NHTSA-2018-0067-11813, at 37-38.
                 \2815\ See id. at 37.
                 \2816\ See id.
                ---------------------------------------------------------------------------
                 Similarly, seventeen States and the District of Columbia submitted
                a joint comment requesting that the agencies reinstate the hearing
                locations that were initially listed in the NPRM, with the stated goal
                of maximizing the number of public participants.\2817\ Similarly, a
                group of environmental organizations jointly submitted a comment
                stating that the new hearing locations failed to maximize the potential
                participants for the agencies' public hearings.\2818\ That group also
                asserted that the agencies failed to provide a reason for the agencies'
                denial of requests to hold more than three public hearings.\2819\
                ---------------------------------------------------------------------------
                 \2817\ See comments from the State of California et al., Request
                for an extension, Docket No. NHTSA-2018-0067-3458.
                 \2818\ See comments from the Center for Biological Diversity,
                Conservation Law Foundation, Environmental Defense Fund,
                Earthjustice, Environmental Law and Policy Center, Natural Resources
                Defense Council, Public Citizen, Inc., Sierra Club, and Union of
                Concerned Scientists, Appendix A--Coalition Comment Letter (10-26-
                2018), Docket No. NHTSA-2018-0067-12000, at 213. A number of other
                commenters also requested that the Agencies hold additional public
                hearings. See, e.g., comments from the Georgetown Climate Center,
                20180906--GCC Comments to NHTSA and EPA, Docket No. NHTSA-2018-0067-
                3610; The City of Los Angeles, Docket No. NHTSA-2018-0067-4159, at
                2-3; California Air Resources Board, 2018-09-11 SAFE Rule DEIS--CARB
                Req Add Info, Docket No. NHTSA-2018-0067-4166, at 1; Northeast
                States for Coordinated Air Use Management, NESCAUM SAFE rule request
                for comment extension and hearing_20180824, Docket No. NHTSA-2018-
                0067-2158, at 1-2.
                 \2819\ Id.
                ---------------------------------------------------------------------------
                 The agencies more than satisfied their legal obligation with
                respect to holding public hearings, and the three hearings provided
                substantial additional opportunity for public participation. While the
                agencies understand that some commenters were disappointed with some
                aspects of the process, those commenters did not demonstrate that the
                agencies' process was legally deficient, nor that any party suffered
                prejudice from the changes the agencies made to their public hearing
                arrangement.
                [[Page 25156]]
                 The APA does not require agencies to hold public hearings during
                the rulemaking process, unless the opportunity for a public hearing is
                required by a governing statute.\2820\ NHTSA's governing fuel economy
                statute does not require a public hearing during the rulemaking
                process.\2821\ The Clean Air Act requires EPA to ``give interested
                persons an opportunity for the oral presentation of data, views, or
                arguments, in addition to an opportunity to make written submissions .
                . . .'' 42 U.S.C. 7607(d)(5)(ii). The agencies' three joint public
                hearings satisfied this statutory requirement.
                ---------------------------------------------------------------------------
                 \2820\ See 5 U.S.C. 553(c). Absent a statutory requirement, the
                APA gives agencies the discretion whether or not to hold a public
                hearing, stating that ``the agency shall give interested persons an
                opportunity to participate in the rule making through submission of
                written data, views, or arguments with or without opportunity for
                oral presentation.'' Id.
                 \2821\ See 49 U.S.C. 32902.
                ---------------------------------------------------------------------------
                 The agencies note that it was clear from the NPRM that the hearings
                were not yet finalized. No addresses or dates were announced for the
                hearings, and the NPRM indicated that information on the hearings would
                be forthcoming in a supplemental Federal Register notice. The NPRM
                (signed by the EPA Administrator) indicated that three hearings would
                be held, and the fact that specific details about those hearings were
                announced in a later notice signed by a different political appointee
                does not itself make the hearings themselves invalid. The Clean Air Act
                does not mandate hearings in any particular location and the public was
                aware from the NPRM that additional information on the hearings would
                be forthcoming. To the extent that any individual person or group was
                inconvenienced by the change in location announced in the supplemental
                notice, they still had ample time to submit public comments through any
                of the multiple other available methods indicated in the NPRM.\2822\
                ---------------------------------------------------------------------------
                 \2822\ Executive Order 13,563 offers guidance to agencies with
                respect to how to maximize public participation. The Executive Order
                states that agencies should ``afford the public a meaningful
                opportunity to comment through the internet on any proposed
                regulation . . . .'' The vast majority of the comments the agencies
                received in this rulemaking were submitted through the internet.
                ---------------------------------------------------------------------------
                 The agencies regret any confusion that resulted from publication of
                the NPRM in the Federal Register on the same date as publication of the
                notice that updated the hearing locations and provided additional
                information, including hearing dates. However, because the NPRM did not
                include dates for the hearings, and the NPRM informed interested
                parties to look for an additional notice that would announce specific
                dates and addresses for the hearings, no one could have relied on the
                NPRM to the exclusion of the supplemental notice.\2823\
                ---------------------------------------------------------------------------
                 \2823\ Additionally, as a matter of fairness, the agencies gave
                interested parties notice about the change in public hearing
                locations one month prior to the first public hearing. See 83 FR
                42817 (August 24, 2018).
                ---------------------------------------------------------------------------
                 The agencies ultimately held three public hearings, as was
                originally announced. There is no Clean Air Act requirement for a
                particular number of hearings, and by holding the hearings in locations
                throughout the United States (including in California), the agencies
                offered a meaningful opportunity for participation. Moreover, the
                public docket remained open for two months subsequent to the
                announcement of the final hearing locations, providing any interested
                party who was unable to attend a public hearing ample opportunity to
                submit comments in writing. As evidence of this meaningful opportunity
                to comment on the proposed rule, the agencies received a total of more
                than 750,000 public comments.
                 Several commenters also asserted that the agencies delayed posting
                the hearing transcripts to the public docket until October 25, which
                was one day before the close of the public comment period.\2824\ The
                Environmental Defense Fund claimed that this was inconsistent with the
                Clean Air Act's requirements that ```[t]he transcript of public
                hearings, if any, on the proposed rule shall also be included in the
                docket promptly upon receipt from the person who transcribed such
                hearings.' 42 U.S.C. 7607(d)(4)(B).'' \2825\ As one commenter
                acknowledged, the transcripts were certified by the reporters on
                September 26, 2018 (Pittsburgh hearing), September 27, 2018 (Dearborn
                hearing), and October 1, 2018 (Fresno hearing).\2826\ The agencies made
                the transcripts publicly available within a reasonable period.
                Moreover, it was reasonable for the agencies to have an opportunity to
                review the transcripts for errors prior to making them publicly
                available. While the concern expressed by these commenters was an
                inadequate ability to offer responsive comments to the transcripts, the
                rulemaking process would be never-ending if every commenter had an
                opportunity to respond to every other commenter. There is no such
                requirement in the APA, the Clean Air Act, or otherwise. The public had
                sufficient opportunity to comment on the agencies' proposals, as
                described above.
                ---------------------------------------------------------------------------
                 \2824\ Environmental Defense Fund, NHTSA-2018-0067-12108, NHTSA-
                2018-0067-12327, NHTSA-2018-0067-12371; State of California et al.,
                NHTSA-2018-0067-11735.
                 \2825\ Environmental Defense Fund, NHTSA-2018-0067-12371.
                 \2826\ State of California et al., NHTSA-2018-0067-11735.
                ---------------------------------------------------------------------------
                 A few commenters requested that the agencies host a workshop or
                webinar to help commenters better understand the agencies' modeling and
                analyses.\2827\ The commenters pointed to similar activities undertaken
                by EPA for other complex rulemakings. While the agencies did not
                conduct a live workshop or webinar regarding the proposal, they did
                make extensive information publicly available beyond the contents of
                the NPRM. To assist the public, NHTSA hosted a dedicated web page with
                information on the modeling.\2828\ The web page included a video
                introduction to the CAFE model.\2829\ The web page enabled members of
                the public to download the model software, its system documentation,
                source code, and input files.\2830\ Many commenters commented in detail
                on the modeling and analyses. However, the agencies recognize that
                public stakeholders vary in their experience and understanding of the
                modeling and analyses and will continue to consider ways to facilitate
                public participation in future rulemakings, which could include the use
                of workshops or webinars.
                ---------------------------------------------------------------------------
                 \2827\ See Minnesota Pollution Control Agency (MPCA), NHTSA-
                2017-0069-0528; Minnesota Pollution Control Agency (MPCA) et al.,
                NHTSA-2018-0067-11706.
                 \2828\ https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
                 \2829\ Id.
                 \2830\ Id.
                ---------------------------------------------------------------------------
                 Some comments criticized the agencies for the agencies'
                untimeliness in adding materials to the rulemaking dockets, for
                example, identifying material ``that was not added to the rulemaking
                docket until the end of the original comment period or, in some cases,
                added either after that period already had closed or not at all.''
                \2831\
                ---------------------------------------------------------------------------
                 \2831\ CBD et. al, Supplemental Comments, Docket No. NHTSA-2018-
                0067-12371, at 8.
                ---------------------------------------------------------------------------
                 The critical question is ``whether the final rule changes
                critically from the proposed rule rather than on whether the agency
                relies on supporting material not published for comment.'' \2832\ In
                other words, ``[t]he question is typically whether the agency's final
                rule so departs from its proposed rule as to constitute more surprise
                than notice.'' \2833\ To that end, agencies are allowed--as the
                agencies here did--to
                [[Page 25157]]
                rely on supplemental data that clarified, expanded on, or confirmed
                information in the proposed rule, even if that supplemental data was
                not disclosed in the proposed rule.\2834\ In any event, the commenters
                have failed to show how they were prejudiced by any information posted
                later than they would have preferred.\2835\
                ---------------------------------------------------------------------------
                 \2832\ Air Transp. Ass'n of Am. v. F.A.A., 169 F.3d 1, 7 (D.C.
                Cir. 1999).
                 \2833\ Id. (citing Air Transp. Ass'n of Am., 732 F.2d 219, 225
                n.12 (D.C. Cir. 1984)).
                 \2834\ See Air Transp. Ass'n of Am. v. F.A.A., 169 F.3d 1, 7
                (D.C. Cir. 1999) (citing Solite Corp. v. EPA, 952 F.2d 473, 485
                (D.C. Cir. 1991); Air Transp. Ass'n of Am. v. CAB, 732 F.2d 219, 224
                (D.C. Cir. 1984)).
                 \2835\ See Solite Corp. v. U.S. E.P.A., 952 F.2d 473, 484 (D.C.
                Cir. 1991) (citing Cmty. Nutrition Inst. v. Block, 749 F.2d 50, 57-
                58 (D.C. Cir. 1984)). Parties also could have submitted comments
                after the end of the comment period on any of these materials. See
                49 CFR 553.23 (NHTSA regulation providing that ``[l]ate filed
                comments will be considered to the extent practicable.'').
                ---------------------------------------------------------------------------
                 Some commenters noted that certain aspects of the CAFE model used
                for the proposal were not previously subject to peer review.\2836\
                Certain commenters asserted that the proposal was legally flawed
                because the full CAFE model was not peer reviewed prior to the
                proposal.\2837\ In support of this argument, commenters cited the
                Information Quality Act and related OMB guidance that states that
                ``each agency shall have a peer review conducted on all influential
                scientific information that the agency intends to disseminate.'' \2838\
                Commenters also cited EPA's Peer Review Handbook, which states: ``For
                highly influential scientific assessments, external peer review is the
                expected procedure.'' \2839\
                ---------------------------------------------------------------------------
                 \2836\ See, e.g., Center for Biological Diversity et al., NHTSA-
                2018-0067-12000; Environmental Defense Fund, NHTSA-2018-0067-12327;
                Environmental Defense Fund et al., NHTSA-2018-0067-12371;
                Environmental Defense Fund et al., NHTSA-2018-0067-12406; Center for
                Biological Diversity, Environment America, Environmental Defense
                Fund, Environmental Law Policy Center, Public Citizen, Inc., Sierra
                Club, and Union of Concerned Scientists, NHTSA-2018-0067-12439;
                States of California et al., NHTSA-2018-0067-11735.
                 \2837\ See, e.g., Center for Biological Diversity et al., NHTSA-
                2018-0067-12000.
                 \2838\ See Center for Biological Diversity et al., NHTSA-2018-
                0067-12000.
                 \2839\ See Center for Biological Diversity et al., NHTSA-2018-
                0067-12000.
                ---------------------------------------------------------------------------
                 The agencies agree that peer review is appropriate for the CAFE
                model, and the CAFE model has been peer reviewed. As discussed in the
                NPRM, and as certain commenters acknowledged, the CAFE model was peer
                reviewed in 2017.\2840\ NHTSA included peer review materials in the
                public docket as well as on its web page regarding the model.\2841\ As
                described in those materials: ``In 2017, the Volpe Center arranged for
                a formal peer review of the version of the CAFE model released and
                documented in 2016 . . . . All of the peer reviewers supported much
                about the model's general approach, and supported many of the model's
                specific characteristics. Peer reviewers also provided a variety of
                general and specific recommendations regarding potential changes to the
                model, inputs, outputs, and documentation. NHTSA and Volpe Center staff
                agree with many of these recommendations and have either completed or
                begun work to implement many of them; implementing others would require
                further research, testing, and development not possible at this time,
                but we are considering them for future model versions.''\2842\
                ---------------------------------------------------------------------------
                 \2840\ 83 FR 43000 (Aug. 24, 2018) (``A report available in the
                docket for this rulemaking presents peer reviewers' detailed
                comments and recommendations, and provides DOT's detailed
                responses.''); see Center for Biological Diversity et al., NHTSA-
                2018-0067-12000.
                 \2841\ NHTSA-2018-0067-0055; https://www.nhtsa.gov/corporate-average-fuel-economy/compliance-and-effects-modeling-system.
                 \2842\ NHTSA-2018-0067-0055.
                ---------------------------------------------------------------------------
                 However, certain new elements of the CAFE model were not completed
                at the time of the 2017 peer review.\2843\ NHTSA subsequently obtained
                a peer review of significant new elements added to the model after the
                2017 peer review.\2844\ As described in the new peer review charge,
                included in a July 2019 report included in the rulemaking docket, NHTSA
                explained:
                ---------------------------------------------------------------------------
                 \2843\ NHTSA-2018-0067-0055 (explaining, in responses to 2017
                peer review, that ``[t]he model has been updated to including
                procedures to estimate impacts on new vehicle sales, and on older
                vehicle scrappage'').
                 \2844\ NHTSA-2018-0067-0055.
                 To inform the proposed rule announced in August 2018, DOT staff
                introduced significant new elements to the model, including methods
                to estimate changes in vehicle sales volumes, vehicle scrappage, and
                automotive sector labor usage. Each of these regulatory actions
                involved consideration of and response to significant public comment
                on model results, as well as comments on the model itself. In
                addition to DOT staff's own observations, these comments led DOT
                staff to make a wide range of improvements to the model. Insofar as
                a formal peer review could identify additional potential
                opportunities to improve the model, DOT sponsored a review of the
                entire model in 2017. At this time, DOT seeks review of some of the
                ---------------------------------------------------------------------------
                significant new elements added to the model after that review.
                 This subsequent peer review of the new elements was not complete at
                the time the proposal was published, and therefore materials concerning
                the peer reviewers' comments and NHTSA's responses were not available
                until later.\2845\ Although the comment period on the proposal had
                closed at that time, the agencies continued to receive comments on the
                new peer review materials, which they have considered in issuing this
                final rule.\2846\ Of course, the new elements of the modeling were also
                described in detail in the NPRM and commenters also directly commented
                on them in great detail. Thus, the public was fully apprised of all
                aspects of the modeling and had a robust opportunity to provide
                comment.
                ---------------------------------------------------------------------------
                 \2845\ NHTSA-2018-0067-0055 (July 2019 report).
                 \2846\ See, e.g., Center for Biological Diversity et al., NHTSA-
                2018-0067-12439; Environment America et al., NHTSA-2018-0067-12441.
                ---------------------------------------------------------------------------
                 To the extent commenters are suggesting the Information Quality Act
                required a full peer review of all aspects of the CAFE model prior to
                the proposal, the agencies disagree.\2847\ Peer review of the new
                elements of the CAFE model helped ensure that the model is
                scientifically sound, and the peer reviewers provided feedback that
                helped improve the model and may help develop additional improvements
                to the model in the future. In this sense, the peer review of the new
                elements of the model functioned similarly to public comments from
                commenters with specific scientific expertise. Much of the feedback
                from the peer reviewers were in fact similar in nature to comments
                received from public commenters on the model. By engaging in both peer
                review and notice-and-comment procedures, the agencies ensured that
                they had information from a wide variety of sources, including those
                with specific expertise, to validate and improve the model.\2848\ The
                technical aspects of the model, including improvements made to the
                model following the proposal, are described in detail in this final
                rule. Moreover, as the Center for Biological Diversity noted, the
                Information Quality Act does not create third-party rights.\2849\
                ---------------------------------------------------------------------------
                 \2847\ See, e.g., Center for Biological Diversity et al., NHTSA-
                2018-0067-12000; Environment America et al., NHTSA2018-0067-12441.
                 \2848\ The timing of the peer review of new elements of the
                model also did not require a second cycle of notice and comment.
                See, e.g., Alto Dairy v. Veneman, 336 F.3d 560, 569-70 (7th Cir.
                2003) (``The law does not require that every alteration in a
                proposed rule be reissued for notice and comment. If that were the
                case, an agency could `learn from the comments on its proposals only
                at the peril of subjecting itself to rulemaking without end.''').
                 \2849\ Center for Biological Diversity et al., NHTSA-2018-0067-
                12000.
                ---------------------------------------------------------------------------
                 The agencies also disagree that EPA needed to obtain a separate
                peer review of the CAFE model.\2850\ The peer review addressed aspects
                of the model relevant to the analysis by both agencies under their
                respective statutory schemes. The agencies have expertise in their
                [[Page 25158]]
                statutory requirements and discussed in detail both in the proposal and
                this final rule how the CAFE model was used to inform the decision-
                making under both EPCA and the CAA.
                ---------------------------------------------------------------------------
                 \2850\ Center for Biological Diversity et al., NHTSA-2018-0067-
                12000.
                ---------------------------------------------------------------------------
                (c) Other APA Comments
                 Many commenters suggested that the record of evidence developed for
                the 2016 Draft TAR and EPA's Original Determination was a better basis
                for NHTSA to determine maximum feasible standards than the record of
                evidence for the current rulemaking. These commenters also argued that,
                in the NPRM, NHTSA ignored the findings and analysis in the TAR and the
                Technical Support Document and contradicted the pre-existing record
                without explanation. Lastly, these commenters argued that the NPRM did
                not have a reasoned basis under the APA, particularly in light of the
                agency's change in position and the reliance interests at stake.
                 Agencies always have authority under the Administrative Procedure
                Act to revisit previous decisions in light of new facts, as long as
                they provide notice and an opportunity for comment--as the agencies did
                here. Indeed, it is the best practice to do so when changed
                circumstances so warrant.\2851\
                ---------------------------------------------------------------------------
                 \2851\ See FCC v. Fox Television, 556 U.S. 502 (2009).
                ---------------------------------------------------------------------------
                 ``Changing policy does not, on its own, trigger an especially
                `demanding burden of justification.' '' \2852\ ``Agencies are free to
                change their existing policies as long as they provide a reasoned
                explanation for the change.'' \2853\ Providing this explanation ``would
                ordinarily demand that [the agency] display awareness that it is
                changing position.'' \2854\ Beyond that, however, ``[w]hen an agency
                changes its existing position, it `need not always provide a more
                detailed justification than what would suffice for a new policy created
                on a blank slate.' '' \2855\ The agency ``need not demonstrate to a
                court's satisfaction that the reasons for the new policy are better
                than the reasons for the old one.'' \2856\ For instance, ``evolving
                notions'' about the appropriate balance of varying policy
                considerations constitute sufficiently good reasons for a change in
                position.\2857\ A change in policy is ``well within an agency's
                discretion:'' Agencies are permitted to conduct a ``reevaluation of
                which policy would be better in light of the facts,'' without
                ``rely[ing] on new facts.'' \2858\
                ---------------------------------------------------------------------------
                 \2852\ Mingo Logan Coal Co. v. Envtl. Prot. Agency, 829 F.3d
                710, 718 (DC Cir. 2016) (quoting Ark Initiative v. Tidwell, 816 F.3d
                119, 127 (DC Cir. 2016)).
                 \2853\ Encino Motorcars, LLC v. Navarro, 136 S. Ct. 2117, 2125
                (2016) (citations omitted).
                 \2854\ FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515
                (2009) (emphasis in original) (``An agency may not, for example,
                depart from a prior policy sub silentio or simply disregard rules
                that are still on the books.'').
                 \2855\ Encino Motorcars, LLC v. Navarro, 136 S. Ct. 2117, 2125-
                26 (2016) (quoting FCC v. Fox Television Stations, Inc., 556 U.S.
                502, 515 (2009)).
                 \2856\ FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515
                (2009) (emphasis in original).
                 \2857\ N. Am.'s Bldg. Trades Unions v. Occupational Safety &
                Health Admin., 878 F.3d 271, 303 (D.C. Cir. 2017) (quoting the
                agency's rule).
                 \2858\ Nat'l Ass'n of Home Builders v. E.P.A., 682 F.3d 1032,
                1037-38 (D.C. Cir. 2012).
                ---------------------------------------------------------------------------
                 To be sure, providing ``a more detailed justification'' is
                appropriate in some cases.\2859\ But when ``a more detailed
                justification'' is needed, all that is required is for the agency to
                explain how ``new information arising after'' the previous
                determination ``informed its conclusion'' that a change was
                appropriate: ``Explanations relying on new data are sufficient to
                satisfy the more detailed explanatory obligation.'' \2860\ As one of
                the critical comments itself noted, ``[a]gencies must use `the best
                information available' in reaching their conclusions, and cannot
                lawfully rely on outdated information as circumstances change.'' \2861\
                Accordingly, when new information became available, the agencies relied
                on it expressly, resulting in a fully-explained change in their
                analysis and ultimately their conclusions.
                ---------------------------------------------------------------------------
                 \2859\ FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515
                (2009) (``Sometimes [the agency] must [provide a more detailed
                justification than what would suffice for a new policy created on a
                blank slate]--when, for example, its new policy rests upon factual
                findings that contradict those which underlay its prior policy; or
                when its prior policy has engendered serious reliance interests that
                must be taken into account.'').
                 \2860\ Mingo Logan Coal Co. v. Envtl. Prot. Agency, 829 F.3d
                710, 727 (D.C. Cir. 2016).
                 \2861\ CBD et. al, Appendix A, Docket No. NHTSA-2018-0067-12000,
                at 11 (quoting Flyers Rights Education Fund v. FAA, 864 F. 3d 738,
                745 (D.C. Cir. 2017)).
                ---------------------------------------------------------------------------
                 While ``[i]t would be arbitrary or capricious to ignore such
                matters,''\2862\ the agencies have not ignored them. NHTSA has
                satisfied these standards. The NPRM expressly and repeatedly
                acknowledged that it represented a change from the 2012 final rule, the
                Draft TAR, and EPA's Original Determination, appropriately justifying
                the change by citing shifts in policy priorities or new facts and
                changed circumstances that became apparent since the Original
                Determination.\2863\ The agencies are fully cognizant of the facts and
                circumstances that have changed since the Original Determination,
                expressly acknowledged them in the Revised Determination and SAFE Rule
                NPRM, and adapted to accept them now in the final rule.
                ---------------------------------------------------------------------------
                 \2862\ FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515
                (2009).
                 \2863\ See, e.g., 83 FR at 43213 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 Several commenters invoked requests to the agencies under the
                Freedom of Information Act (``FOIA'') regarding material sought in
                connection with the rulemaking.\2864\ These comments ranged from simple
                references to existing FOIA requests to the agencies, to the actual
                submission of the FOIA requests as a comment posted to the rulemaking
                docket.\2865\ These commenters sought a variety of information, which
                included calendars and internal correspondence of specific agency
                personnel, communications with non-governmental stakeholders, and
                technical materials and clarifications relating to aspects of the
                agencies' analysis.\2866\
                ---------------------------------------------------------------------------
                 \2864\ See, e.g., Environmental Defense Fund, NHTSA-2018-0067-
                12371.
                 \2865\ Compare, e.g., Joint Submission from the States of
                California et al. and the Cities of Oakland et al., NHTSA NHTSA-
                2018-0067-11735, with, e.g., Office of the Attorney General of the
                State of New York, NHTSA-2018-0067-3613.
                 \2866\ See, e.g., Environmental Defense Fund, NHTSA-2018-0067-
                12397; Office of the Attorney General of the State of New York,
                NHTSA-2018-0067-3613; California Air Resources Board, NHTSA-2018-
                0067-4166.
                ---------------------------------------------------------------------------
                 To the extent these requests sought substantive material, those
                matters are addressed in other sections herein that pertain to the
                respective underlying issues implicated. Although the submission of
                FOIA requests through an online rulemaking docket is a very unusual
                form of submitting a FOIA request to an agency, the agencies
                nevertheless processed the comments that requested materials by
                invoking FOIA as formal FOIA requests. As such, once identified, those
                comments were forwarded to the agencies' respective FOIA offices, which
                commenced the intake process of the letters as FOIA requests. In turn,
                the agencies' FOIA offices transmitted receipt acknowledgement letters
                to the requestors and conducted searches for the applicable material.
                The agencies responded to the requestors by producing the responsive
                non-exempt records identified, applying the appropriate FOIA standards
                applicable to the records and requests. Like all other typical FOIA
                requests, the requestors were provided with an opportunity to
                administratively appeal the FOIA decision and, if desired, subsequently
                seek judicial review of the agencies' decisions. Several commenters
                availed themselves of this procedure.\2867\
                ---------------------------------------------------------------------------
                 \2867\ See generally, e.g., New York v. U.S. Envtl. Prot. Agency
                and Nat'l Highway Traffic Safety Admin., Case No. 1:19-cv-00712
                (S.D.N.Y.) (FOIA litigation concerning a FOIA request submitted as a
                comment from the Office of the Attorney General of the State of New
                York, NHTSA-2018-0067-3613).
                ---------------------------------------------------------------------------
                [[Page 25159]]
                 Thus, the agencies fully satisfied their obligations under the
                governing FOIA provisions. In fact, other commenters noted the
                agencies' responses to these FOIA requests and incorporated information
                disclosed in the responses into their comments.\2868\ Moreover, several
                of the FOIA requests submitted as comments requested information that
                had already been published on the agencies' websites for the rulemaking
                or in the rulemaking dockets.
                ---------------------------------------------------------------------------
                 \2868\ See James H. Stock, Kenneth Gillingham & Wade Davis, EPA-
                HQ-OAR-2018-0283-6220, at p. 6.
                ---------------------------------------------------------------------------
                 Although the agencies fulfilled their obligations under all
                applicable FOIA law, the agencies also stress that FOIA compliance is
                wholly irrelevant to conformity to governing APA standards in the
                rulemaking process. FOIA arises from an independent statutory
                framework, which contains unique provisions for judicial review.\2869\
                These provisions for judicial review provide ``an adequate form of
                relief'' such that the APA is not typically even an appropriate
                mechanism to seek the disclosure of further information requested under
                FOIA.\2870\ Likewise, the APA's principles governing rulemaking
                procedures, including disclosures of information for such rulemakings,
                exist as autonomous statutory and jurisprudential concepts totally
                untethered from the principles of disclosure under FOIA.
                ---------------------------------------------------------------------------
                 \2869\ 5 U.S.C. 552(a)(4)(B).
                 \2870\ See, e.g., Feinman v. FBI, 713 F. Supp. 2d 70, 76 (D.D.C.
                2010) (``This court and others have uniformly declined jurisdiction
                over APA claims that sought remedies made available by FOIA.'').
                ---------------------------------------------------------------------------
                 Similarly, as an independent statutory framework from the APA, the
                susceptibility of materials and records for production under FOIA has
                no bearing on whether such materials should have been made public under
                the APA as part of a rulemaking. The scope of materials for production
                under FOIA arises from the Agency's reasonable interpretation of the
                language of the FOIA request, as well as the exemptions potentially
                applicable to the records under the applicable FOIA statutes and
                implementing regulations.\2871\ In contrast, in an APA review of
                rulemaking procedures, separate standards exist to govern the scope of
                materials an agency must make available during the rulemaking
                process.\2872\ Thus, records may be responsive to a FOIA request, but
                not appropriate for publication under the APA--even if the FOIA request
                concerns the proposed rule in question. The FOIA requests at issue here
                are illustrative of this distinction. For example, one of the specific
                FOIA requests identified by commenters describes the requests as
                pertaining to the NPRM, but seeks Outlook calendars of DOT and NHTSA
                personnel.\2873\ While such materials may be responsive to the
                underlying FOIA requests, which expressly mention the calendars, an
                employee's entire list of calendar appointments--including appointments
                unrelated to the rulemaking--is clearly not contemplated by the APA as
                material necessary for publication along with a proposed rule. Thus,
                while the agencies sought to comply with their independent statutory
                obligations under FOIA, to the extent commenters invoke purported FOIA
                noncompliance, the agencies consider such arguments irrelevant to the
                rulemaking analysis. Likewise, any production of records in connection
                with any FOIA request that invokes the proposed rule is not a
                recognition by the agencies that the material should have also been
                made available during the rulemaking under the APA.
                ---------------------------------------------------------------------------
                 \2871\ See 5 U.S.C. 552. See also, e.g., Weisberg v. U.S. Dep't
                of Justice, 745 F.2d 1476, 1485 (DC Cir. 1984) (discussing standards
                applicable to the scope of an Agency's search for records under
                FOIA).
                 \2872\ See Air Transp. Ass'n of Am. v. F.A.A., 169 F.3d 1, 7 (DC
                Cir. 1999) (discussing the scope of materials for an agency to make
                available during a notice and comment period).
                 \2873\ See Environmental Defense Fund, NHTSA-2018-0067-12397.
                ---------------------------------------------------------------------------
                 Several commenters also criticized the agencies, and specifically
                the EPA, for not publishing an updated version of the Optimization
                Model for Reducing Emissions of Greenhouse Gases from Automobiles
                (``OMEGA'') along with the proposed rule.\2874\ As described in further
                detail in Section IV herein, OMEGA is a fleet compliance model
                developed by the EPA and used in previous rulemakings. While many
                commenters raised technical arguments comparing the OMEGA model to the
                CAFE Model utilized in this rulemaking, such technical analysis and
                comments are addressed elsewhere in this final rule analysis. See
                Section IV. Likewise, while several comments refer to FOIA requests for
                OMEGA model materials, the Agencies' discussion of FOIA comments are
                addressed above.
                ---------------------------------------------------------------------------
                 \2874\ See, e.g., International Council on Clean Transportation,
                NHTSA-2018-0067-11741.
                ---------------------------------------------------------------------------
                 Most other commenters who raised more procedural arguments
                concerning the unavailability of an updated version of the OMEGA model
                argued that an updated version of the model should have been released
                because the EPA utilized the model during an interagency review of the
                proposed rule.\2875\ In considering these comments, the agencies
                emphasize that neither NHTSA, the EPA, nor any other interagency
                reviewer relied upon the OMEGA model for the preparation of either the
                proposed or the final versions of the SAFE Vehicles Rule. Instead, as
                clearly expressed in rulemaking descriptions and documents accompanying
                both this final rule and the proposed rule, the agencies relied on a
                separate model to perform the analysis that helped to inform the
                agencies regarding potential effects of various fuel economy standards.
                This independent model, the CAFE Model, was developed by the Department
                of Transportation's Volpe National Transportation Systems Center.
                ---------------------------------------------------------------------------
                 \2875\ See, e.g., Sallie E. Davis, NHTSA-2018-0067-12430.
                ---------------------------------------------------------------------------
                 In fact, most commenters discussing the OMEGA model understood and
                expressly acknowledged that the agencies relied upon the CAFE Model
                rather than the OMEGA model for this rulemaking.\2876\ Several
                commenters even paradoxically argued both that the agencies
                unreasonably failed to utilize the OMEGA model and that the agencies
                denied meaningful opportunity for comment by utilizing but failing to
                publish an updated OMEGA model.\2877\ Nevertheless, the analysis and
                universe of documents published for the proposed rule made abundantly
                clear that the CAFE Model--not the OMEGA model--performed the
                applicable analysis for this rulemaking. Likewise, the agencies'
                proposed rule published voluminous analyses and supporting documents to
                describe the CAFE Model and explain the underlying methodologies
                incorporated into the model's operation for this rulemaking. The
                agencies also released the full version of the CAFE Model employed in
                this rulemaking, as well as its respective inputs and outputs, in order
                to provide commenters with ample opportunities to understand the
                model's function and operation.
                ---------------------------------------------------------------------------
                 \2876\ See, e.g., Union of Concerned Scientists, NHTSA-2018-
                0067-12303-016; Center for Biological Diversity, NHTSA-2018-0067-
                12000.
                 \2877\ See, e.g., Environmental Defense Fund, NHTSA-2018-0067-
                12108.
                ---------------------------------------------------------------------------
                 The extensive comments on the modeling conducted for this
                rulemaking confirm that the agencies provided the public with
                sufficient information to comment on the modeling process for the
                rulemaking. Comments regarding the OMEGA and CAFE models were
                expansive, spanning hundreds of pages of technical analysis and
                submissions from a variety of commenters. Many of these comments even
                consisted of detailed and technical comparisons of
                [[Page 25160]]
                the CAFE model used in this rulemaking with past versions of OMEGA
                models used for prior rulemakings.\2878\ Even if certain of these
                commenters disagreed with the Agencies' ultimate approach to the
                modeling, they evidently understood the applicable methodologies and
                performance of the CAFE Model for this rulemaking sufficiently to
                substantively engage with the Agencies on these topics through their
                comments. Therefore, the agencies consider the detailed comments on the
                OMEGA and CAFE models as clear indicia that the extensive information,
                materials, and explanations provided by the agencies in the proposed
                rule enabled significant opportunity for the public to comment on the
                modeling for the rule.
                ---------------------------------------------------------------------------
                 \2878\ See, e.g., California Air Resources Board, NHTSA-2018-
                0067-11873; Union of Concerned Scientists, NHTSA-2018-0067-12039;
                Alliance of Automobile Manufacturers, NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 To the extent that commenters allege an insufficient opportunity to
                comment by claiming that the EPA actually utilized the OMEGA model in
                the rulemaking process, the agencies consider such comments
                unfounded.\2879\ The agencies did not rely on the OMEGA model during
                the rulemaking process, including during the analysis for the proposed
                and final rules. In past rulemakings, the EPA developed a complete
                final version of the OMEGA model to perform the rulemaking analysis.
                Here, the EPA did not even finalize a completed updated version of the
                OMEGA model, much less rely on such a model in the course of the
                rulemaking. Therefore, no completed version of an updated OMEGA model
                even existed for the agencies to publish as part of the notice of
                proposed rulemaking.
                ---------------------------------------------------------------------------
                 \2879\ See, e.g., Center for Biological Diversity, NHTSA-2018-
                0067-12000.
                ---------------------------------------------------------------------------
                 To the extent commenters argue that the EPA should have updated the
                model for this rulemaking, the APA's facilitation of a meaningful
                opportunity to comment neither requires nor contemplates a mandate that
                the agencies develop computational modeling alternatives for the
                public, which were not even incorporated into the agencies' own
                rulemaking analysis.\2880\ In fact, doing so would actually detract
                from the notice and comment process because it would convolute the
                rulemaking docket and inhibit the public's ability to identify the
                modeling materials actually used in the rulemaking process. Thus, such
                extraneous materials would only dilute the rulemaking docket with
                voluminous and complex materials, such as modeling files, input files,
                and statistical figures, that had no influence on the rulemaking in
                question. Indeed, several commenters already claimed that the
                voluminous and complex supporting materials in the rulemaking docket
                required significant time for review, so the introduction of extensive
                totally extraneous material would have been only counterproductive to
                the process.\2881\
                ---------------------------------------------------------------------------
                 \2880\ See, e.g., Center for Biological Diversity et al., NHTSA-
                2018-0067-12000.
                 \2881\ See, e.g., Institute for Policy Integrity, NHTSA-2018-
                0067-5641; Northeast States for Coordinated Air Use Management,
                NHTSA-2018-0067-2158.
                ---------------------------------------------------------------------------
                 Moreover, requiring the EPA to perform the work necessary to fully
                update the OMEGA model solely for a public release--when it did not
                otherwise intend to consider the model in the rulemaking--would divert
                valuable and finite agency resources away from actual rulemaking
                analyses in favor of efforts that further no progress in the
                rulemaking.\2882\ Such an approach would detract from the agencies'
                opportunities to devote time to other considerations that actually
                influenced the rulemaking, such as the substantive analysis
                incorporated into the proposed rule and the drafting of extensive
                language to explain to the public the methodologies applied by the
                agencies for the proposal. Such an inefficient allocation of resources
                undermines both the rulemaking process envisioned by the APA and the
                very notice and comment procedures utilized by these commenters.
                ---------------------------------------------------------------------------
                 \2882\ See, e.g., Environmental Defense Fund, NHTSA-2018-0067-
                12108.
                ---------------------------------------------------------------------------
                 Several commenters also argued that even if the agencies did not
                rely on the model for this rulemaking, the OMEGA model still informed
                the EPA's analysis and interagency review by providing general
                background experience in regulating greenhouse gas emissions--either
                through the agency's work with prior versions of the model or ongoing
                efforts to update the OMEGA model for purposes unrelated to this
                rulemaking. However, even assuming the model provided background
                experience to the EPA in regulating in this arena, federal
                jurisprudence makes clear that ``[t]he Administrative Procedure Act
                does not require that every bit of background information used by an
                administrative agency be published for public comment.'' See B. F.
                Goodrich Co. v. Dep't of Transp., 541 F.2d 1178, 1184 (6th Cir. 1976).
                This is particularly the case when, as here, ``[t]he basic data upon
                which the agency relied in formulating the regulation was available . .
                . for comment.'' Id.; see also Am. Min. Cong. v. Marshall, 671 F.2d
                1251, 1261 (10th Cir. 1982) (``These documents consist of background
                information and data as well as several internal memoranda. There is
                nothing to indicate that the Secretary actually relied on any of these
                documents in promulgating the rule or that the data they contain was
                critical to the formulation of the rule.''). In fact, publishing such
                background information not only exceeds the requirements of the APA,
                but would actually affirmatively undermine the APA's notice and comment
                procedure. If every piece of information ever referenced by the
                agencies or upon which the Agencies drew regulatory experience were
                required to be published, rulemaking dockets would expand to an absurd
                scope of nearly infinite materials, spanning arguably back to even the
                school textbooks the rulemaking personnel used to learn the underlying
                disciplines employed in the rulemaking analysis. Clearly such a scope
                would frustrate rather than further the provision of proper notice to
                the public about a proposed rule.\2883\
                ---------------------------------------------------------------------------
                 \2883\ To the extent commenters seek to understand the manner in
                which the OMEGA model informed prior rulemaking efforts, the EPA has
                released the full versions of prior OMEGA models and applicable
                materials along with the prior rulemakings. In fact, several
                commenters referenced such materials in submitting detailed comments
                comparing the CAFE Model with the OMEGA model. Manufacturers of
                Emission Controls Association, NHTSA-2018-0067-11994. Thus, any
                commenters that were interested in such extraneous background
                information had ample opportunity to access the material.
                ---------------------------------------------------------------------------
                 Moreover, even assuming the premise of several commenters'
                challenges--that the EPA consulted updates to the OMEGA model during
                the interagency review--such a predicate still would not require the
                publication of the model during the rulemaking process.\2884\ As the
                agencies have made clear, the OMEGA model did not affect any part of
                the rule, including the methodologies and analysis underlying the
                formulation of the rule. Therefore, even if consulted, the OMEGA model
                would exist as, at most, supplementary material which had no influence
                on the rulemaking methodologies, all of which were fully disclosed.
                See, e.g., Chamber of Commerce of U.S. v. SEC., 443 F.3d 890, 900 (DC
                Cir. 2006) (``When the agency relies on supplementary evidence without
                a showing of prejudice by an interested party, the procedural
                requirements of the APA are satisfied without further opportunity for
                comment, provided that the agency's response constitutes a logical
                outgrowth
                [[Page 25161]]
                of the rule initially proposed'') (internal citations omitted).
                ---------------------------------------------------------------------------
                 \2884\ See, e.g., Environmental Defense Fund, NHTSA-2018-0067-
                12406.
                ---------------------------------------------------------------------------
                3. National Environmental Policy Act
                 As discussed above, EPCA requires NHTSA to determine the level at
                which to set CAFE standards for each model year by considering the four
                factors of technological feasibility, economic practicability, the
                effect of other motor vehicle standards of the Government on fuel
                economy, and the need of the United States to conserve energy. The
                National Environmental Policy Act (NEPA) directs that environmental
                considerations be integrated into that process.\2885\ To explore the
                potential environmental consequences of this rulemaking action, NHTSA
                prepared a Draft Environmental Impact Statement (``DEIS'') for the NPRM
                and a Final Environmental Impact Statement (``FEIS'') for the final
                rule. The purpose of an EIS is to ``provide full and fair discussion of
                significant environmental impacts and [to] inform decisionmakers and
                the public of the reasonable alternatives which would avoid or minimize
                adverse impacts or enhance the quality of the human environment.''
                \2886\
                ---------------------------------------------------------------------------
                 \2885\ NEPA is codified at 42 U.S.C. 4321-47. The Council on
                Environmental Quality (CEQ) NEPA implementing regulations are
                codified at 40 CFR parts 1500-08.
                 \2886\ 40 CFR 1502.1.
                ---------------------------------------------------------------------------
                 As explained in the NPRM, NEPA is ``a procedural statute that
                mandates a process rather than a particular result.'' \2887\ The
                agency's overall EIS-related obligation is to ``take a `hard look' at
                the environmental consequences before taking a major action.'' \2888\
                Significantly, ``[i]f the adverse environmental effects of the proposed
                action are adequately identified and evaluated, the agency is not
                constrained by NEPA from deciding that other values outweigh the
                environmental costs.'' \2889\ The agency must identify the
                ``environmentally preferable'' alternative but need not adopt it.\2890\
                ``Congress in enacting NEPA . . . did not require agencies to elevate
                environmental concerns over other appropriate considerations.'' \2891\
                Instead, NEPA requires an agency to develop and consider alternatives
                to the proposed action in preparing an EIS.\2892\ The statute and
                implementing regulations do not command the agency to favor an
                environmentally preferable course of action, only that it make its
                decision to proceed with the action after taking a hard look at the
                potential environmental consequences and consider the relevant factors
                in making a decision among alternatives.\2893\
                ---------------------------------------------------------------------------
                 \2887\ Stewart Park & Reserve Coal., Inc. v. Slater, 352 F.3d
                545, 557 (2d Cir. 2003).
                 \2888\ Baltimore Gas & Elec. Co. v. Natural Resources Defense
                Council, Inc., 462 U.S. 87, 97 (1983).
                 \2889\ Robertson v. Methow Valley Citizens Council, 490 U.S.
                332, 350 (1989).
                 \2890\ 40 CFR 1505.2(b).
                 \2891\ Baltimore Gas, 462 U.S. at 97.
                 \2892\ 42 U.S.C. 4332(2)(C)(iii).
                 \2893\ 40 CFR 1505.2(b).
                ---------------------------------------------------------------------------
                 NHTSA received many comments on the DEIS. Among the comments
                received, many commenters stated that the baseline/no-action standards
                were the environmentally preferable alternative and argued that the
                environmental benefits of the proposal were (1) insufficient and/or (2)
                incorrectly assessed in a variety of ways. Comments regarding the
                environmental analyses presented in this preamble are addressed in
                Section VI above, while those regarding the DEIS are addressed in
                Chapter 10 of the FEIS.
                 When preparing an EIS, NEPA requires an agency to compare the
                potential environmental impacts of its proposed action and a reasonable
                range of alternatives. In the DEIS, NHTSA analyzed a No Action
                Alternative and eight action alternatives. In the FEIS, NHTSA analyzed
                the same No Action Alternative and seven action alternatives, including
                a new alternative (the Preferred Alternative) within the range of the
                alternatives considered in the DEIS and FEIS.\2894\ The alternatives
                represent a range of potential actions the agency could take, and they
                are described more fully in Section V above, below in this section, and
                Chapter 2 of the FEIS. The environmental impacts of these alternatives,
                in turn, represent a range of potential environmental impacts that
                could result from NHTSA's setting maximum feasible fuel economy
                standards for passenger cars and light trucks.
                ---------------------------------------------------------------------------
                 \2894\ In its scoping notice, NHTSA indicated that the action
                alternatives analyzed would bracket a range of reasonable annual
                fuel economy standards, allowing the agency to select an action
                alternative in its final rule from any stringency level within that
                range. 82 FR 34740, 34743 (July 26, 2017).
                ---------------------------------------------------------------------------
                 To derive the direct and indirect impacts of the action
                alternatives, NHTSA compared each action alternative to the No Action
                Alternative, which reflects baseline trends that would be expected in
                the absence of any further regulatory action other than finalizing the
                augural standards. More specifically, the No Action Alternative in the
                DEIS and FEIS assumed that NHTSA would not amend the CAFE standards for
                MY 2021 passenger cars and light trucks. In addition, the No Action
                Alternative assumed that NHTSA would finalize the MY 2022-2025 augural
                CAFE standards that were described in the 2012 final rule. Finally, for
                purposes of its analysis, NHTSA assumed that the MY 2025 augural
                standards would continue indefinitely. The augural standards also serve
                as a proxy for EPA's CO2 standards for MYs 2022-2025, which
                were also finalized in the 2012 final rule. The No Action Alternative
                provides an analytical baseline against which to compare the
                environmental impacts of other alternatives presented in the EIS.\2895\
                ---------------------------------------------------------------------------
                 \2895\ See 40 CFR 1502.2(e), 1502.14(d). CEQ has explained that
                ``[T]he regulations require the analysis of the no action
                alternative even if the agency is under a court order or legislative
                command to act. This analysis provides a benchmark, enabling
                decision makers to compare the magnitude of environmental effects of
                the action alternatives [See 40 CFR 1502.14(c).] . . . Inclusion of
                such an analysis in the EIS is necessary to inform Congress, the
                public, and the President as intended by NEPA. [See 40 CFR
                1500.1(a).]'' Forty Most Asked Questions Concerning CEQ's National
                Environmental Policy Act Regulations, 46 FR 18026 (Mar. 23, 1981).
                ---------------------------------------------------------------------------
                 For the DEIS, NHTSA analyzed eight action alternatives,
                Alternatives 1 through 8, which ranged from amending the MY 2021
                standards to match the MY 2020 standards and holding those standards
                flat for passenger cars and light trucks through MY 2026 (Alternative
                1) to maintaining the existing MY 2021 standards and subsequently
                requiring average annual increases in fuel economy by 2.0 percent
                (passenger cars) and 3.0 percent (light trucks) (Alternative 8). The
                action alternatives analyzed in the DEIS also reflected different
                options regarding air conditioning efficiency and off-cycle technology
                adjustment procedures, with some alternatives phasing out these
                adjustments in MYs 2022-2026. For the FEIS, NHTSA analyzed seven action
                alternatives, Alternatives 1 through 7, which range from amending the
                MY 2021 standards to match the MY 2020 standards and holding those
                standards flat for passenger cars and light trucks through MY 2026
                (Alternative 1) to maintaining the existing MY 2021 standards and
                subsequently requiring average annual increases in fuel economy by 2.0
                percent (passenger cars) and 3.0 percent (light trucks) (Alternative 7)
                from year to year. The primary differences between the action
                alternatives for the DEIS and FEIS is that the FEIS did not analyze
                alternatives that phased out the air conditioning efficiency and off-
                cycle technology adjustments (see Section V above for further
                discussion), and the FEIS added an alternative under which fuel economy
                increased at 1.5 percent per year for both cars and light trucks
                (Alternative 3). Both of the ranges of action alternatives, as well as
                the No
                [[Page 25162]]
                Action Alternative, in the DEIS and FEIS encompassed a spectrum of
                possible standards NHTSA could determine was maximum feasible based on
                the different ways the agency could weigh EPCA's four statutory
                factors. Throughout the FEIS, estimated impacts were shown for all of
                these action alternatives, as well as for the No Action Alternative.
                For a more detailed discussion of the environmental impacts associated
                with the alternatives, see Chapters 3-8 of the FEIS, as well as Section
                VII above.
                 NHTSA's FEIS describes potential environmental impacts to a variety
                of resources, including fuel and energy use, air quality, climate, land
                use and development, hazardous materials and regulated wastes,
                historical and cultural resources, noise, and environmental justice.
                The FEIS also describes how climate change resulting from global carbon
                emissions (including CO2 emissions attributable to the U.S.
                light duty transportation sector under the alternatives considered)
                could affect certain key natural and human resources. Resource areas
                are assessed qualitatively and quantitatively, as appropriate, in the
                FEIS, and the findings of that analysis are summarized here.\2896\
                ---------------------------------------------------------------------------
                 \2896\ The impacts described in this section come from NHTSA's
                FEIS, which is being publicly issued simultaneously with this final
                rule. As described in Section VII.A.4.c.1 above, the FEIS is based
                on ``unconstrained'' modeling rather than ``standard setting''
                modeling; NHTSA conducts modeling both ways in order to reflect the
                various statutory requirements of EPCA and NEPA. The preamble
                employs the ``standard setting'' modeling in order to ensure that
                the decision-maker does not consider things that EPCA/EISA prohibit,
                but as a result, the impacts reported here may differ from those
                reported elsewhere in this preamble. However, NHTSA considers the
                impacts reported in the FEIS, in addition to the other information
                presented in this preamble, as part of its decision-making process.
                ---------------------------------------------------------------------------
                 As the stringency of the alternatives increases, total U.S.
                passenger car and light truck fuel consumption for the period of 2020
                to 2050 decreases. Total light-duty vehicle fuel consumption from 2020
                to 2050 under the No Action Alternative is projected to be 3,371
                billion gasoline gallon equivalents (GGE). Light-duty vehicle fuel
                consumption from 2020 to 2050 under the action alternatives is
                projected to range from 3,598 billion GGE under Alternative 1 to 3,456
                billion gallons GGE under Alternative 7. Under the Alternative 3,
                light-duty vehicle fuel consumption from 2020 to 2050 is projected to
                be 3,571 GGE. All of the action alternatives would increase fuel
                consumption compared to the No Action Alternative, with fuel
                consumption increases that range from 226 billion GGE under Alternative
                1 to 85 billion GGE under Alternative 7.
                 The relationship between stringency and air pollutant emissions is
                less straightforward, reflecting the complex interactions among the
                tailpipe emissions rates of the various vehicle types, the technologies
                assumed to be incorporated by manufacturers in response to the CAFE
                standards, upstream emissions rates, the relative proportions of
                gasoline and diesel in total fuel consumption, and changes in VMT from
                the rebound effect. In general, emissions of criteria and toxic air
                pollutants increase across all action alternatives, with some
                exceptions. Further, the action alternatives would result in increased
                incidence of PM2.5-related adverse health impacts (including
                increased incidences of premature mortality, acute bronchitis,
                respiratory emergency room visits, and work-loss days) due to the
                emissions increases.\2897\
                ---------------------------------------------------------------------------
                 \2897\ As discussed in Section X.E.1, NHTSA also performed a
                national-scale photochemical air quality modeling and health benefit
                assessment for the FEIS, which is included as Appendix E. This
                analysis affirms the estimates that appeared in the DEIS and
                explains conclusions that may be drawn from the FEIS air quality
                discussion.
                ---------------------------------------------------------------------------
                 For CO (in 2025), NOX (in 2025), and SO2,
                emissions generally decrease under the action alternatives compared to
                the No Action Alternative. For CO in 2025, the largest decrease occurs
                under Alternative 1 and the emissions decreases get smaller from
                Alternative 1 through Alternative 7. For NOX in 2025, the
                largest decrease occurs under Alternative 6. For SO2 in
                2025, the largest decrease occurs under Alternative 6; however,
                SO2 emissions under Alternative 7 are greater than under the
                No Action Alternative. For SO2 in 2035, the largest decrease
                occurs under Alternative 2. For SO2 in 2050, the largest
                decrease occurs under Alternative 1 and the emissions decreases get
                smaller from Alternative 1 through Alternative 7. Across all criteria
                pollutants, action alternatives, and analysis years, the smallest
                decrease in emissions is less than 0.1 percent and occurs for
                NOX under Alternative 7 in 2025; the largest decrease is 12
                percent and occurs for SO2 under Alternative 2 in 2050.
                 For CO (in 2035 and 2050), NOX (in 2035 and 2050),
                PM2.5, and VOCs, emissions show increases across action
                alternatives compared to the No Action Alternative, with the largest
                increases occurring under Alternative 1 (except CO in 2035, for which
                the largest increase occurs under Alternative 4). The emissions
                increases get smaller from Alternative 1 through Alternative 7.
                Exceptions to this trend are for PM2.5 and VOCs in 2025,
                which show the smallest emissions increase under Alternative 6. Across
                all criteria pollutants, action alternatives, and analysis years, the
                smallest increase in emissions is 0.1 percent and occurs for
                SO2 under Alternative 7 in 2025; the largest increase is 12
                percent and occurs for VOCs under Alternative 1 in 2050.
                 Under each action alternative in 2025 compared to the No Action
                Alternative, decreases in emissions would occur for all toxic air
                pollutants except for DPM, for which emissions would increase by as
                much as 2 percent. For 2025, the largest relative decreases in
                emissions would occur for 1,3,-butadiene, for which emissions would
                decrease by as much as 0.5 percent. Percentage reductions in emissions
                of acetaldehyde, acrolein, benzene, and formaldehyde would be less.
                Under each action alternative in 2035 and 2050 compared to the No
                Action Alternative, increases in emissions would occur for all toxic
                air pollutants. The largest relative increases in emissions would occur
                for DPM, for which emissions would increase by as much as 9 percent.
                Percentage increases in emissions of acetaldehyde, acrolein, benzene,
                1,3,-butadiene, and formaldehyde would be less.
                 In addition, the action alternatives would result in increased
                incidence of PM2.5-related adverse health impacts due to the
                emissions increases. Increases in adverse health outcomes include
                increased incidences of premature mortality, acute bronchitis,
                respiratory emergency room visits, and work-loss days. In 2025 and
                2035, all action alternatives except for Alternative 6 would result in
                increased adverse health impacts nationwide compared to the No Action
                Alternative as a result of increases in emissions of NOX,
                PM2.5, and DPM. The increases in adverse health impacts are
                largest for the least stringent alternative (Alternative 1). The
                increases get smaller from Alternative 1 to Alternative 4, get larger
                from Alternative 4 to Alternative 5, then smaller from Alternative 5 to
                Alternative 6, and larger again from Alternative 6 to Alternative 7. In
                2050, all action alternatives would result in decreased adverse health
                impacts nationwide compared to the No Action Alternative as a result of
                decreases in emissions of SOX. The decreases in adverse
                health impacts get smaller from Alternative 1 to Alternative 7.
                 The action alternatives would increase U.S. passenger car and light
                truck fuel consumption and CO2 emissions compared with the
                No Action
                [[Page 25163]]
                Alternative, resulting in minor increases to the anticipated increases
                in global CO2 concentrations, temperature, precipitation,
                and sea level, and minor decreases in ocean pH that would otherwise
                occur, as described below. They could also, to a small degree, increase
                the impacts and risks of climate change. Uncertainty exists regarding
                the magnitude of impact on these climate variables, as well as to the
                impacts and risks of climate change. Still, the impacts of the action
                alternatives on global mean surface temperature, precipitation, sea
                level, and ocean pH would be extremely small in relation to global
                emissions trajectories. This is because of the global and multi-
                sectoral nature of climate change. These effects would be small, would
                occur on a global scale, and would not disproportionately affect the
                United States.
                 According to the FEIS, passenger cars and light trucks are
                projected to emit 85,900 million metric tons of carbon dioxide
                (MMTCO2) from 2021 through 2100 under the No Action
                Alternative. Alternative 1 would increase these emissions by 10 percent
                through 2100 (approximately 8,800 MMTCO2). Alternative 7
                would increase these emissions by 4 percent through 2100 (approximately
                3,100 MMTCO2). Emissions increases would be highest under
                Alternative 1 and would decrease across the action alternatives, with
                emissions being the lowest under the No Action Alternative.
                 In the FEIS, NHTSA presented two different analyses based on these
                emissions changes to illustrate potential impacts to certain climate
                variables. In the first analysis, to represent the direct and indirect
                impacts of this action, NHTSA used the Global Change Assessment Model
                (GCAM) Reference scenario (i.e., future global emissions assuming no
                additional climate policy [``business-as-usual'']) to represent the
                reference case emissions scenario. Under that analysis, total global
                CO2 emissions from all sources are projected to be 4,950,865
                MMTCO2 under the No Action Alternative from 2021 through
                2100, which means that the action alternatives are expected to increase
                global CO2 emissions between 0.06 (Alternative 7) and 0.17
                (Alternative 1) percent by 2100. The estimated CO2
                concentrations in the atmosphere for 2100 would range from 789.89 parts
                per million (ppm) under Alternative 1 to approximately 789.11 ppm under
                the No Action Alternative, indicating a maximum atmospheric
                CO2 increase of approximately 0.78 ppm compared to the No
                Action Alternative.
                 Changes in CO2 emissions translate to changes in global
                mean surface temperature, sea levels, global mean precipitation, and
                ocean pH, among other things. Under the first analysis, global mean
                surface temperature is projected to increase by approximately
                3.48[deg]C (6.27 [deg]F) under the No Action Alternative by 2100.
                Implementing the lowest-emissions action alternative (Alternative 7)
                would increase this projected temperature rise by 0.001[deg]C (0.002
                [deg]F), while implementing the highest-emissions alternative
                (Alternative 1) would increase projected temperature rise by
                0.003[deg]C (0.005 [deg]F). Projected sea-level rise in 2100 ranges
                from a low of 76.28 centimeters (30.03 inches) under the No Action
                Alternative to a high of 76.35 centimeters (30.06 inches) under
                Alternative 1. Alternative 1 would result in an increase in sea level
                equal to 0.07 centimeter (0.03 inch) by 2100 compared with the level
                projected under the No Action Alternative, compared to an increase
                under Alternative 7 of 0.02 centimeter (0.001 inch) compared with the
                No Action Alternative. Global mean precipitation is anticipated to
                increase by 5.85 percent by 2100 under the No Action Alternative. Under
                the action alternatives, this increase in precipitation would be
                increased further by 0.01 percent. Finally, ocean pH in 2100 is
                anticipated to be 8.2715 under Alternative 7, about 0.0001 less than
                the No Action Alternative. Under Alternative 1, ocean pH in 2100 would
                be 8.2712, or 0.0004 less than the No Action Alternative.
                 In the second analysis, NHTSA used the GCAM6.0 scenario instead of
                the default scenario to represent the reference case emissions
                scenario. The GCAM6.0 scenario assumes a moderate level of global GHG
                reductions and corresponds to stabilization, by 2100, of total
                radiative forcing and associated CO2 concentrations at
                roughly 678 ppm. By assuming a moderate level of global GHG reduction,
                NHTSA attempts to capture the cumulative impacts of this action (i.e.,
                the impact on the environment which results from the incremental impact
                of the action when added to other past, present, and reasonably
                foreseeable future actions). In the FEIS, NHTSA documented a number of
                domestic and global actions that indicate that a moderate reduction in
                the growth rate of global GHG emissions is reasonably foreseeable in
                the future.
                 Under the second analysis, compared with projected total global
                CO2 emissions of 4,044,005 MMTCO2 from all
                sources from 2021 to 2100, the incremental impact of this rulemaking is
                expected to increase global CO2 emissions between 0.08
                (Alternative 7) and 0.22 (Alternative 1) percent by 2100. Estimated
                atmospheric CO2 concentrations in 2100 range from a low of
                687.3 ppm under the No Action Alternative to a high of 688.04 ppm under
                Alternative 1. Alternative 7, the lowest CO2 emissions
                alternative, would result in CO2 concentrations of 687.55
                ppm, an increase of 0.26 ppm compared with the No Action Alternative.
                Global mean surface temperature increases for the action alternatives
                compared with the No Action Alternative in 2100 range from a low of
                0.001[deg]C (0.002 [deg]F) under Alternative 7 to a high of 0.004[deg]C
                (0.007 [deg]F) under Alternative 1. Global mean precipitation is
                anticipated to increase by 4.77 percent by 2100 under the No Action
                Alternative. Under the action alternatives, this increase in
                precipitation would be increased further by 0.01 percent. Projected
                sea-level rise in 2100 ranges from a low of 70.22 centimeters (27.65
                inches) under the No Action Alternative to a high of 70.30 centimeters
                (27.68 inches) under Alternative 1, indicating a maximum increase of
                sea-level rise of 0.07 centimeter (0.03 inch) by 2100. Sea-level rise
                under Alternative 7 would be 70.25 centimeters (27.66 inches), a 0.03
                centimeter (0.01-inch) increase compared to the No Action Alternative.
                Ocean pH in 2100 is anticipated to be 8.2721 under Alternative 7, about
                0.0001 less than the No Action Alternative. Under Alternative 1, ocean
                pH in 2100 would be 8.2719, or 0.0004 less than the No Action
                Alternative.
                 For several other resources, NHTSA is unable to provide a
                quantitative measurement of potential impacts. Instead, the FEIS
                presents a qualitative discussion on potential impacts. In most cases,
                NHTSA presents the findings of a literature review of scientific
                studies, such as in Chapter 6, where NHTSA provides a literature
                synthesis focusing on existing credible scientific information to
                evaluate the most significant lifecycle environmental impacts from some
                of the fuels, materials, and technologies that may be used to comply
                with the alternatives. In Chapter 7, NHTSA discusses land use and
                development, hazardous materials and regulated waste, historical and
                cultural resources, noise, and environmental justice. Finally, in
                Chapter 8, NHTSA discusses cumulative impacts related to energy, air
                quality, and climate change, and provides a literature synthesis of the
                impacts on key natural and human resources of changes in climate change
                variables. In these chapters, NHTSA concludes that impacts would be
                proportional to changes in emissions that would result
                [[Page 25164]]
                under the alternatives. As a result, among the action alternatives,
                Alternative 1 would have the highest impact on these resources while
                Alternative 7 would have the lowest.
                 Based on the foregoing, NHTSA concludes from the FEIS that the No
                Action Alternative is the overall environmentally preferable
                alternative because, assuming full compliance were achieved regardless
                of the agency's assessment of the costs to industry and society, it
                would result in the largest reductions in fuel use and CO2
                emissions among the alternatives considered. In addition, the No Action
                Alternative would result in the lowest overall emissions levels of
                criteria air pollutants (with the exception of sulfur dioxide) and of
                the toxic air pollutants studied by NHTSA. Impacts on other resources
                (especially those described qualitatively in the FEIS) would be
                proportional to the impacts on fuel use and emissions, as further
                described in the FEIS, with the No Action Alternative expected to have
                the fewest negative impacts.\2898\ Although the CEQ regulations require
                NHTSA to identify the environmentally preferable alternative,\2899\ the
                agency need not adopt it, as described above. The following section
                (Section VIII.B.4) explains how NHTSA balanced the relevant factors to
                determine which alternative represented the maximum feasible standards,
                including why NHTSA does not believe that the environmentally
                preferable alternative is maximum feasible.
                ---------------------------------------------------------------------------
                 \2898\ Among the action alternatives considered, Alternative 7
                would be the environmentally preferable alternative, as it is
                closest in stringency to the No Action Alternative.
                 \2899\ 40 CFR 1505.2(b).
                ---------------------------------------------------------------------------
                4. Evaluating the EPCA Factors and Other Considerations To Arrive at
                the Proposed Standards
                 As discussed in this section, NHTSA is required to consider four
                enumerated factors when establishing maximum feasible CAFE standards
                under 49 U.S.C. chapter 329: ``technological feasibility, economic
                practicability, the effect of other motor vehicle standards of the
                Government on fuel economy, and the need of the United States to
                conserve energy.'' \2900\ For this final rule, NHTSA has considered a
                wide range of potential CAFE standards (Baseline/No Action Alternative
                and Alternatives 1 through 7), ranging from the augural standards set
                forth in 2012 (Baseline/No Action Alternative), through a number of
                less stringent alternatives, including the proposed preferred
                alternative (Alternative 1, 0 percent per year stringency improvement)
                and what has been chosen as the final standards (Alternative 3, 1.5
                percent per year stringency improvement). NHTSA has determined that
                Alternative 3, which would increase the stringency of the MY 2020
                standards by 1.5 percent per year for both passenger cars and light
                trucks from MY 2021 through 2026, represents the maximum feasible CAFE
                standards under 49 U.S.C. 39202. In addition to technological
                feasibility, economic practicability, the effects of other motor
                vehicle standards of the Government on fuel economy, and the need of
                the United States to conserve energy, NHTSA has also considered the
                impact of the standards on safety and the environment.
                ---------------------------------------------------------------------------
                 \2900\ 49 U.S.C. 32902(f).
                ---------------------------------------------------------------------------
                How did the Agency balance the factors for the NPRM?
                 In the NPRM, NHTSA began its discussion of the tentative balancing
                of factors by explaining that ``NHTSA well recognizes that the decision
                it proposes to make in today's NPRM is different from the one made in
                the 2012 final rule that established standards for MY 2021 and
                identified ``augural'' standard levels for MYs 2022-2025. Not only do
                we believe that the facts before us have changed, but we believe that
                those facts have changed sufficiently that the balancing of the EPCA
                factors and other considerations must also change. The standards we are
                proposing today reflect that balancing.'' \2901\ NHTSA highlights this
                discussion at the outset in response to the number of commenters who
                claimed that NHTSA had not acknowledged or explained in the NPRM how or
                why the proposal was different from past work or policy decisions.
                ---------------------------------------------------------------------------
                 \2901\ 83 FR at 43213.
                ---------------------------------------------------------------------------
                 The NPRM balancing discussion went on to explore the definition of
                ``to conserve'' in the context of what ``energy conservation'' and
                ``the need of the U.S. to conserve energy'' should be interpreted to
                mean, in recognition of the major structural changes in global oil
                markets since EPCA was originally passed, and even since the 2012 final
                rule that set forth the augural standards. NHTSA examined these changes
                from both a demand perspective and a supply perspective. On the demand
                side, U.S. demand and global demand have both changed over time. The
                NPRM discussed the fact that the U.S. consumes a much smaller share of
                global oil output than it did at the CAFE program's outset, both
                because U.S. fleet fuel economy has improved, and because other
                countries that were not major petroleum consumers in the 1970s have
                rapidly increased their share of consumption, and continue to do so. A
                more globalized market means that risk of price spikes is spread
                around--making the U.S. in particular less likely to bear a
                disproportionate burden of price spikes. The NPRM also discussed the
                decreasing energy intensity of the U.S. economy over time and the
                improving balance of payments in petroleum, including the likelihood
                that the U.S. is poised to become a net petroleum exporter in the near
                future. Related to the decreasing energy intensity of the U.S. economy,
                on the demand side, the NPRM discussed the proliferation of fuel-
                efficient vehicle options in the market in response to CAFE increases
                over time, and the fact that consumers who wish to purchase more fuel
                efficient vehicles have largely done so, and may continue to do so over
                time if they wish.
                 On the supply side, the NPRM explained, vast increases in U.S.
                petroleum production, largely from shale formations, have introduced a
                major new stable supply into the global market. Shale oil production
                costs may be higher than the cost (for example, to OPEC members) to
                produce traditional oil, but that itself acts as a lever on global
                prices. Prices of goods like oil are affected by demand and supply--
                given that global demand trends increase relatively steadily, if OPEC
                States want to increase revenues by selling more of the total oil
                consumed globally, they have to try to control global supply volume by
                controlling production volumes (to avoid shale production increasing in
                response to higher prices). In short, the higher global prices trend,
                the more U.S. shale production increases in response, and as supply
                increases, prices fall. The NPRM discussed the responsiveness of U.S.
                shale production and suggested it could be higher than traditional
                producers in some instances. Traditional oil producers seeking to
                maintain market share have a new incentive to keep prices below a
                certain threshold, and U.S supply helps to buffer the impact of
                geopolitical events. The NPRM looked at then-current EIA oil price
                forecasts, under which U.S. gasoline prices were not forecast to exceed
                $4/gallon through 2050, and acknowledged that while price shocks could
                still occur, NHTSA tentatively concluded that from the supply side, it
                is possible that the oil market conditions that created the price
                shocks in the 1970s may no longer exist.
                 In light of these changes in global oil markets, the NPRM
                tentatively concluded that many aspects of the need of the U.S. to
                conserve energy had
                [[Page 25165]]
                improved enough over time to merit further consideration of what the
                need of the United States is to conserve oil today and going forward.
                With regard to environmental considerations, the NPRM returned to the
                definition of ``to conserve'' and suggested that differences of
                thousandths of a degree Celsius in 2100 resulting from higher levels of
                carbon dioxide emissions under the proposal as compared to the augural
                standards might not rise to the level of ``wasteful,'' given the other
                considerations discussed. With regard to consumer costs, the NPRM
                discussed the interplay of oil market conditions with prior arguments
                about consumer ``myopia'' with regard to the benefits of fuel savings,
                and tentatively concluded that U.S. consumers may be valuing fuel
                savings appropriately and purchasing the vehicles they want to
                purchase--i.e., that using CAFE standards as a tool to compel consumers
                to save money may not be necessary.
                 Given the discussion above, NHTSA tentatively concluded that the
                need of the U.S. to conserve energy may no longer function as assumed
                in previous considerations of what CAFE standards would be maximum
                feasible. In that discussion, NHTSA stated that the overall risks
                associated with the need of the U.S. to conserve oil have entered a new
                paradigm with the risks substantially lower today and projected into
                the future than when CAFE standards were first issued and in the recent
                past. NHTSA explained that the effectiveness of CAFE standards in
                reducing the demand for fuel combined with the increase in domestic oil
                production have contributed significantly to the current situation and
                outlook for the near- and mid-term future. NHTSA tentatively concluded
                that the world has changed, and the need of the U.S. to conserve
                energy, at least in the context of the CAFE program, has also changed.
                 Of two other factors under 32902(g), the NPRM explained that the
                changes were perhaps less significant. NHTSA suggested that all of the
                alternatives appear as though they could narrowly be considered
                technologically feasible, in that they could be achieved based on the
                existence or the projected future existence of technologies that could
                be incorporated on future vehicles. With regard to the effect of other
                motor vehicle standards of the Government on fuel economy, the NPRM
                explained that it was similarly not heavily limiting during this
                rulemaking time frame. The NPRM analysis projected that neither safety
                standards nor Tier 3 compliance obligations appeared likely to make it
                significantly harder for industry to comply with more stringent CAFE
                standards, and that EPA's CO2 standards should have no
                greater effect on difficulty in meeting CAFE standards than already
                existed.
                 For economic practicability, the NPRM considered the traditional
                definition used by the agency, and expressed concern that all of the
                alternatives considered in the proposal could raise economic
                practicability concerns. NHTSA stated that it believed there could be
                potential for unreasonable elimination of consumer choice, loss of U.S.
                jobs, and a number of adverse economic consequences under nearly all if
                not all of the regulatory alternatives considered in the NPRM. NHTSA
                explored consumer choice issues given a foreseeable future of
                relatively low fuel prices and the likelihood that more stringent CAFE
                standards could cause automakers to add technology to new vehicles that
                consumers do not want, or prevent the addition of technology to new
                vehicles that consumers do want, and suggested that there could be risk
                that such elimination of consumer choice could be unreasonable. NHTSA
                explained its assumption, based on repeated manufacturer input, that
                fuel-saving technologies that paid for themselves within 2.5 years
                would be added regardless of CAFE stringency, meaning that the power of
                CAFE standards (by themselves) to compel fuel savings was reduced.
                NHTSA suggested that requiring more technology to be added than
                consumers were willing to pay for could have dampening effects on
                vehicle sales, particularly given forecasted relatively low gas prices,
                increasing the likelihood of automaker non-compliance with more
                stringent standards due to difficulty in selling higher-fuel-economy
                models. NHTSA examined the levels of electrification necessary to meet
                the various regulatory alternatives evaluated in the NPRM and compared
                them with information about consumers' willingness to purchase vehicles
                with these technologies and even to spend money on fuel economy
                improvements generally. NHTSA suggested that if the market for higher
                fuel-economy vehicles exists and is already possibly saturated,
                increasing fuel economy requirements could create economic
                practicability concerns by affecting sales and consumer choice.
                 NHTSA recognized that automakers cross-subsidize regulation-driven
                cost increases and expressed concern about their ability to do that
                under sustained, ongoing increases over many years, and the
                corresponding concern that continued cross-subsidizing could create
                affordability problems for lower-income consumers if manufacturers pass
                costs forward to consumers more broadly rather than concentrating them
                in high-volume, higher-profit vehicles. NHTSA suggested that higher
                vehicle prices and monthly vehicle payments could outweigh, for at
                least some new vehicle purchasers, the benefit of fuel savings, because
                vehicle payments are fixed costs and fuel costs may be less fixed.
                NHTSA expressed concern that as vehicles get more expensive in response
                to higher CAFE standards, it will become more and more difficult for
                finance companies and dealers to continue creating loan terms that keep
                monthly payments low and do not result in consumers' still owing
                significant amounts of money on the vehicle by the time they can be
                expected to be ready for a new vehicle. This situation may imply a
                bubble in new vehicle sales, the effects of which could fall
                disproportionately on new and low-income buyers. NHTSA suggested that
                these effects could impact both fleet-wide safety (by slowing fleet
                turnover) and consumer choice. The NPRM also expressed concern that the
                sales and employment analyses were unable to capture (1) the risk that
                manufacturers and dealers may not be able to continue keeping monthly
                new vehicle payments low, or (2) the risk that manufacturing could
                shift overseas as manufacturing costs rise.
                 NHTSA also examined the net benefits of the various regulatory
                alternatives, and noted that the analysis showed that consumers recoup
                only a portion of the costs associated with increasing stringency under
                all of the alternatives, because the fuel savings resulting from each
                of the alternatives was substantially less than the costs associated
                with the alternative, meaning that net savings for consumers improved
                as stringency decreased. NHTSA explained that it recognized that this
                was a significantly different analytical result from the 2012 rule,
                which showed the opposite trend, and explained that the result was
                different because the facts and analysis underlying the result were
                also different, and enumerated the noteworthy differences, such as
                payback assumptions; fleet composition; what levels of technologies had
                already been applied; the costs and effectiveness values for some of
                those technologies; fuel price forecasts; the value of the rebound
                effect; the value of the social cost of carbon; accounting for price
                impacts on fleet turnover; not limiting mass reduction to only the
                largest vehicles; and the value of a statistical life having increased.
                NHTSA explained that all of these changes, together, meant
                [[Page 25166]]
                that the standards under any of the regulatory alternatives (compared
                to the preferred alternative) were more expensive and had lower
                benefits than if they had been calculated using the inputs and
                assumptions of the 2012 analysis. This assessment, in turn, contributed
                to the agency's decision to reevaluate what standards might be maximum
                feasible in the model years covered by the rulemaking. NHTSA explained
                that it had thus both relied on new facts and circumstances in
                developing the proposal and reasonably rejected prior analyses relied
                on in the 2012 final rule.\2902\
                ---------------------------------------------------------------------------
                 \2902\ See FCC v. Fox Television Stations, 556 U.S. at 514-515;
                see also NAHB v. EPA, 682 F.3d 1032 (D.C. Cir. 2012).
                ---------------------------------------------------------------------------
                 NHTSA then considered that ``maximum feasible'' may change over
                time as the agency assessed the relative importance of each factor that
                Congress requires it to consider, and tentatively concluded that
                proposing CAFE standards that hold the MY 2020 curves for passenger
                cars and light trucks constant through MY 2026 would be the maximum
                feasible standards for those fleets and would fulfill EPCA's
                overarching purpose of energy conservation in light of the facts before
                the agency and as the agency expected them to be in the rulemaking time
                frame. NHTSA recognized that this was a different interpretation from
                the 2012 final rule and explained that the context of that rulemaking
                was meaningfully different from the current context, because the facts
                had changed the importance of the need of the U.S. to conserve energy,
                and NHTSA recognized that under that circumstance, while more stringent
                standards may be possible, insofar as production-ready technology
                exists that the industry could physically employ to reach higher
                standards, it was not clear that higher standards would be economically
                practicable in light of current U.S. consumer needs to conserve energy.
                Therefore, NHTSA stated, it viewed the determination of maximum
                feasible standards as a question of the appropriateness of standards
                given that their need--either from the societal-benefits perspective in
                terms of risk associated with fuel price shocks or other related
                catastrophes, or from the private-benefits perspective in terms of
                consumer willingness to purchase new vehicles with expensive
                technologies that may allow them to save money on future fuel
                purchases--seems likely to remain low for the foreseeable future. NHTSA
                also considered the effects of the standards on highway safety and
                expressed concern that because more stringent standards could depress
                sales and slow fleet turnover, and because higher fuel economy leads to
                more driving and more exposure to crash risk, all regulatory
                alternatives would improve safety as compared to the augural standards.
                (b) What comments did NHTSA receive regarding how it balanced the
                factors in the NPRM?
                 In addition to comments on each of the factors NHTSA considered
                discussed above, comments also were received on how NHTSA should
                balance these factors in determining the maximum feasible final
                standards. Hundreds of thousands of comments addressed stringency and,
                thus, the agency's evaluation of what standards were maximum feasible.
                Most of those focused on the augural standards: Many individual
                commenters supported reducing the stringency of the standards from
                augural levels--some citing estimates of cost, and some citing concerns
                about consumer choice. Many comments by other individual commenters
                supported retaining stringency at augural levels or increasing
                stringency beyond that level--generally citing concerns about climate
                change and increased fuel costs under less stringent standards. A few
                commenters, like CEI, expressly supported the proposal, and even
                suggested that stringency should be decreased further. Many other
                commenters, including environmental and consumer groups, health
                advocacy organizations, and a number of State organizations, argued
                that the proposal was flawed and/or that the augural standards should
                be finalized because more stringent standards help to reduce climate
                change and address other air quality issues.\2903\ The Congressional
                Tri-Caucus commenters supported maintaining the augural standards,
                stating that they contribute to employment and protect low income
                communities and communities of color.\2904\
                ---------------------------------------------------------------------------
                 \2903\ See, e.g., Harvard Environmental Law Clinic, EPA-HQ-OAR-
                2018-0283-5486, at 1; University of San Francisco graduate students,
                EPA-HQ-OAR-2018-0283-2676, at 1-2; Vanderbilt student organizations,
                EPA-HQ-OAR-2018-0283-4189, at 1-2; Blue Planet Foundation, EPA-HQ-
                OAR-2018-0283-4207, at 1; Green Energy Institute (Lewis and Clark
                Law School), et al., EPA-HQ-OAR-2018-0283-4193, at 1-3; CBD et al.,
                NHTSA-2018-0067-12057, at 2; NESCAUM, NHTSA-2018-0067-11691, at 3-4.
                 \2904\ Congressional Tri-Caucus, NHTSA-2018-0067-1424, at 1.
                ---------------------------------------------------------------------------
                 The Alliance and Global Automakers both supported final standards
                that increased in stringency year over year. The Alliance stated that
                it could support stringency increases between 0 percent per year and 2-
                3 percent per year ``along with the inclusion of appropriate
                flexibilities.'' \2905\ Global stated that increases should be
                ``meaningful'' \2906\ and suggested that ``[i]n order for the U.S. auto
                industry to remain competitive and continue to export vehicles to the
                rest of the world, industry is best served by a reasonable, steady ramp
                rate that accounts for investments made and the global nature of the
                market. Steady increases allow for long-term planning and create an
                environment of security that fosters ongoing investment in vehicle
                technology and consumer confidence in purchasing new vehicles. It also
                provides a level playing field upon which automakers can compete.''
                \2907\ Toyota made similar points, and argued that while the standards
                set in 2012 are beyond maximum feasible today, the ``statutes support
                an adjustment to those standards that reflect the realities of the
                market, consumer choice, and the pace of technological advancement
                acceptable to consumers.'' \2908\ Mazda stated that it supported
                ``increasing requirements for fuel efficiency. . ., if they are
                sensible and achievable under changing market conditions.'' \2909\
                ---------------------------------------------------------------------------
                 \2905\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 8.
                 \2906\ Global, NHTSA-2018-0067-12032, at 3.
                 \2907\ Global, NHTSA-2018-0067-12032, Attachment A, at A-11.
                 \2908\ Toyota, NHTSA-2018-0067-12150, at 31.
                 \2909\ Mazda, NHTSA-2018-0067-11727, at 2.
                ---------------------------------------------------------------------------
                 NADA commented that it was willing to support standards that
                increased in stringency (i.e., more stringent than the proposal) if
                they were economically practicable and technologically feasible, based
                on the evidence before the agencies; if they ensured consumer choice
                and ``the strongest possible rate of fleet turnover;'' and if passenger
                car and light truck standards increased at the same rate.\2910\ The
                Alliance for Vehicle Efficiency (AVE) argued that compliance shortfalls
                are evidence that the current rate of stringency increase is beyond
                maximum feasible, and that the assumptions that enabled those rates to
                be chosen ``are no longer feasible based on consumer adoption.'' \2911\
                AVE suggested that a rate of increase of 2.5 percent per year for both
                cars and trucks, retroactively imposed beginning in MY 2018, would be
                feasible given sufficient flexibilities.\2912\
                ---------------------------------------------------------------------------
                 \2910\ NADA, NHTSA-2018-0067-12064, at 12.
                 \2911\ AVE, NHTSA-2018-0067-11696, at 6-8.
                 \2912\ Id., at 10.
                ---------------------------------------------------------------------------
                 NADA also stressed the importance of flexibilities as a compliance
                tool for meeting standards that increase faster
                [[Page 25167]]
                than the proposal.\2913\ The Minnesota agencies supported maintaining
                standards at the augural levels, commenting that automakers has simply
                ``requested additional flexibility . . ., not a wholesale rollback of
                the standards,'' and suggesting that additional flexibilities would
                enable augural levels.\2914\ IPI disagreed with the suggestion in the
                NPRM that heavy automaker reliance on credits for compliance might
                indicate that standards were beyond maximum feasible, arguing that
                automakers must be either using credits about to expire, or counting on
                future standards being cheaper to meet due to rising consumer demand
                for fuel economy, technology costs decreasing over time, and the cost-
                effectiveness of EPA's EV multiplier incentive.\2915\
                ---------------------------------------------------------------------------
                 \2913\ NADA, NHTSA-2018-0067-12064, at 12.
                 \2914\ 1 Minnesota agencies, NHTSA-2018-0067-11706, at 6-7.
                 \2915\ IPI, NHTSA-2018-0067-12213, Appendix, at 25-26.
                ---------------------------------------------------------------------------
                 With regard to analysis of costs and benefits, IPI argued that the
                final rule needed, like the 2012 rule, to cite costs and benefit
                expressly in discussing balancing of statutory factors, but with a
                ``proper'' accounting of costs and benefits. IPI claimed that in the
                NPRM the factors were balanced ``in a way that conflicts with the . .
                .controlling statute[ ] and weighed . . .without regard for the
                accuracy of the accompanying cost-benefit analysis.'' \2916\ IPI stated
                that ``. . . the agencies' analysis produced biased and irrational
                results at each of the steps in that causal chain, leading to a
                Proposed Rule that vastly overstates the benefits of the rollback and
                understates the benefits society foregoes with the rollback,'' and that
                ``[a] full and balanced analysis of all the costs and benefits that the
                agencies are charged with considering would reveal--as the midterm
                review recently confirmed--that the baseline standards will deliver
                massive net social benefits, and the proposed rollback is
                unjustified.'' \2917\
                ---------------------------------------------------------------------------
                 \2916\ Id.
                 \2917\ IPI, NHTSA-2018-0067-12213, Appendix, at 1-2.
                ---------------------------------------------------------------------------
                 With regard to net benefits, the States and Cities commenters
                stated that prior analyses had concluded that the net benefits of the
                augural standards were extremely high,\2918\ while the Alliance stated
                that ``[t]he NERA-Trinity Assessment confirms the Agencies' findings
                that Alternatives 1, 5, and 8 result in increased net benefits relative
                to the no-action alternative augural CAFE standards.''\2919\ Michalek
                and Whitefoot commented that ``maximizing net benefits is among the
                most important factors to consider in policy selection because it is an
                effort to weigh a variety of policy implications on a common basis and
                seek decisions that are beneficial to society overall,'' but also
                cautioned that estimates are inherently uncertain and should be
                transparent and clearly justified; that sensitivity analysis is
                necessary; that a net benefits analysis will not be able to capture
                distributional effects or changes in behavior caused by the policy; and
                that ``it is not clear that there is necessarily any relationship
                between MNB and setting the `maximum feasible' criteria while
                considering `economic practicability.' ''\2920\ IPI disagreed with the
                NPRM's suggestion that feasibility concerns could lead NHTSA not to
                maximize net benefits, stating that ``if a standard were truly not
                feasible, then its costs would be prohibitively high, and a full and
                fair cost-benefit analysis would reflect that.'' \2921\
                ---------------------------------------------------------------------------
                 \2918\ States and Cities, NHTSA-2018-0067-11735, Detailed
                Comments, at 6.
                 \2919\ Alliance, NHTSA-2018-0067-12073, Full Comment Set, at 13.
                 \2920\ Michalek and Whitefoot, NHTSA-2018-0067-11903, at 14-15.
                 \2921\ IPI, NHTSA-2018-0067-12213, Appendix, at 11.
                ---------------------------------------------------------------------------
                 CARB argued that ``[a]lthough EPCA provides NHTSA with some
                discretion with respect to balancing the four factors, that discretion
                is nevertheless constrained by EPCA's overriding mandate of conserving
                energy.'' \2922\ CARB further stated that EPCA ``envision[s] the
                promulgation of increasingly stringent requirements to ensure the
                continued reductions of both emissions and fuel consumption from motor
                vehicles.'' \2923\ Michalek and Whitefoot similarly commented that the
                requirement that standards be maximum feasible necessarily means that
                stringency must increase over time, because technology capabilities and
                cost are constantly improving; international regulations are constantly
                increasing in stringency; and if standards are held constant,
                automakers will always exceed them.\2924\ The States and Cities
                commenters cited the CAS language from the D.C. Circuit that ``[i]t is
                axiomatic that Congress intended energy conservation to be a long term
                effort that would continue through temporary improvements in energy
                availability,'' and argued that ``[w]hile NHTSA purports to acknowledge
                this purpose and the importance of improving fuel economy over time,
                NHTSA proposes to do the opposite: roll back fuel economy standards for
                a period of at least six years.'' \2925\ The States and Cities
                commenters further argued that NHTSA had ``departed sharply from its
                past interpretations and practice without an adequate explanation,
                often without even an acknowledgement,'' citing Fox Television, insofar
                as the 2012 final rule justification had noted that less stringent
                regulatory alternatives would have conserved less energy than the then-
                finalized standards, as compared to ``[w]ith the Proposed Rollback,
                NHTSA has radically changed positions--assuming energy conservation
                provides little, if any, benefits, for example--without explaining or
                even acknowledging this complete reversal of course.'' \2926\ The
                States and Cities commenters concluded that it was ``impermissible''
                for NHTSA to balance ``the factors in a manner that contravenes EPCA's
                central purpose of energy conservation.'' \2927\
                ---------------------------------------------------------------------------
                 \2922\ CARB, NHTSA-2018-0067-11783, Detailed Comments, at 78.
                 \2923\ Id., at 80.
                 \2924\ Michalek and Whitefoot, NHTSA-2018-0067-11903, at 3-4.
                 \2925\ States and Cities, NHTSA-2018-0067-11735, Detailed
                Comments, at 64-65.
                 \2926\ Id., at 65.
                 \2927\ Id.
                ---------------------------------------------------------------------------
                 ACEEE commented that NHTSA did not have discretion to assess
                whether the need of the U.S. to conserve energy was as great as when
                EPCA was first passed, arguing that ``[t]he statute does not ask for a
                determination on whether the nation needs to save energy. It assumes
                the need and directs that the need be taken into account along with
                other considerations.'' \2928\ Securing America's Energy Future
                commented that the need of the U.S. to conserve energy continued, and
                that ``[a]lthough the nation is undoubtedly more energy secure than it
                was before the start of the U.S. shale oil revolution ten years ago,''
                \2929\ ``[u]ntil the U.S. transportation sector is no longer beholden
                to oil, the country will be vulnerable to oil price volatility.
                Improving the fuel efficiency of the U.S. vehicle fleet is a valuable
                insurance policy against this volatility.'' \2930\ IPI also commented
                that fuel efficiency standards act as insurance, but against
                unpredictable future fuel prices.\2931\ IPI stated that anticipating
                relatively low future fuel prices was not an appropriate basis for
                finalizing the proposal, both because fuel costs may rise in the
                future, and also because
                [[Page 25168]]
                EPA's Final Determination ``found that that even with the lowest prices
                projected in AEO 2016 of close to $2, the `lifetime fuel savings
                significantly outweigh the increased lifetime costs' of the GHG
                standards.'' \2932\ IPI further argued that ``[i]n ignoring the [FD]
                analysis, the Proposed Rule has failed to provide a `reasoned
                explanation' for dismissing the `facts and circumstances that underlay'
                the original rule, rendering its analysis arbitrary and capricious.''
                \2933\ IPI also argued that NHTSA had not adequately explained its
                ``shift since 2012 in its interpretation and application of the need to
                conserve energy factor,'' stating that ``[a]ctual fuel savings, and the
                associated benefits to consumers, the environment, and society, were at
                the heart of NHTSA's analysis of the need to conserve energy factor
                back in 2012. Now the agency ignores those conclusions from 2012 and
                relies on mistaken and inconsistent interpretations of petroleum import
                projections and the urgency of climate change to justify ignoring this
                statutory factor and giving primacy instead to economic practicability
                and safety effects. The failure to explain this shift in approach is
                arbitrary.'' \2934\
                ---------------------------------------------------------------------------
                 \2928\ ACEEE, NHTSA-2018-0067-12122, main comments, at 1.
                 \2929\ Securing America's Energy Future, NHTSA-2018-0067-12172,
                at 17.
                 \2930\ Id., at 7, 8.
                 \2931\ IPI, NHTSA-2018-0067-12213, Appendix, at 31.
                 \2932\ Id., at 32.
                 \2933\ Id.
                 \2934\ Id., at 6.
                ---------------------------------------------------------------------------
                 UCS argued that the need of the United States to conserve energy is
                ``the most important of the four required factors'' according to CBD v.
                NHTSA, and claimed that ``NHTSA has manipulated the evaluation of the
                factors to produce a result that supports the preferred option in the
                NPRM.'' \2935\ The States and Cities commenters argued that it was
                ``[c]ynical. . .'' for NHTSA to justify the proposal on the basis that
                ``the oil intensity of U.S. GDP has continued to decline'' in part as a
                result of increasingly stringent CAFE standards, and on the basis that
                ``[m]anufacturers have responded to fuel economy standards and to
                consumer demand over the last decade to offer a wide array of fuel-
                efficient vehicles in different segments and with a wide array of
                features.'' \2936\
                ---------------------------------------------------------------------------
                 \2935\ UCS, NHTSA-2018-0067-12039, at 3, 7.
                 \2936\ States and Cities, NHTSA-2018-0067-11735, Detailed
                Comments, at 64-65.
                ---------------------------------------------------------------------------
                 CARB and CBD et al. argued that if NHTSA's analysis indicates that
                automakers will voluntarily exceed the standards, then the standards
                cannot be maximum feasible.\2937\ Robertson commented relatedly that
                standards should not be set below augural levels because ``Much higher
                fuel economy and reduced emissions have been achieved by several lower
                priced makes and models using hybrid technology.'' \2938\ Blue Planet
                Foundation stated that the augural standards are feasible because
                automakers have already invested in technologies, and electrification
                is projected to continue to grow cheaper over time, so that ``even the
                up-front cost of an EV will begin to reach parity with gas-powered cars
                by 2024.'' \2939\ ACEEE also cited the voluntary overcompliance in the
                NPRM analysis as evidence that there could not be diminishing returns
                from higher fuel efficiency standards, because ``the list of [cost-
                effective] technology [must] continually regenerate itself'' if
                manufacturers would continue applying it in the absence of future
                standards. Moreover, ACEEE argued, past analyses had always found
                plenty of available cost-effective technologies, and automakers would
                find a way to apply them.\2940\
                ---------------------------------------------------------------------------
                 \2937\ CARB, NHTSA-2018-0067-11873, Detailed Comments, at 84;
                CBD et al., NHTSA-2018-0067-12057, at 2.
                 \2938\ Robertson, EPA-HQ-OAR-2018-0283-0787, at 3.
                 \2939\ Blue Planet Foundation, EPA-HQ-OAR-2018-0283-4207, at 1-
                2.
                 \2940\ ACEEE, NHTSA-2018-0067-12122, main comments, at 9.
                ---------------------------------------------------------------------------
                c) How is NHTSA Balancing the Factors to Determine the Maximum Feasible
                Final CAFE Standards?
                 EPCA/EISA grants the Secretary (by delegation, NHTSA) discretion in
                how to balance the relevant statutory factors, while bearing in mind
                EPCA's overarching purpose of energy conservation. Many commenters
                cited the Ninth Circuit's language in CBD v. NHTSA that ``the
                overarching purpose of EPCA is energy conservation,'' \2941\ and the
                D.C. Circuit's language in CAS v. NHTSA that ``[i]t is axiomatic that
                Congress intended energy conservation to be a long term effort that
                would continue through temporary improvements in energy availability.''
                \2942\ NHTSA has considered those comments and those court decisions
                carefully as it made the decision set forth in the final rule. Based on
                the information before the agencies and considering carefully the
                comments received, NHTSA has determined that the preferred alternative
                identified in the proposal--amending the MY 2021 standards to match MY
                2020, and holding those standards flat through MY 2026--does not
                represent the maximum feasible standards, and that the maximum feasible
                standards for MYs 2021-2026 passenger cars and light trucks increase in
                stringency by 1.5 percent per year from the MY 2020 standards. The
                following discussion walks through NHTSA's evaluation and balancing of
                the relevant factors in light of the information before it.
                ---------------------------------------------------------------------------
                 \2941\ CBD, 508 F.3d 508, 537 (9th Cir. 2007), opinion vacated
                and superseded on denial of reh'g, 538 F.3d 1172 (9th Cir. 2008).
                 \2942\ CAS, 793 F.2d 1322, 1340 (D.C. Cir. 1986).
                ---------------------------------------------------------------------------
                (1) Need of the U.S. to Conserve Energy
                 NHTSA agrees with commenters that energy conservation remains
                important, and that changed conditions, even significantly changed
                conditions, do not obviate NHTSA's obligation to set maximum feasible
                CAFE standards as directed by Congress. Many commenters disagreed
                strongly with NHTSA's suggestion in the NPRM that increased U.S.
                petroleum production, and the U.S.'s likely imminent status as a net
                petroleum exporter, decreased the need of the U.S. to conserve energy.
                NHTSA agrees that there is still a need to conserve energy, and oil in
                particular. Like an insurance policy or a savings account, continuing
                to move the needle forward on CAFE helps position Americans better to
                weather certain types of possible future uncertainty. NHTSA believes
                that it is reasonable to be somewhat conservative about this risk, and
                thus to set CAFE standards that increase in stringency year over year
                through MY 2026.
                 That said, NHTSA believes that there are limits to how much
                uncertainty the CAFE program can mitigate--continuing to make progress
                is important, but it is also important to be transparent and realistic
                about what is being accomplished, even if NHTSA were able to set
                standards beyond levels that NHTSA considers maximum feasible. NHTSA
                also continues to believe that structural changes in global oil markets
                over the last 10 years, driven in part by changes in demand both in the
                U.S. and abroad, and in part by the significant growth in U.S.
                petroleum production, have led to a fundamental shift in the dynamics
                of global oil prices, which has in turn improved U.S. (and possibly,
                global) energy security. NHTSA believes that this shift is important to
                consider as NHTSA weighs the need of the Nation to conserve energy.
                 NHTSA acknowledges that price shocks can still happen. The large
                scale attack on Saudi Arabia's Abqaiq processing facility--the world's
                largest crude oil processing and stabilization plant--on September 14,
                2019 caused ``the largest single-day [crude oil] price increase in the
                past decade,'' of between $7 and $8, according to EIA.\2943\ The Abqaiq
                facility has a capacity to process
                [[Page 25169]]
                7 million barrels per day, or about 7 percent of global crude oil
                production capacity. By September 17, however, also according to EIA,
                ---------------------------------------------------------------------------
                 \2943\ https://www.eia.gov/todayinenergy/detail.php?id=41413.
                 Saudi Aramco reported that Abqaiq was producing 2 million
                barrels per day, and they expected its entire output capacity to be
                fully restored by the end of September. In addition, Saudi Aramco
                stated that crude oil exports to customers will continue by drawing
                on existing inventories and offering additional crude oil production
                from other fields. Tanker loading estimates from third-party data
                sources indicate that loadings at two Saudi Arabian export
                facilities were restored to the pre-attack levels. Likely driven by
                news of the expected return of the lost production capacity, both
                Brent and WTI crude oil prices fell on Tuesday, September 17.\2944\
                ---------------------------------------------------------------------------
                 \2944\ Id.
                 Thus, the largest single-day oil price increase in the past decade
                was largely resolved within a week, and assuming very roughly that
                average crude oil prices were $70/barrel in September 2019 (slightly
                higher than actual), an increase of $7/barrel would represent a 10
                percent increase as a result of the Abqaiq attack. Contrast this with
                the 1973 Arab oil embargo, which lasted for months and raised prices
                350 percent.\2945\ Saudi Arabia could have benefited, revenue-wise,
                from higher prices following the Abqaiq attack, but instead moved
                rapidly to restore production and tap reserves to control the risk of
                resulting price increases, likely recognizing that long-term sustained
                price increases would reduce their ability to control global supply
                (and thus prices, and thus their own revenues) by relying on their
                lower cost of production.\2946\ Even if the NPRM discussion was perhaps
                overconfident about the ability of U.S. shale producers to act as
                ``swing'' supply, as some commenters suggested, it seems clear from
                events that the existence of U.S. production has a stabilizing effect
                on global oil prices. This has played out in important ways in the
                first quarter of 2020, with the dissolution of the ``OPEC+'' coalition
                as Russia and Saudi Arabia compete for market share in response to U.S.
                shale production and also in the wake of global demand downturn.\2947\
                ---------------------------------------------------------------------------
                 \2945\ See Jeanne Whalen, ``Saudi Arabia's oil troubles don't
                rattle the U.S. as they used to,'' Washington Post, September 19,
                2019, available at https://www.washingtonpost.com/business/2019/09/19/saudi-arabias-oil-troubles-dont-rattle-us-like-they-used/.
                 \2946\ See, e.g., ``Dynamic Delivery: America's Evolving Oil and
                Natural Gas Transportation Infrastructure,'' National Petroleum
                Council (2019) at 18, available at: https://dynamicdelivery.npc.org/downloads.php. See also ``Oil prices plunge as Trump speech eases
                Iran fears,'' CNN, available at https://www.cnn.com/2020/01/07/business/oil-prices-iran-attack-iraq/index.html.
                 \2947\ See, e.g., EIA, ``This Week in Petroleum--OPEC shift to
                maintain market share will result in global inventory increases and
                lower prices,'' March 11, 2020, https://www.eia.gov/petroleum/weekly/; DOE, ``DOE Responds to Recent Oil Market Activity,'' March
                9, 2020, https://www.energy.gov/articles/doe-responds-recent-oil-market-activity; Reid Standish, Keith Johnson, ``No End in Sight to
                the Oil Price War Between Russia and Saudi Arabia,'' March 14, 2020,
                https://foreignpolicy.com/2020/03/14/oil-price-war-russia-saudi-arabia-no-end-production/; Alex Ward, ``The Saudi Arabia-Russia oil
                war, explained,'' March 9, 2020, https://www.vox.com/2020/3/9/21171406/coronavirus-saudi-arabia-russia-oil-war-explained.
                ---------------------------------------------------------------------------
                 Even though the effect of significant supply disruptions appears
                much lower than was the case several years ago, the analysis for this
                final rule (like the NPRM analysis) does, in fact, explicitly account
                for the possible occurrence of price shocks. The cost penalty used in
                the analysis to represent the consequences of those shocks attempts to
                quantify the negative impact on U.S. GDP created by abrupt, short-term
                increases in the world oil price. The values used in the NPRM were
                based on arguably outdated work, and commenters cited more recent
                studies of relevance in their comments on the NPRM--one of which formed
                the basis for the estimates in today's analysis. The final rule
                estimate of this cost are based on a recent study which states that
                ``[i]n recent years, the United States has become much more self-
                reliant in producing oil, and a newer economics literature suggests
                that oil demand may be more elastic and U.S. GDP may be less sensitive
                to world oil price shocks than was previously estimated. These
                developments suggest somewhat lower security costs may be associated
                with U.S. oil consumption.'' \2948\ These more recent studies concede
                that the fact that ``the world has not seen a major oil supply
                disruption since 2003,'' and that therefore ``we have no reliable
                method to quantify the effects of these disruptions,'' \2949\ but even
                the range of uncertainty suggests that the risk has decreased relative
                to prior estimates. The price shock cost estimate employed in the NPRM
                was at least twice as large as the upper bound of the range in Brown's
                new estimates, and consistently close to the upper bound of the range
                of his more conservative estimates. The approach taken today, which
                relies on median estimates in Brown's study, implies that risk is more
                properly estimated here than in the NPRM.
                ---------------------------------------------------------------------------
                 \2948\ Brown, Stephen, ``New estimates of the security costs of
                U.S. oil consumption,'' Energy Policy 113 (2018) 171-192, at 171.
                Cited in Securing America's Energy Future, NHTSA-2018-0067-12172, at
                29.
                 \2949\ Brown, at 181.
                ---------------------------------------------------------------------------
                 Commenters (Bordoff, SAFE, CARB, IPI) argued that increased U.S.
                petroleum production, which improves the stability of the global supply
                and reduces the probability of supply interruptions, does not reduce
                U.S. exposure to petroleum price shocks, which are still determined by
                the dynamics of the global market. By reducing the probability of
                supply disruptions in the global market, the U.S. does reduce its
                vulnerability to price shocks. However, to the extent that the
                vulnerability to price shocks is a function of exposure, commenters are
                correct that looming petroleum independence does not entirely insulate
                the U.S. economy from the consequences of global oil price shocks. Some
                commenters further argued that the proposed standard would leave the
                U.S. more exposed to oil price shocks, which would harm consumers.
                Basic mathematics means that a less efficient on-road fleet necessarily
                would spend more on fuel than a more efficient on-road fleet in the
                event of a sudden, unexpected, and dramatic increase in oil price. The
                suggestion in these comments, however, is that finalizing the augural
                standards would sufficiently insulate U.S. consumers from harm during
                such an event, while finalizing any other regulatory alternative would
                not. NHTSA disagrees that finalizing the augural standards, as compared
                to the standards we are finalizing, would make a meaningful difference
                in this case.
                 A continuous, but slow, price increase over several years is
                fundamentally different from the kinds of acute price shocks over which
                commenters have expressed understandable concern. Long-term price
                increases signal consumers to make investments in fuel economy, in both
                the new and used vehicle markets, and to diversify the vehicles in
                their household fleets. In a side analysis using outputs from the CAFE
                Model, the agencies examined the consequences of a gasoline price spike
                in 2030--increasing the price from $3.40/gallon to $6/gallon for eight
                months, then reverting back to $3.40/gallon.\2950\ By choosing a year
                so far in the future, the agencies consider a larger gap in fleet fuel
                efficiency than is attributable to this action. If the agencies
                increase stringency again after MY 2026, the efficiency gap between the
                on-road fleet in the final standards and baseline would be smaller than
                simulated here. This side analysis showed that even a nearly doubling
                of the fuel price, sustained for more than half a year, would result in
                less than 1 percent savings in fuel expenditures for that
                [[Page 25170]]
                year under the final standards (relative to the proposal), compared to
                about 5 percent reduction in expenditures under the augural standards.
                This demonstrates that even though finalizing the augural standards
                would mitigate American drivers' increase in fuel expenditures by more
                than the standards the agencies are finalizing today, it would only do
                so by a few percent. This is important to understanding concerns about
                differences in the amount of fuel saved under today's final standards
                versus if the augural standards were finalized, as will be discussed
                more below. And as also discussed below, NHTSA believes the augural
                standards are beyond maximum feasible at this time.
                ---------------------------------------------------------------------------
                 \2950\ Docketed in NHTSA-2018-0067.
                ---------------------------------------------------------------------------
                 Some commenters raised the possibility that the U.S. might ban
                fracking at some point in the future, and suggested that therefore the
                need of the U.S. to conserve energy could not be assumed away. NHTSA
                acknowledges that the future is uncertain. Without the supply of U.S.
                oil in the global market, NHTSA agrees that it is foreseeable that
                conditions could revert somewhat to how global oil market conditions
                were before the ramp-up in U.S. supply--i.e., that the global market as
                a whole could be somewhat less stable and thus fuel prices could be
                somewhat more prone to change unexpectedly and for longer periods.
                Pulling out of the market on the supply side means that the agencies
                would lose the ability to influence the market on that side.
                Presumably, part of the policy objective of banning fracking would be
                to accelerate a transition to a post-oil transportation system. In that
                scenario, presumably decision-makers would consider higher fuel prices
                to be an acceptable tradeoff for less driving and lower emissions. That
                said, the availability of shale oil resources does exist today, and is
                not realistically in question. And, even if the future availability of
                that capacity was realistically doubtful, any increase in fuel economy
                above current levels, like the final rule will require, will help
                somewhat to mitigate the economic pain to drivers of that event were it
                to occur, as shown above.\2951\ To the extent that current events cause
                pauses or consolidation in the shale industry's development, while that
                may lead to transitory difficulty for the shale industry, the resources
                will continue to exist, and U.S. shale will continue to be able to act
                as a lever to keep global prices from rising very high for very long.
                ---------------------------------------------------------------------------
                 \2951\ See also Letter from Alliance for Automotive Innovation,
                NADA, and MEMA to Congress, Mar. 23, 2020, available at https://www.autosinnovate.org/wp-content/uploads/2020/03/COVID-19-Letter-to-Congress-NADA-MEMA-AAI-March-23.pdf.
                ---------------------------------------------------------------------------
                 As noted above, Securing America's Energy Future commented that
                ``[a]lthough the nation is undoubtedly more energy secure than it was
                before the start of the U.S. shale oil revolution ten years ago,''
                \2952\ ``[u]ntil the U.S. transportation sector is no longer beholden
                to oil, the country will be vulnerable to oil price volatility.
                Improving the fuel efficiency of the U.S. vehicle fleet is a valuable
                insurance policy against this volatility.'' \2953\ (Emphasis added.)
                NHTSA agrees fully with this comment. Energy security concerns were the
                driving force behind the creation of the CAFE program, as discussed in
                the NPRM. U.S. energy security has improved, but the only way to
                resolve petroleum-related energy security concerns entirely would be
                for the U.S. vehicle fleet to stop using oil. And doing so would not
                avoid energy-related concerns entirely, but rather shift them away from
                petroleum (and the Middle East) and toward battery-related security
                (and lithium-, nickel-, cobalt-, and other metals-producing
                countries).\2954\
                ---------------------------------------------------------------------------
                 \2952\ Securing America's Energy Future, NHTSA-2018-0067-12172,
                at 17.
                 \2953\ Id., at 7, 8.
                 \2954\ While progress is being made on developing and improving
                domestic sources for many of the minerals necessary for battery
                development, the U.S. is still heavily dependent on imports of both
                raw materials and batteries. Regarding minerals production and
                import dependence, see Schulz, K.J., DeYoung, J.H., Jr., Seal, R.R.,
                II, and Bradley, DC, eds., Critical mineral resources of the United
                States--Economic and environmental geology and prospects for future
                supply: U.S. Geological Survey Professional Paper 1802 (see
                particularly Chapter K, p. K1-K21 on lithium), available at https://www.commerce.gov/sites/default/files/2020-01/Critical_Minerals_Strategy_Final.pdf and https://pubs.usgs.gov/pp/1802/k/pp1802k.pdf. Regarding vehicle battery supply chains, see
                Coffin, D., and J. Horowitz, ``The Supply Chain for Electric Vehicle
                Batteries,'' Journal of International Commerce and Economics,
                December 2018, available at https://www.usitc.gov/publications/332/journals/the_supply_chain_for_electric_vehicle_batteries.pdf.
                ---------------------------------------------------------------------------
                 Our relationship to the global energy market has changed
                significantly since the CAFE program was created, with most of this
                change occurring over the last decade. The United States has become
                energy independent, and is currently a net exporter of petroleum
                products. Rising world oil prices no longer only mean a financial
                burden on U.S. drivers and a wealth transfer to foreign nations. While
                rising prices continue to affect U.S. motorists, we have taken steps to
                insulate our transportation system from exogenous price shocks. CAFE
                standards (and, recently, CO2 standards) have increased the
                efficiency of new vehicles for more than a decade, and these
                increasingly efficient vehicles are still working their way into the
                on-road fleet as older models are retired. Accompanying any increase in
                the global oil price is an increase in revenue to the U.S. oil
                industry. To the extent that motorists are spending more on oil
                everywhere, the dollars spent on domestically produced petroleum
                products stay within the U.S. and additional revenue from foreign
                buyers flows into our domestic energy industry. To the extent that the
                U.S. transportation system is able to further reduce its dependence on
                petroleum in a cost-effective manner, it is sensible to do so. But in
                the current environment, in which motorized transportation is
                increasingly energy efficient and U.S. energy producers are not only
                supplying our demand but exporting petroleum products to other nations,
                the nationwide benefits of reducing petroleum consumption are
                substantially diminished.
                 There is also the opposite concern to bear in mind--that energy
                security is not just about oil becoming more expensive, but also about
                other changes in oil prices. Major fluctuations in either direction, as
                well as oil price collapse, can potentially have seriously
                destabilizing geopolitical effects. Many major oil producing countries
                (some of whom are allies) rely heavily on oil revenues for public
                revenue, and sustained losses in public revenue in certain countries
                and regions can foreseeably create new energy-related security risks,
                not only for the U.S. As the world works toward transitioning away from
                oil for transportation, keeping prices reasonably stable may best help
                that transition remain peaceful and steady. In short, energy security
                can cut both ways, and the current estimates of price shock that we
                model inherently do not account for the longer-term stabilizing effect
                of steady global oil consumption (of which the U.S. is a part) on
                global security. Steady trends in consumption can facilitate steady
                changes in production, which can facilitate a steady security
                situation.
                 NHTSA does not interpret EPCA/EISA to mean that Congress expected
                the CAFE program to take the U.S. auto fleet off of oil entirely--
                indeed, EISA renders doing so impossible because it amended EPCA to
                prohibit NHTSA from considering the fuel economy of dedicated
                alternative fuel vehicles, including electric vehicles, when setting
                maximum feasible standards. This means that standards cannot be set
                that assume increased usage of full electrification for compliance.
                Reading that prohibition together with the obligation to set maximum
                feasible standards by considering (which is hard
                [[Page 25171]]
                to do without balancing) factors like economic practicability with the
                need of the U.S. to conserve energy, NHTSA believes that Congress
                intended CAFE to try to mitigate the risk of gas lines, but not to
                shift the fleet entirely off of oil. Moreover, the EISA-added
                requirement that standards ``increase ratably'' for MYs through 2020
                ceases to apply beginning in MY 2021. While NHTSA unquestionably has
                discretion to determine that standards should continue to increase
                post-MY 2020, NHTSA does not interpret EPCA/EISA as requiring that they
                do, as long as they are maximum feasible. Several commenters suggested
                that standards that do not continue to increase, by definition, cannot
                be maximum feasible, but NHTSA believes that this interpretation does
                not account for the clear requirement that maximum feasible standards
                be determined with reference to the four statutory factors. The statute
                does not preclude an interpretation that non-increasing standards could
                be maximum feasible, depending on the facts before the agency. Neither
                does the statute preclude an interpretation that amending standards
                downward can be maximum feasible, as has occurred in the past in
                response to changes in consumer demand.\2955\
                ---------------------------------------------------------------------------
                 \2955\ See, Center for Auto Safety v. NHTSA (CAS), 793 F.2d 1322
                (D.C. Cir. 1986).
                ---------------------------------------------------------------------------
                 Nevertheless, for purposes of this final rule, NHTSA does believe
                that standards that increase in stringency are maximum feasible; the
                question remains by how much those standards should increase. While
                NHTSA agrees that CAFE standards must conserve energy, the improvement
                in energy security discussed above is entirely relevant to how much
                energy should be conserved. If the marginal improvement in energy
                security of increasing CAFE stringency from one regulatory alternative
                to another is very small, as it appears to be based on the above
                discussion, then other aspects of the need of the U.S. to conserve
                energy must be considered next to see what effect they have.
                 Consumer costs, as discussed above, is another aspect of the need
                of the U.S. to conserve energy. The final rule analysis estimates that
                all alternatives besides the baseline/augural standards would result in
                higher fuel costs for consumers than the baseline/augural standards
                would result in, as follows:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.735
                 A number of commenters stated that the 2012 rulemaking had relied
                on fuel savings as part of its justification, and argued that the NPRM
                had not adequately grappled with the fact that the proposal would have
                cost consumers more in fuel expenditures than if NHTSA finalized the
                augural standards. In fact, NHTSA explained in the NPRM that while fuel
                costs would be higher, NHTSA believed that the higher upfront (and
                ongoing, if financed) costs of new vehicles and associated taxes and
                registration fees--as well as the opportunity cost associated with
                those upfront costs--would outweigh, for many consumers, the additional
                fuel costs that would be incurred if standards were less stringent than
                augural. That continues to be the case under the final rule analysis,
                as discussed below. In addition, Section VI.D. discusses how past
                rulemaking analyses assumed that consumers were `myopic' and/or did not
                have adequate information about the benefits of fuel savings, which led
                them to choose to purchase less efficient vehicles than they otherwise
                would if they better understood the costs or savings they would accrue.
                As Section VI.D. explains, the agencies are less certain today that
                consumers improperly value fuel savings. Vehicle buyers today have more
                information about fuel costs than ever before, including right on the
                window sticker when considering a new vehicle purchase, and it is
                ultimately a private choice whether consumers prefer improvements in
                other vehicle attributes over additional fuel economy. When fuel costs
                are expected to rise manageably over time, it may be that consumers are
                comfortable choosing to absorb an additional $1,375 over the vehicle's
                lifetime, the estimated difference in lifetime expenditures between the
                proposal and if NHTSA was choosing to finalize the augural standards,
                and are even more comfortable choosing to absorb an additional $1,125,
                the estimated difference in lifetime expenditures between the final
                standards and what
                [[Page 25172]]
                the augural standards would have required. If fuel prices rise less
                than anticipated, as they have done since the 2012 final rule, or even
                decrease over time, buyers face an even smaller tradeoff between
                foregone fuel savings and the value of improvements in other aspects of
                new cars.
                 Consumer expenditures on fuel are important to understanding the
                benefits (and net benefits) of CAFE and CO2 standards. Every
                analysis of CAFE/CO2 standards relies on hundreds of
                assumptions, and estimates of costs and benefits developed as part of
                those analyses, by their very nature, depend on those assumptions.
                Specifically, the net benefits associated with each alternative result
                from the assumptions used and the relationships between vehicle
                production, ownership, and usage in which the assumptions interact. Put
                more simply, inputs affect outputs. As discussed in the section above
                on economic practicability, net benefits may be a consideration in the
                determination of maximum feasible standards, among the many other
                things the agency considers. While some commenters have asserted that
                the analysis for this rulemaking has ``put a thumb on the scale by
                undervaluing the benefits and overvaluing the costs of more stringent
                standards,'' \2956\ this final rule has identified a number of critical
                assumptions in the 2012 final rule that were problematic in the other
                direction (i.e., undervaluing the costs and overvaluing the benefits),
                for a variety of reasons. For example, the projected fuel prices in the
                2012 analysis inflated the value of fuel savings relative to what has
                actually occurred. That assumption about how fuel prices were projected
                to rise over time was solidly grounded at the time, but is no longer
                so, and continuing to use it would not be reasonable, even if that
                means that the benefits of all of the regulatory alternatives decrease
                as compared to what the 2012 analysis showed. Lower oil prices mean
                that fuel savings benefits for consumers are lower under any CAFE
                standards, whether the augural standards or the standard being
                finalized today--consumers may yet spend less on fuel under more
                stringent standards, but how much less matters.
                ---------------------------------------------------------------------------
                 \2956\ See CBD v. NHTSA, 538 F.3d 1172, 1189 (9th Cir. 2008).
                ---------------------------------------------------------------------------
                 Additionally, the assumption in 2012 that no market exists for fuel
                economy improvements at any fuel price or technology cost artificially
                inflated the value of fuel savings attributable to the standards in
                each regulatory alternative. The combination of assumptions and
                relationships (the examples above, and others) in the 2012 final rule
                produced estimates of net benefits that continued to increase with
                stringency from 1 percent per year through 6 percent per year.\2957\
                Under some alternatives, benefits actually would have appeared to be
                infinite, growing faster than the discount rate, if the analysis had
                been extended far enough into the future. No market works this way, and
                there is no reasonable set of assumptions under which costs could never
                exceed benefits no matter how much technology was deployed or how much
                stringency was required. Rather than demonstrating meaningfully that
                more stringent standards are always more beneficial to society, the
                result from the 2012 analysis suggests that that analysis was
                critically flawed. That said, while the 2012 analysis appeared to show
                that more technology, at a faster pace, is always preferable from the
                perspective of net benefits, the agencies ultimately relied on other
                features of the analysis and considerations of impacts in choosing a
                preferred alternative. While today's analysis produces an inflection
                point at a 3 percent discount rate--a level of stringency where further
                increases reduce net benefits as the tradeoff between regulatory costs
                and resulting net benefits tips the other way \2958\--the agencies
                similarly rely on considerations beyond net benefits in choosing the
                preferred alternative.\2959\
                ---------------------------------------------------------------------------
                 \2957\ The 7 percent per year alternative happened to be
                indistinguishable from the 6 percent alternative in that analysis.
                 \2958\ See Table VII-95.
                 \2959\ See CBD v. NHTSA, 538 F.3d 1172, 1188 (9th Cir. 2008).
                ---------------------------------------------------------------------------
                 NHTSA also agrees with many commenters that environmental (both
                climate and air quality) concerns are relevant to the need of the U.S.
                to conserve oil, as explained above. As the Supreme Court stated in
                Massachusetts v. EPA, ``[a] reduction in domestic emissions would slow
                the pace of global emissions increases,'' \2960\ and there is no
                question that CAFE standards directly affect CO2 emissions.
                Besides providing information on differences between the regulatory
                alternatives in terms of million metric tons of CO2 emitted,
                the NPRM also provided a chart illustrating the difference between the
                estimated atmospheric CO2 concentration (789.76 ppm) in 2100
                under the proposal as compared to the estimated level under the augural
                standards (789.11 ppm) in a scenario where no CO2 emissions
                reduction measures are implemented throughout the planet.\2961\ The
                NPRM noted that this translated to 3/1000ths of a degree Celsius
                increase in global average temperatures by 2100, relative to the
                augural standards. Many commenters strongly objected to the framing of
                these findings, as discussed above in the section on the environmental
                implications of the need of the U.S. to conserve energy. Changing the
                framing does not change the agency's findings.\2962\ For this final
                rule, the Preferred Alternative would result in 922.5 million metric
                tons of CO2 more than the estimated emissions if the augural
                standards were to be finalized (for MY 2017-MY 2029 vehicles between
                calendar years 2017 and 2070), which is 160.2 million fewer tons than
                if the proposed Preferred Alternative were to be finalized. It is
                reasonable to consider these raw million-metric-ton estimates in terms
                of their effects, namely, on estimated temperature change and sea level
                rise, which are the primary climate effects referred to and estimated.
                The FEIS accompanying today's rule estimates that, by 2100, global mean
                surface temperature will increase by 3.487 degrees (Celsius) under
                either the proposed or final standards, versus 3.484 degrees under the
                augural standards. The FEIS shows corresponding sea level rise in 2011
                reaching 76.34 cm under the final standards, 76.35 cm under the
                proposed standards, and 76.28 cm under the augural standards. This is
                accounted for in economic terms (i.e., translated from fractions of a
                degree temperature rise and from millimeters of sea level rise, among
                other things, into dollar-based effects) in the measure of the social
                cost of carbon, described in Section VI.D.1.b)(13).
                ---------------------------------------------------------------------------
                 \2960\ Mass. v. EPA, 549 U.S. at 526.
                 \2961\ 83 FR at 42996-97 (Aug. 24, 2018).
                 \2962\ In fact, NHTSA's analysis in Section 8.6.4.2 of the FEIS
                illustrates that the differences between alternatives are similar in
                reference to other GCAM scenarios. Regardless of whether there will
                be widespread global efforts to mitigate climate change, the impacts
                of this action are roughly the same.
                ---------------------------------------------------------------------------
                 NHTSA is mindful of the language in Massachusetts v. EPA that
                ``[a]gencies . . . do not generally resolve massive problems in one
                fell regulatory swoop,'' \2963\ and acknowledges the concerns of many
                commenters that standards less stringent than augural may result in
                higher CO2 emissions. In response, it is important to
                remember that even under the proposal, sales of new vehicles would,
                over time, have continued to improve the fuel economy and reduce the
                CO2 emissions of the on-road fleet through fleet turnover
                effects, as discussed in Section IV. Under the final rule, those rates
                of improvement will likely be faster than they would have been if NHTSA
                were finalizing the
                [[Page 25173]]
                proposal. Emissions are still being reduced under the final rule, and
                the on-road fleet will be less energy and carbon intensive than it is
                today. NHTSA is taking the impacts of CO2 emissions into
                account, while also considering the other statutory factors in its
                balancing.
                ---------------------------------------------------------------------------
                 \2963\ Mass. v. EPA, 549 U.S. at 524.
                ---------------------------------------------------------------------------
                 It is also important to note that the science of climate change and
                the models used to assess effects on climate variables (and other
                effects discussed in Section VII.A.4.b, and in the DEIS/FEIS) are
                subject to various types and degrees of uncertainty. In light of this,
                NHTSA also conducted climate sensitivity analyses in the FEIS.\2964\ In
                these analyses, NHTSA considered a range of climate sensitivities (1.5
                [deg]C, 2.0 [deg]C, 2.5 [deg]C, 3.0 [deg]C, 4.5 [deg]C, and 6.0 [deg]C)
                for a doubling of CO2 compared to preindustrial atmospheric
                concentrations (278 ppm CO2). Even under the least stringent
                alternative considered (the proposal) and assuming the highest level of
                climate sensitivity (6.0 [deg]C), the global mean surface temperature
                increase in 2100 was 0.006 [deg]C higher than under the augural
                standards. Thus, accounting for some of this uncertainty, impacts on
                global mean surface temperature resulting from this action remain very
                small.
                ---------------------------------------------------------------------------
                 \2964\ See Sections 5.4.2.3 and 8.6.4.2 of the FEIS.
                ---------------------------------------------------------------------------
                 NHTSA received many comments about the costs of delaying
                CO2 emissions reductions and the potential of crossing
                climate tipping points and triggering abrupt climate change. Many of
                these costs and risks are factored in to the social cost of carbon, and
                are therefore considered as part of the agency's cost-benefit analysis.
                And many of these costs and risks cannot be quantified at all: The
                current state of science does not allow for quantifying how increased
                emissions from a specific policy or action might affect the probability
                and timing of abrupt climate change. However, NHTSA does recognize that
                while these costs cannot be quantified, they do exist and must also be
                taken into account. Ultimately, the costs of delaying CO2
                emissions reductions (both the ones that can be accounted for
                quantitatively and those that can only be considered qualitatively)
                must also be balanced against the costs associated with more stringent
                alternatives. Some of the costs associated with more stringent
                alternatives are direct, such as the additional costs passed on to
                consumers for technology that improves fuel economy. Other costs are
                indirect, such as environmental costs associated with more stringent
                fuel economy standards. For example, the increased electrification of
                motor vehicles can result in localized impacts associated with the
                production and recycling of lithium-ion batteries. Similarly, the
                increased reliance on material substitution for vehicle mass reduction
                could result in various environmental impacts associated with
                manufacture and recycling. Certainly, the benefits of these
                technologies in reducing carbon emissions outweighs the other life-
                cycle environmental impacts, but that does not mean NHTSA can just
                ignore those impacts, either.
                 Many commenters claimed that NHTSA ignored the effects of climate
                change or determined they were inevitable, not urgent enough to act
                upon, or not worth the effort to address at all. NHTSA makes none of
                those determinations here. On the contrary, NHTSA has considered the
                material on this subject in the administrative record and the plethora
                of public comments we received on the topic. The agency recognizes what
                is at stake, but we also recognize that NHTSA is not charged by
                Congress to single-mindedly address carbon emissions at the expense of
                all other considerations. The question before NHTSA is not whether to
                conserve energy (and thereby reduce carbon emissions, which drive
                climate change) but by how much each year. Taking climate change into
                account elevates the importance of the ``need of the United States to
                conserve energy'' criterion in NHTSA's balancing. However, in light of
                the limits in what the agency can achieve, the potential offsetting
                impacts to the environment, and the statutory requirement to consider
                other factors, the impacts of carbon emissions alone cannot drive the
                outcome of NHTSA's decision-making.
                 NHTSA also recognizes the potential impacts of this rulemaking on
                air quality. To be clear, this final rule does not directly involve the
                regulation of pollutants such as carbon monoxide, smog-forming
                pollutants (nitrogen oxides and unburned hydrocarbons), or ``air
                toxics'' (e.g., formaldehyde, acetaldehyde, benzene). Nevertheless,
                NHTSA recognizes that this rule is expected to impact such emissions
                indirectly (by reducing travel demand and accelerating fleet turnover
                to newer and cleaner vehicles on one hand while, on the other,
                increasing activity at refineries and in the fuel distribution system).
                Based on a review of Section VII.A.4.c. above and the FEIS, NHTSA
                believes these impacts are much smaller than impacts on fuel use and
                CO2 emissions, and therefore factor in less to the need of
                the U.S. to conserve energy.\2965\
                ---------------------------------------------------------------------------
                 \2965\ For an explanation of how NHTSA considers environmental
                impacts and the differences between the preamble and FEIS analyses,
                see Section VII.A.4.c.1 above.
                ---------------------------------------------------------------------------
                 For criteria pollutants, NHTSA estimates that emissions over the
                lifetimes of vehicles through MY 2029 under the alternatives will not
                change significantly. Tailpipe emissions of most pollutants will
                generally decrease, while upstream emissions will generally increase.
                Overall emissions under the action alternatives for most pollutants
                will increase over time. Changes are not uniform year-to-year, however,
                reflecting the complex interaction of the amount of highway travel, the
                distribution of that travel among different vehicles, upstream
                processes, etc. Generally, tailpipe air toxic emissions decrease while
                upstream air toxic emissions increase. Over the long term, however, the
                upstream emissions increase further while the decreases in tailpipe
                emissions become less pronounced. Overall, NHTSA anticipates that air
                toxic emission will increase over time under the action alternatives.
                Most alternatives result in cumulative increases in adverse health
                impacts associated with total upstream and tailpipe pollutant
                emissions. Although some alternatives would have resulted in decreases,
                the differences among alternatives across the lifetime of vehicles
                through MY 2029 are not large.
                 NHTSA also considered the various impacts reported qualitatively in
                the FEIS and described briefly above in Section VIII.B.3. Although the
                agency cannot compare the impacts of the alternatives quantitatively
                (except to the degree that they are otherwise covered by the agency's
                monetary cost-benefit analysis, such as through the social cost of
                carbon), NHTSA recognizes that such impacts would generally increase
                under all the action alternatives compared to the augural standards. In
                Chapter 8 of the FEIS, for example, NHTSA provides a qualitative
                discussion of the long-term impacts of climate change on key natural
                and human resources. While these impacts would be expected to increase
                under the action alternatives, the change is expected to be very small.
                In contrast, the FEIS also discusses some environmental impacts that
                would decrease with the lower stringencies considered in this
                rulemaking. For example, in Chapter 6 of the FEIS, NHTSA provides a
                literature review of potential lifecycle impacts as a result of
                manufacturer use of various materials and technologies to meet the
                standards. NHTSA can account for the benefits to
                [[Page 25174]]
                tailpipe emissions of these technologies as part of its evaluation of
                technology effectiveness. However, as discussed in the FEIS, accounting
                for the upstream emissions associated with the processes used in the
                manufacture of these technologies can be complicated. Because the
                adoption of these materials and technologies would vary across
                alternatives, and each has varying upstream impacts, the agency cannot
                provide meaningful comparisons across alternatives. Still, any benefit
                to tailpipe CO2, criteria pollutant, or air toxic emissions
                of more stringent alternatives would be offset by the increased
                upstream impacts reported in that section.\2966\
                ---------------------------------------------------------------------------
                 \2966\ In most cases, tailpipe emissions benefits offset
                upstream environmental impacts associated with materials and
                technologies NHTSA considered in its analysis. However, in some
                cases, results may not align with conventional wisdom. For example,
                while EVs can offer significant life-cycle GHG emissions savings
                over conventional vehicles, this is highly dependent on the time and
                location of charging. In some regions, life-cycle impacts are
                similar for EVs and conventional vehicles.
                ---------------------------------------------------------------------------
                 In total, environmental impacts factor into the need of the U.S. to
                conserve energy and potentially elevate that criterion, but those
                impacts cannot be considered in isolation. While some impacts are more
                significant than others, NHTSA must consider how much weight to place
                on this factor as well as the relative weight of other factors.
                 Thus, even if the agency no longer interprets the need of the U.S.
                to conserve energy as necessarily boundless as it once did, as it
                explained in the NPRM and again in the discussion above, NHTSA
                continues to believe that the factor functions in the overall balancing
                to push toward increases in stringency, and notes that any increase in
                stringency over the last binding standards--not in question at this
                point, the standards for MY 2020--does conserve energy and reduce
                negative environmental impacts. In fact, fleet turnover over time means
                that less energy is being consumed by the fleet over time even if
                standards did not increase year over year. Even if new vehicles are not
                all as efficient as would have been required under more stringent
                standards, they are still more efficient on average than the older
                vehicles they are replacing, particularly after a decade of successive
                increases in CAFE standard stringency, as Section IV above discusses.
                The on-road fleet has well over 250 million vehicles, dwarfing the
                roughly 16 million new vehicles sold each year. Comprehensive energy
                savings come from turning over legacy vehicles in the fleet so that
                overall fleet fuel economy increases. If the NPRM's preferred
                alternative were finalized, the fuel consumption of the passenger car
                and light truck fleet would have fallen from roughly 8.5 million
                barrels per day (currently) to roughly 7 million barrels per day by
                2050 as the fleet turned over. Finalizing the 1.5 percent alternative
                reduces that number to 6.3 million barrels per day. That breaks the
                trend of increasing oil consumption over time, and conserves energy.
                (2)Technological Feasibility and the Effect of Other Motor Vehicle
                Standards of the Government on Fuel Economy
                 As in the 2012 final rule, technological feasibility and the effect
                of other motor vehicle standards of the Government on fuel economy do
                weigh in NHTSA's balancing of the relevant factors, but they play a
                less significant role because they vary less across regulatory
                alternatives than the other factors vary. Technological feasibility, as
                explained above and as similarly explained in 2012, relates to whether
                technologies exist and can be commercially applied during the
                rulemaking timeframe. None of the regulatory alternatives under
                consideration today would require brand new technologies to be
                invented--they can all be met with technology that exists currently.
                However, as recognized in the 2012 final rule, ``some technologies that
                currently have limited commercial use cannot be deployed on every
                vehicle model in MY [2021], but require a realistic schedule for
                widespread commercialization to be feasible. . . . Any of the
                alternatives could thus be achieved on a technical basis alone if the
                level of resources that might be required to implement the technologies
                is not considered.'' As explained above in the discussion of economic
                practicability, however, resources must be, and are, considered. The
                2012 final rule further explained that ``If all alternatives are at
                least theoretically technologically feasible in the [rulemaking]
                timeframe, and the need of the nation is best served by pushing
                standards as stringent as possible, then the agency might be inclined
                to select the alternative that results in the very most stringent
                standards considered.'' The 2012 final rule stated, however, that such
                a selection would be inappropriate because ``the agency must also
                consider what is required to practically implement technologies, which
                is part of economic practicability, and to which the most stringent
                alternatives give little weight.''
                 NHTSA considers technological feasibility similarly to how it has
                long considered that factor--for the most part, the question of what
                standards are maximum feasible is less about technological feasibility
                than about economic practicability. All of the regulatory alternatives
                considered in this final rule are likely technologically feasible, but
                that does not mean that any of them could be maximum feasible, just as
                we concluded in evaluating alternatives in 2012. NHTSA must now account
                for how the need of the U.S. to conserve oil has changed, and this
                consideration tips our balancing away from the most stringent
                standards.
                 For the effect of other motor vehicle standards of the Government
                on fuel economy, there is relatively little variation across regulatory
                alternatives, as discussed in the FRIA. As in the 2012 final rule, in
                developing this final rule NHTSA considered the effects of compliance
                with known and possible NHTSA safety standards and known EPA emission
                standards in developing this final rule, and has accounted for those
                effects in the analysis. The effect of other motor vehicle standards of
                the Government does not, therefore, have a noticeable effect on NHTSA's
                balancing of factors to determine maximum feasible standards.
                (3) Economic Practicability
                 Economic practicability remains a complex factor to consider and
                balance, as discussed above, encompassing a variety of different issues
                that are each captured to various degrees through the analysis. As
                NHTSA stated in the 2012 final rule, ``The agency does not necessarily
                believe that there is a bright-line test for whether a regulatory
                alternative is economically practicable, but there are several metrics
                . . . that we find useful for making the assessment.'' \2967\ In 2012,
                as today, NHTSA looks to factors like:
                ---------------------------------------------------------------------------
                 \2967\ 77 FR at 63038 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 Per-vehicle cost, in terms of ``even if the technology
                exists and it appears that manufacturers can apply it consistent with
                their product cadence, if meeting the standards will raise per-vehicle
                cost more than we believe consumers are likely to accept, which could
                negatively impact sales and employment in this sector, the standards
                may not be economically practicable;'' \2968\
                ---------------------------------------------------------------------------
                 \2968\ Id.
                ---------------------------------------------------------------------------
                 Application rate of technologies, because ``even if
                shortfalls are not extensive, whether it appears that a regulatory
                alternative would impose undue burden on manufacturers in either or
                both the near and long term in terms of how much and which technologies
                might be required'' can be
                [[Page 25175]]
                relevant to manufacturers' difficulty with meeting standards; \2969\
                ---------------------------------------------------------------------------
                 \2969\ Id.
                ---------------------------------------------------------------------------
                 Consumer demand, which NHTSA described in 2012 as ``other
                . . . considerations related to the application rate of technologies,
                whether it appears that the burden on several or more manufacturers
                might cause them to respond to the standards in ways that compromise .
                . . other aspects of performance that are important to consumer
                acceptance of new products'' \2970\
                ---------------------------------------------------------------------------
                 \2970\ Id.
                ---------------------------------------------------------------------------
                 Manufacturer compliance shortfalls, because ``If it
                appears, in our modeling analysis, that a significant portion of the
                industry cannot meet the standards defined by a regulatory alternative
                in a model year, given that our modeling analysis accounts for
                manufacturers' expected ability to design, produce, and sell vehicles
                (through redesign cycle cadence, technology costs and benefits, etc.),
                then that suggests that the standards may not be economically
                practicable;'' \2971\
                ---------------------------------------------------------------------------
                 \2971\ Id.
                ---------------------------------------------------------------------------
                 Uncertainty and consumer acceptance of technologies, which
                the 2012 final rule said was ``not accounted for expressly in our
                modeling analysis, but [was] important to an assessment of economic
                practicability given the time frame of this rulemaking.'' \2972\
                ---------------------------------------------------------------------------
                 \2972\ Id.
                ---------------------------------------------------------------------------
                 Thus, estimated impacts on per-vehicle cost are one issue;
                estimated sales and employment impacts are issues; uncertainty
                surrounding future market conditions and consumer demand for fuel
                economy (versus consumer demand for other vehicle attributes) are other
                issues. Consumers may respond to per-vehicle cost increases by choosing
                to keep their current vehicle or buy used vehicles instead of new
                vehicles, with consequent effects on new vehicle sales and the overall
                fleet makeup; consumers may respond to new fuel-economy-improving
                technologies on certain models by choosing to buy other models,
                especially when fuel costs are not expected to increase significantly
                in the ownership timeframe and consumers value other vehicle attributes
                more than they value fuel economy. Either of these responses may cause
                manufacturers both to lose money and to face further difficulties in
                meeting the CAFE standards. While there are significant benefits for
                both manufacturers and consumers under attribute-based standards,
                manufacturers must still sell enough ``target-beaters'' to balance out
                sales of less-fuel-efficient vehicles and meet their overall fleet-
                average compliance obligations. If consumer demand shifts strongly away
                from target-beaters, CAFE compliance will be a struggle, even if the
                target-beaters are widely available. Section IV above discusses this
                phenomenon in more detail. And if consumers buy fewer new vehicles in
                response to per-vehicle cost increases, which the agencies are
                beginning to see already, \2973\ the fleet as a whole will turn over
                more slowly, and fuel conservation gains may also be slowed. NHTSA does
                not believe that that is EPCA's goal. Manufacturers struggling to sell
                new vehicles will have less capital to devote to further technological
                improvements; may choose to move manufacturing jobs outside the U.S. to
                places with lower labor costs; and so forth. A net benefits analysis
                may be informative to attempting to quantify some of the issues
                described above, but not all of these issues lend themselves to clear
                quantification. The following discussion will evaluate what the
                agencies believe has been reasonably accounted for.
                ---------------------------------------------------------------------------
                 \2973\ See, e.g., Jackie Charniga, ``Prime buyers flood used-
                vehicle market in Q4,'' Automotive News, March 4, 2020, https://www.autonews.com/finance-insurance/prime-buyers-flood-used-vehicle-market-q4.
                ---------------------------------------------------------------------------
                (a) Per-Vehicle Costs, Sales, and Employment as Part of Economic
                Practicability
                 Per-vehicle cost estimates are relevant to NHTSA's consideration of
                economic practicability because, when cost increases associated with
                more stringent standards are passed through to consumers as price
                increases, they affect consumers' willingness and ability to purchase
                new vehicles, and thus influence vehicle sales and fleet turnover. A
                similar effect occurs in reverse when stringency is decreased. Table
                VIII-7 below shows the estimated effects on per-vehicle costs by
                regulatory alternative in MY 2029:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.736
                 Generally speaking, per-vehicle costs increase as stringency
                increases. The agencies estimate that, by MY 2029, costs for additional
                fuel-saving technology (beyond that present on vehicles in MY 2017)
                would average about $2,800 under the augural CAFE standards, as
                compared to about $1,400 under the proposed CAFE standards,
                [[Page 25176]]
                and about $1,650 under the final CAFE standards for MYs 2021-2026. The
                next most stringent alternative beyond the 1.5 percent alternative is
                the ``2%/3%'' alternative. Under 2%/3%, the agencies estimate that
                costs would increase by $2,000 per vehicle on average. NHTSA
                understands that many readers may not find an extra $350 per vehicle to
                be a compelling reason to reject the 2%/3% alternative, or even find an
                additional $1,125 per vehicle a reason to reject the baseline/augural
                standards. As the NPRM discussed, ``. . . the corresponding up-front
                and monthly costs may pose a challenge to low-income or credit-
                challenged purchasers. . . . such increased costs will price many
                consumers out of the market--leaving them to continue driving an older,
                less safe, less efficient, and more polluting vehicle, or purchasing
                another used vehicle that would likewise be less safe, less efficient,
                and more polluting than an equivalent new vehicle.'' \2974\ This
                continues to be a concern: For example, the average MY 2025 prices
                estimated here under the baseline, final, and 2%/3% CAFE standards are
                about $38,100, $36,850, and $37,150, respectively. The buyer of a new
                MY 2025 vehicle might thus avoid the following purchase and first-year
                ownership costs under the final standards as compared to the baseline
                standards or 2%/3% standards:
                ---------------------------------------------------------------------------
                 \2974\ 83 FR at 43222 (Aug. 24, 2018).
                 [GRAPHIC] [TIFF OMITTED] TR30AP20.737
                
                 While the buyer of the average vehicle would also purchase somewhat
                more fuel under the final standards than the baseline standards, this
                difference might average less than four gallons per month during the
                first year of ownership. Some purchasers may consider it more important
                to avoid these very certain (e.g., being reflected in signed contracts)
                cost savings than the comparatively uncertain (because, e.g., some
                owners drive considerably less than others, and may purchase fuel in
                small increments as needed) fuel savings. For some low-income
                purchasers or credit-challenged purchasers, the cost savings may make
                the difference between being able or not to purchase the desired
                vehicle. As vehicles get more expensive in response to higher CAFE
                standards, it will get more and more difficult for manufacturers and
                dealers to continue creating loan terms that both keep monthly payments
                low and do not result in consumers still owing significant amounts of
                money on the vehicle by the time they can be expected to be ready for a
                new vehicle. These considerations were discussed in the NPRM and they
                remain true for this final rule.
                ---------------------------------------------------------------------------
                 \2975\ Edmunds estimates that the average down payment for a new
                vehicle in 2019 was 11.7% of the vehicle's price, see https://www.edmunds.com/car-buying/how-much-should-a-car-down-payment-be.html.
                ---------------------------------------------------------------------------
                 Per-vehicle cost and fuel economy both affect sales estimates in
                the final rule analysis. Table VIII-9 below shows the estimated effects
                on fleet-wide sales by regulatory alternative from 2017-2030, where the
                augural standards represent absolute sales and all other alternatives
                represent increases relative to the augural sales:
                BILLING CODE 4910-59-P
                [[Page 25177]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.738
                BILLING CODE 4910-59-C
                 The final rule analysis indicates that industry sales decrease as
                stringency increases, and increase as stringency decreases. While sales
                under both the
                [[Page 25178]]
                proposal and the final rule are comparable, each represents about a 1.5
                percent reduction in total sales over the period from 2017--2030. In
                the context of 16-17 million new vehicle sales annually, NHTSA does not
                believe that the sales volume effects here, while significant, are
                necessarily determinative for economic practicability, even after
                accounting for fuel economy effects in the sales analysis as some
                commenters recommended. That said, NHTSA recognizes that the final rule
                sales analysis does not account for a number of factors that could
                cause differences between alternatives to result in changes in new
                vehicle sales (perhaps greater). For example, as explained above, NHTSA
                remains concerned that significant increases in fixed upfront prices
                (which for many people translate to monthly financing costs) are harder
                for certain segments of new vehicle buyers to manage than fuel costs,
                which can be managed to some extent through vehicle switching or travel
                decisions. The sales analysis for this final rule indicates that more
                stringent standards tend to result in higher light truck sales and
                lower passenger car sales. While NHTSA does not have specific
                information (or a vehicle choice model) to inform the agency about
                which consumers (by income) buy which vehicles, and while NHTSA
                acknowledges that it does not account for price cross-subsidization by
                manufacturers to keep ``entry-level'' new vehicle (often, passenger
                car) prices low, NHTSA continues to be concerned about the possibility
                of a bubble in the market for new vehicles. As the Wall Street Journal
                reported in November 2019, ``Some 33% of people who traded in cars to
                buy new ones in the first nine months of 2019 had negative equity,
                compared with 28% five years ago and 19% a decade ago, according to
                car-shopping site Edmunds . . . . Rising car prices have exacerbated an
                affordability gap that is increasingly getting filled with auto debt.''
                \2976\ The sales analysis for this final rule does not directly account
                for these effects, but NHTSA is concerned that they may be
                considerable. NHTSA notes that this analysis does not take into account
                potential economic turmoil or recession, which may have a significant
                impact on vehicle sales and industry viability.\2977\
                ---------------------------------------------------------------------------
                 \2976\ AnnaMaria Andriotis and Ben Eisen, ``A $45,000 Loan for a
                $27,000 Ride: More Borrowers are Going Underwater on Car Loans,''
                Wall Street Journal, November 9, 2019.
                 \2977\ Letter from Alliance for Automotive Innovation, NADA, and
                MEMA to Congress, Mar. 23, 2020, available at https://www.autosinnovate.org/wp-content/uploads/2020/03/COVID-19-Letter-to-Congress-NADA-MEMA-AAI-March-23.pdf.
                ---------------------------------------------------------------------------
                 The final rule analysis also looked at employment effects under the
                different regulatory alternatives. A number of commenters argued that
                more stringent standards improved employment opportunities, as shown in
                the NPRM analysis and in other analyses, due to the need for workers to
                manufacture the additional technology needed to meet those more
                stringent standards. Similar to the NPRM analysis, the agencies'
                updated analysis shows labor utilization, on balance, increasing
                slightly with stringency, as this effect outweighs the opposing effect
                of changes in vehicle sales. Table VIII-11 below shows the estimated
                effects on U.S. auto industry employment by regulatory alternative in
                MY 2029:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.739
                [GRAPHIC] [TIFF OMITTED] TR30AP20.740
                 It is important to note, however, that the reduction in person-
                years described in this table merely reflects the fact that, when
                compared to the standards set in 2012, fewer jobs will be specifically
                created to meet infeasible regulatory requirements. It is also
                important to note that the $15 billion in avoided required technology
                costs (in MY 2029) can be invested by manufacturers into other areas,
                or passed on to consumers. Moreover, consumers can either take those
                cost savings in the form of a reduced vehicle price, or used toward the
                purchase of specific automotive features that they desire (potentially
                including a more-efficient vehicle or optional safety features that can
                reduce risk of injury or death for all vehicle occupants on the road),
                which would increase employment among suppliers and manufacturers.
                 Generally speaking, the agencies' analysis shows net labor
                utilization increasing with stringency, because the additional labor
                utilization involved with producing additional fuel-saving
                [[Page 25179]]
                technology outweighs the foregone labor utilization involved with the
                foregone sales. As indicated above, for the scope of labor utilization
                accounted for in today's analysis, the agencies show about 1.20 million
                person-years under the augural CAFE standards and about 1.19 million
                person-years under either the proposed or final standards. As for
                sales, it is arguably instructive to consider these estimates in the
                broader context of U.S. employment. BLS data indicates that roughly 129
                million people in the U.S. are employed full-time at the time of
                writing,\2978\ and that roughly 1.4 million people were employed in
                motor vehicle and motor vehicle equipment manufacturing in 2018.\2979\
                The agencies estimate that, compared to the augural standards, the
                final standards will reduce automotive labor utilization associated
                with production of the MY 2029 fleet by about 1.1%, a slightly smaller
                reduction than the 1.4% estimated to occur under the proposed
                standards. For comparison, the Synapse Report cited often by commenters
                concluded that vehicle standards result in ``nationwide employment
                increases of more than 100,000 in 2025 and more than 250,000 in 2035. .
                . these increases represent less than 0.2 percent of current U.S.
                employment levels.'' \2980\ Even at these levels, which NHTSA does not
                necessarily agree are accurate, the employment effects of standards are
                in the range of the average of more than 216,000 jobs added to the U.S.
                economy during each month of 2018.\2981\ That said, as for sales, NHTSA
                recognizes that the final rule labor utilization analysis does not
                account for a number of factors that could cause differences between
                alternatives to be different (perhaps greater), as discussed further
                below.
                ---------------------------------------------------------------------------
                 \2978\ https://www.bls.gov/cps/cpsaat08.htm.
                 \2979\ https://www.bls.gov/cps/cpsaat18b.htm.
                 \2980\ https://www.synapse-energy.com/sites/default/files/Cleaner-Cars-and%20Job-Creation-17-072.pdf, at ES-2.
                 \2981\ Payroll employment increased by 2.6 million jobs in 2018,
                an average of 216,667 per month. ``The Employment Situation--
                December 2018,'' Bureau of Labor Statistics, available at: https://www.bls.gov/news.release/archives/empsit_01042019.pdf.
                ---------------------------------------------------------------------------
                (b) Application Rates for New Technologies as Part of Economic
                Practicability
                 The sales analysis for this final rule also does not account for
                the potential consumer acceptance issue of more stringent standards
                effectively requiring the application of technologies not yet ready for
                widespread deployment. As widely noted, the 2012 rule assumed extremely
                high penetration of dual-clutch transmissions in response to standards.
                While the agencies stated throughout that final rule that the analysis
                was not meant to represent the expected response to the standards, Ford
                did apply DCTs to a number of vehicle models in its fleet, that
                resulted in major customer satisfaction issues and ultimately caused
                extensive buyback campaigns, customer service programs, and class-
                action litigation.\2982\ Sales can be impacted as a result of standards
                if technologies applied in response to those standards have
                operational, maintenance, or customer acceptance problems, or if
                consumers are unwilling to pay for it. Manufacturer capital to develop
                and add new technologies and manage these rollout issues is finite, as
                discussed. Insufficient capital can also cause quality problems. The
                cost effects modeled in this final rule analysis, that drive the sales
                and scrappage analyses, only include technology costs and RPE--they do
                not include the cost of stranded capital or lost consumer surplus,
                which are things that could drive up costs, drive down benefits, and
                therefore impact sales and scrappage beyond what today's analysis
                shows.
                ---------------------------------------------------------------------------
                 \2982\ See https://www.autonews.com/technology/dual-clutch-gearbox-complaints-haunt-ford.
                ---------------------------------------------------------------------------
                 As Section IV above notes, a great deal of fuel economy-improving
                technology has already been added to the fleet since 2012, which means
                that the amount of fuel economy-improving technology left to be added
                in response to higher standards is less than it was assumed to be in
                2012. Looking at the technology penetration rates modeled in today's
                analysis, it appears that the augural standards are projected to
                require nearly 20 percent total electrification in MY 2029, while the
                proposal would have required nearly 7 percent and the final standards
                would require nearly 8 percent. Table VIII-11 below shows projected
                electrification rates by 2029 for the regulatory alternatives--
                electrification refers to all models with strong hybrids, PHEVs, or
                full EVs:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.741
                [[Page 25180]]
                 As the table shows, the analysis projects that meeting the augural
                standards could require over twice as much electrification as the final
                rule standards could require.\2983\ The current market penetration for
                all such vehicles is only approximately 4 percent even though the
                technology is well-established, with hybrids having been first
                introduced with the Honda Insight in 1999 and Toyota Prius in 2000,
                plug in hybrids with the Chevrolet Volt in late-2010 and electric
                vehicles with the Tesla Roadster in 2008 and Nissan Leaf in late 2010.
                As Mr. Kreucher commented, and as Figure VIII-2 shows, consumers appear
                to be driven by fuel price. Given anticipated fuel prices during this
                timeframe and evidence in the market today of cannibalization within
                these vehicle segments (not to mention the continued phasing out of
                government incentives for these vehicles),\2984\ NHTSA is concerned
                that there could be consumer acceptance problems associated with
                further electrification under more stringent alternatives, which could
                have sales impacts.
                ---------------------------------------------------------------------------
                 \2983\ While NHTSA is prohibited by statute from considering
                battery electric vehicles as a compliance mechanism, we are aware
                that many OEMs will likely opt to produce a smaller number of fully
                electric vehicles rather than a large number of strong hybrid
                models.
                 \2984\ 26 U.S.C. Section 30D provides for tax credits ranging
                from $2,500 to $7,500 for purchasers of qualifying plug-in hybrid
                (PHEV) and battery electric (BEV) vehicles, with a phaseout applying
                to vehicle manufactured by an automaker once they sell 200,000
                qualifying vehicles. Both Tesla and General Motors have reached this
                threshold and the tax credit applicable to purchasers of new PHEV
                and BEV vehicles from those manufacturers has been reduced gradually
                and will phase out completely on January 1, 2020 for Tesla, and
                April 1, 2020 for General Motors.
                 The California Clean Vehicle Rebate Project was launched in 2010
                to provide incentives of up to $5,000 for purchasers or lessees of
                qualifying PHEV, BEV, and certain other alternative fuel vehicles.
                Since then, the program has undergone significant changes, including
                the addition of income eligibility criteria for certain incentives,
                and excluding eligibility toward the purchase or lease of a vehicle
                with an MSRP exceeding $60,000.
                 Separately, in 2005, California passed a law allowing hybrid
                electric vehicle (HEV), plug in hybrid electric vehicle (PHEV), and
                battery electric vehicle (BEV), and other qualifying alternative
                fuel vehicle owners to apply for a sticker allowing single-occupant
                access to High Occupancy Vehicle (HOV) lanes. HEV access was phased
                out in 2011, with eligibility being limited to PHEV, BEV and other
                qualifying alternative fuel vehicle owners. Access is now limited to
                a four-year period, and only to individuals who do not receive a
                rebate under the California Clean Vehicle Rebate Project (unless
                meeting income eligibility requirements).
                ---------------------------------------------------------------------------
                 We underscore that the table above simply shows the analytical
                results of the modeling for today's final rule based upon the most
                cost-effective means of achieving a given standard--it does not show
                how manufacturers would, or could, comply with the CAFE standards
                represented by the different regulatory alternatives. The discussion
                below covers the topic of manufacturer compliance shortfalls, and this
                discussion and that one are connected: The final rule analysis does not
                show significant compliance shortfalls under any regulatory
                alternative, but NHTSA believes that this is in large part because the
                CAFE model is not programmed with assumptions about consumer acceptance
                of strong hybrid technologies. In effect, the model lets manufacturers
                lean on hybridization to achieve compliance at a lower cost than if
                manufacturers instead pursued, for example, more advanced engine
                technologies. If cost-effectiveness is the only concern, that may be a
                valid compliance choice. If consumer acceptance of hybrid vehicles is
                accounted for, especially in a time of foreseeably low fuel prices, it
                may not be a valid compliance choice.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.742
                [[Page 25181]]
                 As Figure VIII-2 illustrates, the market share of strong hybrids in
                the new vehicle market has mostly tracked fuel prices. The bars
                represent the market share (left axis) and the line tracks the price of
                fuel (on the right axis). The light numbers inside of each bar
                represent the number of unique strong hybrid models offered for sale in
                that year. Initially, we see rapid growth that continues during the
                fuel price increases of the mid-2000s and peaking at around 3.5 percent
                market share. The figure shows that neither the passage of time, where
                consumers become more familiar with the technology over successive
                vehicle purchases, nor the number of models offered for sale have much
                of an impact on the market share for strong hybrids. Despite a doubling
                of the number of models offered for sale in subsequent years, market
                share continued to track fuel price closely, and fell dramatically as
                prices fell in 2015 and 2016. At fuel prices at or above $3.50/gallon,
                strong hybrids were able to capture additional market share. However,
                the current projection does not show prices returning to those levels
                for quite some time--leaving manufacturers uncertain about their
                ability to sell strong hybrids in the numbers estimated to be needed to
                comply with CAFE and CO2 standards before MY 2026.
                 The agencies conducted a sensitivity analysis to evaluate the
                impact of compliance pathways that did not rely on strong hybrids (see
                Chapter 7 of the Final RIA). As we discuss in the sensitivity analysis,
                in the absence of strong hybrids, compliance pathways tend toward a
                greater reliance on advanced engines and transmissions, and more
                aggressive exploitation of opportunities to reduce vehicles' mass.
                These alternative technology pathways carry with them additional
                technology costs that increase compliance costs in the baseline and
                increase the savings associated with the preferred alternative.
                 Under the CAFE program, where battery electric vehicles are not a
                compliance option (due to statutory restrictions on their consideration
                for rulemaking), the additional cost of advanced engine technology in
                the baseline increases baseline technology cost by about $800 per
                vehicle, and increases the cost savings under the preferred
                alternative, which has a much smaller reliance on strong hybrids to
                achieve compliance, by about $600 per vehicle. This difference is
                sufficient to change the sign on net social benefits for the preferred
                alternative to being slightly negative, to being very positive (nearly
                $80 billion at a 3 percent discount rate). The magnitude of this impact
                is comparable to the impact of varying fuel price projections.
                 As shown in, Figure VIII-2 even the preferred alternative requires
                levels of strong hybridization (and PHEV share) that would be about
                twice what has been observed at the market, even at its peak. Both the
                baseline and the 2%/3% alternative have even greater reliance on
                hybridization--more than twice as much in the baseline. The compliance
                costs associated with each alternative in today's rule depend upon the
                estimated levels of hybridization in the compliance scenarios being
                possible to achieve in the new vehicle market. The sensitivity analysis
                shows that manufacturers can still reach comparable levels of fuel
                economy without additional reliance on hybridization, but at
                significantly higher per-vehicle costs. Those higher costs have
                implications for the sales response, vehicle retirement rates in the
                existing vehicle population, and the penetration rate of emerging
                safety features.
                (c) Consumer Demand as Part of Economic Practicability
                 As discussed above, NHTSA's consideration of consumer demand as
                relevant to economic practicability has been upheld by the D.C. Circuit
                in Center for Auto Safety v. NHTSA. A number of commenters argued that
                consumers do, in fact, demand more fuel economy than the NPRM analysis
                assumed; that consumers will appreciate more widespread application of
                fuel economy-improving technologies that NHTSA appears to believe they
                will tolerate; that NHTSA was wrong to assume that fuel prices will
                remain relatively low in the future and continue to dampen consumer
                demand for fuel economy; and that vehicle manufacturers will not make
                tradeoffs between investments in fuel economy improvements and
                investments in other vehicle characteristics which consumers also
                demand, such that requiring manufacturers to meet more stringent
                standards will not impair consumer demand for new vehicles because less
                of those other characteristics will be available. Those commenters also
                often highlighted the CAS language stating that consideration of
                consumer demand may not undermine EPCA's goal of energy conservation.
                 NHTSA agrees with commenters that some consumers seek out vehicle
                models with higher fuel efficiency, and notes that those consumers have
                increasing numbers of relatively high-efficiency vehicle models to
                choose from in the current new-vehicle market, as shown in the previous
                section. CAFE does not affect fuel economy improvements that are
                supported by consumer demand--market forces will take care of that.
                Instead, it specifically addresses fuel economy improvements that are
                not preferred by consumers, and the agency sets standards that require
                manufacturers to make fuel economy improvements that consumers are not
                otherwise seeking. Section IV.B.3 discusses at some length the fact
                that alternative powertrains and higher fuel-efficiency vehicle models
                have proliferated widely since 2011--consumers no longer lack for
                choice if fuel economy is what they want. NHTSA's concern regarding
                consumer demand is that in an era of relatively low gasoline prices--as
                EIA currently projects and NHTSA has no basis to second-guess, and
                which may be even lower than currently projected--it does not appear
                likely that the market for higher fuel-economy vehicles and alternative
                powertrains in particular will increase significantly in the rulemaking
                timeframe, beyond the 30-month payback period that the agencies
                currently use as a proxy for market demand for fuel economy. It is
                worth citing the CAS case at greater length here in light of its
                parallels: As the D.C. Circuit stated in that case,
                 [T]he petitioners do not challenge the consideration of consumer
                demand per se, but rather the weight the agency has given the factor
                in downgrading standards when, they argue, the principal
                impracticability is paying a civil penalty [note: today, using or
                purchasing credits]. Until the model years at issue here, there has
                been little tension between consumer demand and the fuel
                conservation goals of EPCA. The agency now relies on market
                projections in a setting in which falling gas prices have relaxed
                consumer demand for fuel efficiency. Earlier consideration of
                consumer demand in setting standards could not have alerted Congress
                to the agency's current application of this factor. Because Congress
                has not spoken clearly on the issue before us, it must be determined
                whether the agency's interpretation represents a reasonable
                accommodation of the policies embodied in the statute.
                 . . .
                 The agency concluded that if manufacturers had to restrict the
                availability of larger trucks and engines in order to adhere to CAFE
                standards, the effects ``would go beyond the realm of `economic
                practicability' as contemplated in the Act.'' [Citation omitted.]
                The original projections of technological feasibility for the 1985
                model year standards were based on the assumption that gasoline
                prices would remain high and consumer demand for fuel-efficient
                vehicles would remain strong. No one disputes that actual
                circumstances have deviated from these assumptions. NHTSA acted
                within the reasonable range of interpretations of the statute in
                correcting the 1985 standards to
                [[Page 25182]]
                account for these changed conditions. Consideration of product mix
                effects was also reasonable in setting the standards for 1986, as
                there is no evidence that the same trends in consumer demand will
                not continue.
                 . . .
                 In short, while it may be disheartening to witness the erosion
                of fuel conservation measures in the face of changes in consumer
                priorities, this court is nonetheless compelled to uphold the
                agency's standards. They are the result of a balancing process
                specifically committed to the agency by Congress, and, in this case,
                the weight given to consumer demand was not outside the range
                permitted by EPCA.
                 CAS, 793 F.2d 1322, 1340-41 (D.C. Cir. 1986). As in the situation
                presented in the CAS case, the agencies believed in 2012 based on the
                evidence then before them that fuel prices would be significantly
                higher than the fuel prices currently projected today. Using the fuel
                prices currently projected, which are lower because of the structural
                changes to the global oil market described at length above, Figure
                VIII-3 shows the difference in annual fuel consumption for a typical
                driver under the augural standards, proposed standards, and final
                standards. As the figure shows, the difference in annual consumption
                (for a user that drives 14K miles per year) \2985\ is fewer than 40
                gallons by MY 2030--the largest difference between the alternatives.
                Rising fuel prices over time increase the value of those forty gallons,
                but the diminishing returns to successive increases in fuel economy are
                nonetheless evident.\2986\
                ---------------------------------------------------------------------------
                 \2985\ Parts of the central analysis assume a typical new
                vehicle is driven 14,000 miles per year, for each of the first three
                years it is owned. In practice, the average is slightly higher,
                through affected by a smaller number of users that drive much more
                than average. There is no single value that is representative of all
                households, and the National Household Travel Survey has shown lower
                annual usage estimates than 14,000 miles per year for a typical new
                vehicle.
                 \2986\ In general, because fuel savings are subject to
                diminishing returns as CAFE standards become more stringent, and
                per-vehicle costs increase as CAFE standards become more stringent,
                the relationship between per-vehicle costs and the value of fuel
                savings is more of a curve than a line, although the slope of the
                curve is reduced by the fact that we rely on EIA's forecast of
                rising fuel prices over time.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.743
                 Thus, on the supply side, greater and more stable global oil
                supply, which reduces projected fuel prices, means that the benefits of
                more stringent CAFE standards are lower than they appeared to be in
                2012 when the agencies believed oil supply would be scarcer and less
                stable, and projected fuel prices were consequently higher.
                 On the demand side, as already explained, while NHTSA agrees that
                some consumers do seek out higher fuel economy, those consumers have
                vastly more higher fuel-economy-vehicle options than they did when the
                agencies wrote the 2012 final rule, as shown in Section IV above. For
                the other consumers who are driven more by the economics of their
                vehicle-purchasing decisions, NHTSA believes that they are likely
                making reasonably informed decisions about the new vehicle attributes
                they want in light of expectations about future fuel costs. This can be
                illustrated by examining estimated payback periods under the different
                regulatory alternatives, because payback period directly compares
                estimated future fuel savings with estimated vehicle purchase and
                ownership costs. A number of commenters suggested that per-vehicle cost
                was not a meaningful metric in isolation, because consumers would also
                be saving money on fuel under more stringent standards. The agencies
                discuss affordability issues further below, but the rulemaking presents
                Table VIII-12 here as a comparison of per-vehicle costs to lifetime
                fuel savings to illustrate the point raised by commenters:
                [[Page 25183]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.744
                 Table VIII-12 shows the differences in regulatory costs, other
                registration costs (taxes and financing, though the cost of insurance
                also increases to cover more expensive vehicles), lifetime fuel
                savings, and the payback relative to a MY 2017 vehicle. It is important
                to compare apples to apples, so in this case, because the agencies are
                considering fuel costs over a vehicle's full lifetime, this rulemaking
                needs to compare that against a broader lifetime cost of ownership,
                instead of comparing it simply to the estimated increase in initial
                purchase price. Under the augural standards, the analysis projects that
                it would take a full five years for the undiscounted value of fuel
                savings to offset the estimated upfront increase in purchase cost
                (relative to a MY 2017 vehicle). For reference, the average new car
                buyer holds on to that car for about six or seven years.\2987\
                Naturally, this payback period, and the fuel savings on which it is
                based, depend upon fuel prices. Higher fuel prices shorten payback
                periods, while declining fuel prices lengthen them. For this analysis,
                the agencies have employed fuel prices estimated using the version of
                NEMS used to produce AEO 2019, as discussed in Section VI.
                ---------------------------------------------------------------------------
                 \2987\ IHS Markit estimates the average length of new vehicle
                ownership at about 79 months, see https://www.forbes.com/sites/jimgorzelany/2018/01/12/the-long-haul-15-vehicles-owners-keep-for-at-least-15-years/#4e971b576237.
                ---------------------------------------------------------------------------
                 Thus, all of the regulatory alternatives considered in today's
                analysis result in significantly longer payback periods than the 2.5
                years assumed by the agencies, the industry, and the NAS--i.e., while
                fuel economy would foreseeably improve in the rulemaking timeframe in
                the absence of regulation, it would do so at a rate slower even than
                the proposal would have required.\2988\ NHTSA thus does not expect that
                consumer demand for fuel-efficient vehicles will grow significantly in
                the rulemaking timeframe without regulation to prop up manufacturer
                sales of significantly larger volumes of more fuel-efficient models.
                This increases the economic practicability of regulatory alternatives
                that represent less stringent standards, as compared to those that
                represent more stringent standards.
                ---------------------------------------------------------------------------
                 \2988\ While presented at the industry level, technology
                application and compliance simulation occur at the level of each
                individual manufacturer's respective fleets. Some OEMs and fleets
                are able to increase CAFE more easily than others--starting from
                more favorable positions and adding less expensive technology, or
                taking advantage of credit provisions, to improve the fuel economy
                of their fleets. However, for several OEMs, even the proposed
                standards are binding, and the costs associated with bringing their
                fleets into compliance are significant. At the level of the industry
                average, the cost of compliance with the proposal--and as a
                corollary, with the other alternatives--exceeds the 2.5 year payback
                for fuel economy technology, even while a small amount of
                overcompliance occurs at the industry level.
                ---------------------------------------------------------------------------
                (d) Manufacturer Compliance Shortfalls as Part of Economic
                Practicability
                 Manufacturer compliance shortfalls given the pace of increase in
                standard stringency over time are also relevant to economic
                practicability, and were considered as part of the 2012 final rule.
                Some commenters argued that it was not reasonable for NHTSA to
                interpret automakers' fuel economy improvements over time as evidence
                that less stringent standards might be maximum feasible, suggesting
                that evidence of improvements means that improvements are possible, and
                that automakers' stated difficulties with meeting more stringent
                standards may be overstated. Fleet fuel economy improvements over time
                have been possible, NHTSA agrees. NHTSA does not agree, however, that
                improvements thus far constitute de facto evidence of automakers'
                ability to meet rapidly increasing standards indefinitely into the
                future. Section IV above illustrates this clearly--many more very fuel-
                efficient models are available now than in 2012, while fuel prices have
                been trending downward on an absolute basis over the same time period.
                Simultaneously and relatedly, the rate at which various manufacturer
                fleets have been falling short of their standards has been increasing
                steadily. As Section IV explains, at the time of the 2012 analysis,
                most manufacturers were in reasonable shape in terms of compliance. The
                total fleet outperformed CAFE standards by a full mile per gallon--
                reflecting the historical trend that the full fleet always exceeds
                [[Page 25184]]
                the average fuel economy target.\2989\ Of the then 45 import passenger
                car, domestic passenger car, and light truck compliance fleets in the
                2012 model year, 26 of the fleets exceeded their fuel economy targets,
                while 19 failed to meet their standard.\2990\ Of those 19 fleets that
                failed to meet their standard, the total shortfall was 41,033,802
                credits--the equivalent of $225,685,911 in penalties.\2991\ That is no
                longer the case. 2016 marked the first model year in CAFE history that
                the entire light duty fleet failed to meet its target.\2992\ This
                continued in the 2017 model year (the most recent full model year of
                compliance data).\2993\ In the 2017 model year, of the now 42
                compliance fleets, only 14 fleets exceeded their targets.\2994\ 25
                failed to meet their target, with a total shortfall of 166,715,863
                credits--the equivalent of $1,133,430,584 in penalties.\2995\ Required
                manufacturer reporting data shows the situation continuing to get worse
                in the 2018 and 2019 model years,\2996\ despite manufacturers'
                increasing ability to utilize generous credit provisions related to
                alternative fueled vehicles and A/C efficiency and off-cycle
                adjustments.
                ---------------------------------------------------------------------------
                 \2989\ Data from CAFE Public Information Center (PIC), https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Home.htm, last accessed Dec. 27,
                2019.
                 \2990\ NHTSA MY 2011-2019 Industry CAFE Compliance, https://one.nhtsa.gov/cafe_pic/MY%202011-MY_2019_Credit_Shortfall_Report_v08.pdf.
                 \2991\ Id. While we denominate shortfalls in terms of credits,
                that is simply for convenience, and any given manufacturer's
                shortfall is measured in tenths of a mile per gallon for compliance
                purposes.
                 \2992\ Data from CAFE Public Information Center (PIC), https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Home.htm, last accessed Dec. 27,
                2019.
                 \2993\ Id.
                 \2994\ NHTSA MY 2011-2019 Industry CAFE Compliance, https://one.nhtsa.gov/cafe_pic/MY%202011-MY_2019_Credit_Shortfall_Report_v08.pdf.
                 \2995\ Id.
                 \2996\ Id.
                ---------------------------------------------------------------------------
                 Although each year has continued to see improvements in fuel
                economy performance, each successive increase in stringency requires
                many fleets not only to achieve the new level from the resulting
                increase, but to resolve deficits from the prior year as well. The
                problem is particularly marked in the light truck fleet, where sales of
                lower fuel-economy vehicles have proliferated over this time period,
                despite availability of higher fuel-economy models. But the passenger
                car fleet is facing compliance challenges as well, as more consumers
                have shifted away from sedans and into crossover utility vehicles that
                are considered passenger cars for compliance purposes. While the
                agencies' move toward footprint based standards account for vehicle
                length and track width--which certainly affect fuel economy as
                described above--they do not account for mass-intensive increases in
                vehicle ride height that crossover purchasers value, the additional
                frontal area and higher drag at highway speeds, or the additional power
                required to achieve similar performance as the equivalent sedan. These
                issues are further exacerbated by the fact that consumers are demanding
                more powerful engines than the baseline efficient four cylinder
                versions the agencies assumed consumers would find acceptable, instead
                opting to upgrade to more powerful powertrains.\2997\ If the augural
                standards were finalized and energy prices remain as currently
                projected, the shortfall situation could well erase large portions of
                assumed fuel savings/emissions reduction benefits from higher
                standards.
                ---------------------------------------------------------------------------
                 \2997\ Mr. Rykowski's comments for EDF, for example, stated that
                EPA's recent Fuel Economy and CO2 Trends Reports show
                clearly that manufacturers have been improving vehicle performance
                at the expense of fuel economy. See NHTSA-2018-0067-12018, at 31.
                ---------------------------------------------------------------------------
                 In the current analysis, gasoline prices are projected to rise
                steadily from about $2.50/gallon in 2017 to $3.5/gallon by 2035. While
                CAFE can provide some insurance against unexpected and sudden price
                increases, in the case of sustained, consistent increases in gasoline
                prices, market demand for fuel economy would outpace the standards over
                time. In an earlier analysis, the agencies considered the impact of a
                sudden gasoline price shock in a single year, where the price of
                gasoline jumped from $3.50/gallon to $6/gallon for most of a year. If
                instead of that one-year spike, the price of gasoline rose steadily
                from current levels to $6/gallon by 2040, the response of both
                consumers and manufacturers in the marketplace would cause the industry
                to consistently over-comply with even the augural standards.\2998\ The
                payback assumption in this analysis, where consumers are willing to pay
                for any fuel economy improvement that pays for itself in the first 2.5
                years of vehicle usage, would likely be too short in a world with $6/
                gallon gasoline, where the cost of operating a vehicle consumed a
                larger share of a household's budget and even longer payback periods
                could be seen as sound investments. Thus, if it turns out that fuel
                prices rise steadily over the next decade, at a significantly faster
                rate than currently projected, the market will end up demanding more
                efficient vehicles and the gap between the baseline and the preferred
                alternative will shrink further. However, the agencies do not currently
                have information that projects $6/gallon fuel in 2040 is likely, for
                the reasons discussed at length above.
                ---------------------------------------------------------------------------
                 \2998\ We simulated this response in the CAFE Model, where all
                other inputs were identical to the central analysis.
                ---------------------------------------------------------------------------
                 As also discussed above, while the analysis for this final rule
                does not show significant shortfalls under any regulatory alternative,
                that appearance of compliance is predicated on the assumption that
                automakers will be able to sell the hybrids that we simulate them
                producing in response to the standards. Again, given foreseeably low
                fuel prices going forward, it is also foreseeable that selling greater
                volumes of hybrid vehicles will be even more difficult than at present.
                It is very possible that manufacturer compliance shortfalls could end
                up being worse than the agency's analysis currently forecasts for the
                more stringent alternatives.
                 Given the ongoing shortfall problem illustrated above, and given
                the payback period estimates, the proposal might appear to be the
                correct answer in the absence of other considerations. NHTSA believes
                that the bubble concerns may be significant, and the diminishing
                returns of higher standards identified in Section IV above calls into
                question the value of pushing that bubble. Compliance shortfalls
                represent a growing problem with the current standards and will
                continue to be a problem if stringency does not converge at least
                somewhat more closely with what the market appears willing to bear. If
                industry is unable to comply with standards, that non-compliance means
                that the standards are not achieving what they set out to achieve in
                terms of fuel savings or emissions reductions, or at least they are not
                achieving what NHTSA estimated they would achieve. The NPRM disagreed
                with the idea that ``if you build it, they will come''--that
                manufacturers would find a way to market higher fuel-economy vehicles,
                and consumers would eventually buy them. Comments on that topic were
                mixed: some commenters agreed with the NPRM's sentiment, while other
                commenters argued that manufacturers' past ability to exceed standards
                combined with consumers' growing interest in fuel economy/lower
                emissions meant that concerns about the market's ability to bear
                further increases were misplaced. The shortfall discussion above and in
                Section IV suggests that the NPRM's sentiment may be accurate, but this
                difference in perspective highlights the core philosophical question of
                the CAFE program--whether consumers should choose for themselves how
                much fuel
                [[Page 25185]]
                economy they want, or whether the government should choose for them.
                (4) Considering Safety Along With the Other Factors in Determining
                Maximum Feasible Standards
                 In addition to the above, as explained in the NPRM and as discussed
                extensively by commenters, NHTSA considers safety effects in
                determining maximum feasible CAFE standards. A number of commenters
                objected to aspects of the safety analysis, as discussed in Section VI
                above, and some made suggestions for improvement. In response to those
                comments, NHTSA took a very conservative approach in making a number of
                changes to the safety analysis for this final rule:
                 Commenters disagreed with certain aspects of the sales and
                scrappage effects on the safety analysis; in response to those
                comments, changes have been made and the scrappage effect on fatalities
                is lower now than it was in the NPRM;
                 Commenters disagreed with certain aspects of mass
                reduction; in response to those comments, changes have been made there;
                 Commenters argued that additional technologies should be
                accounted for; in response to those comments, many of those
                technologies have been added;
                 Commenters argued that the NPRM did not account for crash
                avoidance technologies; in response to those comments, the final rule
                accounts for the effects of crash avoidance technologies;
                 Commenters argued that the NPRM did not account for the
                mortality/morbidity effects of criteria pollution differences between
                the alternatives; in response, the final rule accounts for these
                effects explicitly in these values.
                 Overall, the final rule analysis suggests that fatalities may be
                lower than the NPRM analysis showed; injuries may be greater; and the
                safety effects overall are less than the NPRM suggested, but they are
                still significant. Less-stringent standards remain better for safety
                and are projected to save thousands of lives and prevent tens of
                thousands of hospitalizations, even if the amount by which they are
                better is lower than previously estimated.
                 EPCA/EISA directs NHTSA to conserve energy and consider the need of
                the U.S. to conserve energy, while simultaneously directing NHTSA to
                set attribute-based standards whose outcome varies depending on what
                consumers choose to buy, and directing NHTSA to consider economic
                practicability. The greater the need of the U.S. to conserve energy,
                the more the government should decide for consumers how much fuel
                economy will be in their new vehicles. Based on the information before
                NHTSA in this final rule, NHTSA agrees with the commenters who
                suggested that increasing CAFE stringency can function as ``insurance''
                against future oil price volatility, although as illustrated above, the
                short-term effects of that insurance may be relatively minor and the
                longer-term effects may be too uncertain to consider meaningfully.
                NHTSA also agrees that environmental considerations necessitate energy
                conservation, though the long-term benefits of emissions reductions
                (even accounting for the increased costs of delayed action) require
                consideration of the immediate costs to consumers, the industry, and
                the environment.
                 Balancing all of the factors and issues identified above, NHTSA
                concludes that standards that increase at 1.5% per year are the maximum
                feasible for passenger cars and light trucks for MYs 2021-2026, based
                on the information currently before the agency. We recognize that more
                stringent standards, including the baseline/augural standards, could
                conserve more energy and might be technologically feasible (in the
                narrowest sense), but the additional incremental fuel savings,
                emissions reductions, and environmental benefits of higher standards is
                not significant enough to outweigh the immediate economic costs. There
                is still risk to the U.S. from circumstances outside our control that
                the CAFE program may be able to mitigate, but there must also be
                recognition of the limited extent to which this program can address
                that risk, certainly without exacerbating considerable challenges
                currently being faced by automakers, dealers, and consumers. Economic
                practicability would be best served by slower increases, as discussed
                above. And while these two factors weigh in different directions, NHTSA
                has discretion to accommodate conflicting statutory priorities in a
                reasonable manner. Beginning with MY 2021, the first MY addressed by
                this rule, Congress eliminated the obligation to increase FE standards
                ratably.\2999\ Thus, the appropriateness of an increase, if any, is
                within NHTSA's discretion based on its balancing of statutory
                factors.\3000\
                ---------------------------------------------------------------------------
                 \2999\ Previously applied for MYs 2011-2020.
                 \3000\ NHTSA also notes that it was expressly anticipated in the
                2012 final rule that the current rulemaking could determine that the
                augural standards were not maximum feasible. NHTSA stated that
                ``Whether different alternatives may be maximum feasible can also be
                influenced by differences and uncertainties in the way in which key
                economic factors (e.g., the price of fuel and the social cost of
                carbon) and technological inputs could be assessed and valued. While
                NHTSA believes that our analysis for this final rule uses the best
                and most transparent technology-related inputs and economic
                assumption inputs that the agencies could derive for MYs 2017-2025,
                we recognize that there is uncertainty in these inputs, and the
                balancing could be different if the inputs were different. When the
                agency undertakes the future rulemaking to develop final standards
                for MYs 2022-2025, for example, we expect that much new information
                will inform that future analysis, which may potentially lead us to
                choose different standards than the augural ones presented today.''
                (emphasis added) 77 FR at 63037 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 In past rulemakings, as discussed above, NHTSA has expressly
                considered the point at which net benefits appear to be maximized as
                potentially relevant to determining maximum feasible CAFE
                standards.\3001\ Whether the standards maximize net benefits has thus
                been a significant, but not dispositive, factor in the past for NHTSA's
                consideration of economic practicability. Executive Order 12866, as
                amended by Executive Order 13563, states that agencies should ``select,
                in choosing among alternative regulatory approaches, those approaches
                that maximize net benefits . . .'' In practice, however, NHTSA must
                consider that the modeling of net benefits does not capture all
                considerations relevant to the EPCA statutory factors. Additionally,
                nothing in EPCA or EISA mandates that NHTSA set standards at the point
                at which net benefits are maximized, and case law confirms that whether
                to maximize net benefits in determining maximum feasible standards is
                within NHTSA's discretion.\3002\ As explained extensively in prior
                rulemakings, even if the agency believed it could quantify enough
                relevant factors to determine the CAFE
                [[Page 25186]]
                levels at which net benefits were maximized with reasonable accuracy,
                there may be other considerations which lead the agency to conclude
                that maximum feasible CAFE standards are not the ones that maximize net
                benefits. For example, in 2012, NHTSA rejected the regulatory
                alternative that appeared to maximize net benefits (and all
                alternatives more stringent than that one) based on the conclusion that
                even though net benefits were maximized, the ``resultant technology
                application and cost'' were simply too high, and thus made those
                standards economically impracticable, and thus beyond maximum
                feasible.\3003\
                ---------------------------------------------------------------------------
                 \3001\ See, e.g., the 2006 final rule, which concluded that the
                point at which net benefits were maximized was the maximum feasible
                CAFE level (71 FR 17566 (Apr. 6, 2006)); the 2010 final rule, which
                considered among the regulatory alternatives one that maximized net
                benefits, but explained that nothing in EPCA or EISA mandated that
                NHTSA choose CAFE standards that maximize net benefits (75 FR 25324,
                at 25606, 25167 (May 7, 2010)); and the 2012 final rule, which also
                considered among the regulatory alternatives one that maximized net
                benefits, and also explained that nothing in EPCA or EISA mandated
                that NHTSA choose CAFE standards that maximize net benefits, in
                fact, directly rejecting the regulatory alternative that maximized
                net benefits as beyond maximum feasible for the MYs 2017-2025
                timeframe (77 FR 62624 (Oct. 15, 2012)).
                 \3002\ The Ninth Circuit has agreed with NHTSA that ``EPCA
                neither requires nor prohibits the setting of standards at the level
                at which net benefits are maximized,'' stating further that ``The
                statute is silent on the precise question of whether a marginal
                cost-benefit analysis may be used. See Chevron, 467 U.S. at 843, 104
                S.Ct. 2778. Public Citizen and Center for Auto Safety persuade us
                that NHTSA has discretion to balance the oft-conflicting factors in
                49 U.S.C. 32902(f) when determining ``maximum feasible'' CAFE
                standards under 49 U.S.C. 32902(a).'' CBD v. NHTSA, 538 F.3d 1172,
                1188 (9th Cir. 2008).
                 \3003\ 77 FR at 63050 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 Table VII-95 and Table VII-96, above, appear to suggest that net
                benefits would be maximized under a 3 percent discount rate by choosing
                the 2%/3% alternative, and under a 7 percent discount rate by choosing
                the 0% (proposed) alternative. Across all alternatives under either
                discount rate, the variation in net benefits is within $20 billion over
                the lifetimes of vehicles produced during the rulemaking timeframe.
                While $20 billion may seem like a large amount of money, it must be
                understood within context--the auto industry accounted for
                approximately $89 billion of U.S. GDP in 2018 alone,\3004\ and
                Americans spent approximately $370 billion on gasoline in 2019
                alone.\3005\ For a program this large, if the difference between the
                net benefits created by different regulatory alternatives is within $20
                billion (over the full lifetimes of six model years), the net benefits
                are relatively small. Furthermore, given how close together the net
                benefits are across the range of regulatory alternatives considered,
                NHTSA does not believe that the point at which net benefits are
                maximized is meaningful for determining maximum feasible CAFE standards
                in this final rule.
                ---------------------------------------------------------------------------
                 \3004\ See Bureau of Economic Analysis, GDP by Industry, ``Value
                Added by Industry,'' Oct. 29, 2019, https://apps.bea.gov/iTable/iTable.cfm?ReqID=51&step=1 (accessed Mar. 18, 2020)
                 \3005\ Using EIA estimates of an average of $2.60/gallon
                gasoline cost in 2019 (https://www.eia.gov/todayinenergy/detail.php?id=42435) and EIA estimates of about 142 billion gallons
                total gasoline consumed (https://www.eia.gov/tools/faqs/faq.php?id=23&t=10).
                ---------------------------------------------------------------------------
                 Important to that conclusion is the fact that the net benefits
                estimates produced by the analysis depend heavily on EIA's future
                forecasts of fuel prices, which were made prior to the recent collapse
                of oil prices. If the former OPEC+ members continue to pursue market
                share, fuel prices will likely continue to drop. If, instead of
                pursuing market share, they try to control prices by restricting
                supply, U.S. shale production can ramp back up and exert downward
                pressure on price. If fuel prices end up even lower than our analysis
                assumes, benefits from saving additional fuel will be worth even less
                to consumers. Our analysis captures none of these effects. Depending
                upon future fuel prices, net benefits estimates described above could
                foreseeably be overstated, possibly by a significant amount. It is
                possible, depending on future fuel prices, that the final rule 1.5
                percent annual increase standards could end up being more stringent
                than standards that would maximize net benefits. Moreover, sustained
                low oil prices can be expected to have real effects on consumer demand
                for additional fuel economy, which will have real effects on sales,
                jobs, and many other things relevant to NHTSA's consideration of what
                standards would be maximum feasible. Choosing a regulatory alternative
                more stringent than the final rule's 1.5 percent annual increases could
                foreseeably either lead to more hybridization than the market is likely
                to bear given foreseeably low fuel prices, or lead to significantly
                more cost than the analysis currently suggests. Neither of those
                outcomes would be beneficial for consumers or for industry, even
                considering the additional fuel savings for consumers.\3006\
                ---------------------------------------------------------------------------
                 \3006\ It is within NHTSA's discretion to adopt an alternative
                based on unquantified/unquantifiable benefits. See, e.g., Inv. Co.
                Inst. v. Commodity Futures Trading Comm'n, 720 F.3d 370, 379 (D.C.
                Cir. 2013) (``The appellants further complain that CFTC failed to
                put a precise number on the benefit of data collection in preventing
                future financial crises. But the law does not require agencies to
                measure the immeasurable. CFTC's discussion of unquantifiable
                benefits fulfills its statutory obligation to consider and evaluate
                potential costs and benefits. See Fox, 556 U.S. at 519, 129 S.Ct.
                1800 (holding that agencies are not required to `adduce empirical
                data that' cannot be obtained). Where Congress has required
                `rigorous, quantitative economic analysis,' it has made that
                requirement clear in the agency's statute, but it imposed no such
                requirement here. American Financial Services Ass'n v. FTC, 767 F.2d
                957, 986 (DCCir.1985); cf., e.g., 2 U.S.C. 1532(a) (requiring the
                agency to `prepare a written statement containing . . . a
                qualitative and quantitative assessment of the anticipated costs and
                benefits' that includes, among other things, `estimates by the
                agency of the [rule's] effect on the national economy').'');
                BellSouth Corp. v. FCC, 162 F.3d 1215, 1221 (D.C. Cir.1999) (`When .
                . . an agency is obliged to make policy judgments where no factual
                certainties exist or where facts alone do not provide the answer,
                our role is more limited; we require only that the agency so state
                and go on to identify the considerations it found persuasive').''
                ---------------------------------------------------------------------------
                 NHTSA concludes that steady increases at 1.5 percent annually, with
                the same rate for cars and trucks as suggested by several commenters,
                are the optimal way to move the needle forward on fuel economy, fuel
                savings, and emissions reductions without imposing excessive cost on
                automakers and consumers and overly reducing vehicle sales. Requiring
                demand changes (through CAFE standards) much faster than what the
                market will bear creates a substantial likelihood of a mis-match
                between what companies produce and what consumers buy. While companies
                can manage that mis-match for short periods through incentivization and
                cross-subsidization, we have seen that over time automakers begin to
                fall short on fuel economy performance relative to the standards. Over
                time, if swaths of the industry continually fall short of fuel economy
                targets, and consumer demand for fuel economy does not significantly
                increase, then continuing to force technology into the fleet does not
                achieve the program's objectives (i.e., energy conservation). This is
                the case regardless of how much manufacturers spend manufacturing
                vehicles that consumers do not purchase (implicating concerns with
                economic practicability) to reduce their compliance liability. This is
                one part of why NHTSA believes that the 1.5 percent alternative is
                maximum feasible during the rulemaking timeframe.
                 While the 1.5 percent alternative being finalized is new for the
                final rule, it is responsive to comments requesting steady increases at
                the same rate for both cars and trucks, and it is within the range of
                rates of increase considered in the NPRM. As both the NPRM analysis and
                the final rule analysis show, after MY 2020 the proposed (0%) standards
                are not binding at the industry level (though some manufacturers, and
                fleets, remain below their standard after that model year) as a
                consequence of market demand for fuel economy at projected gasoline
                prices. However, the preferred (1.5% percent) alternative, while
                producing slightly higher achieved CAFE levels, tracks closely to the
                level produced by the combination of existing CAFE standards (through
                MY 2020) and subsequent market demand for fuel economy represented by
                the proposal. It is also likely close to the point at which net
                benefits will be maximized, even if it remains unclear exactly where
                that point will end up.
                 As a kind of insurance policy against future fuel price volatility,
                standards that increase at 1.5 percent per year for cars and trucks
                will help to keep fleet fuel economy higher than they would be
                otherwise when fuel prices are low, which is not improbable over the
                next several years.\3007\ These standards will
                [[Page 25187]]
                also enable industry to choose how to spend the capital that would
                otherwise be spent meeting more stringent standards on more of what
                consumers are demanding, which could also include more fuel economy if
                the market heads unexpectedly in that direction. As explained above,
                even if more stringent standards might be technologically feasible in a
                narrow sense, and even if the effect of other motor vehicle standards
                of the Government does not vary significantly between regulatory
                alternatives, economic practicability concerns still counsel against
                more stringent standards, and the need of the U.S to conserve energy
                does not, at present, appear to counsel toward higher stringency.
                Standards that increase at 1.5 percent per year represent a reasonable
                balance of additional technology and required per-vehicle costs,
                consumer demand for fuel economy, fuel savings and emissions avoided
                given the foreseeable state of the global oil market and the minimal
                effect on climate between finalizing 1.5 percent standards versus more
                stringent standards. The final standards will also result in year-over-
                year improvements in fleetwide fuel economy, resulting in energy
                conservation that helps address environmental concerns, including
                criteria pollutant, air toxic pollutant, and carbon emissions. All
                things considered, NHTSA determines that an increase of 1.5 percent per
                year is maximum feasible for both passenger cars and light trucks for
                MYs 2021-2026.
                ---------------------------------------------------------------------------
                 \3007\ For example, EIA currently expects U.S. retail gasoline
                prices to average $2.14/gallon in 2020, compared to $2.69/gallon in
                2019 (see https://www.eia.gov/outlooks/steo/archives/mar20.pdf), and
                $3.68/gallon in 2012 (see https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=PET&s=EMM_EPM0_PTE_NUS_DPG&f=A). While gasoline
                prices may foreseeably rise over the rulemaking time frame, it is
                also very foreseeable that they will not rise to the $4-5/gallon
                that many American saw over the 2008-2009 time frame, that caused
                the largest shift seen toward smaller and higher-fuel-economy
                vehicles. See, e.g., Figure VIII-2 above.
                ---------------------------------------------------------------------------
                Compliance and Enforcement
                A. Introduction
                1. Overview
                 The CAFE and CO2 emissions standards are both fleet-
                average standards, and for both programs, determining compliance begins
                by testing vehicles on dynamometers in a laboratory over pre-defined
                test cycles under controlled conditions.\3008\ A machine is connected
                to the vehicle's tailpipe while it performs the test cycle, which
                collects and analyzes the resulting exhaust gases; a vehicle that has
                no tailpipe emissions has its performance measured differently, as
                discussed below. CO2 quantities, as one of the exhaust
                gases, can be evaluated for vehicles that produce CO2
                emissions directly. Fuel economy is determined from the amount of
                CO2 emissions, because the two are directly mathematically
                related.\3009\ Manufacturers generally perform their own testing, and
                EPA confirms and validates those results by testing a sample of
                vehicles at the National Vehicle and Fuel Emissions Laboratory (NVFEL)
                in Ann Arbor, Michigan. The results of this testing form the basis for
                determining a manufacturer's compliance in a given model year, through
                the following steps:
                ---------------------------------------------------------------------------
                 \3008\ For readers unfamiliar with this process, it is similar
                to running a car on a treadmill following a program--or more
                specifically, two programs. 49 U.S.C. 32904(c) states that, in
                testing for fuel economy, EPA must ``use the same procedures for
                passenger automobiles [that EPA] used for model year 1975 (weighted
                55 percent urban cycle and 45 percent highway cycle), or procedures
                that give comparable results.'' Thus, the ``programs'' are the
                ``urban cycle,'' or Federal Test Procedure (abbreviated as ``FTP'')
                and the ``highway cycle,'' or Highway Fuel Economy Test (abbreviated
                as ``HFET''), and they have not changed substantively since 1975.
                Each cycle is a designated speed trace (of vehicle speed versus
                time) that vehicles must follow during testing--the FTP is meant
                roughly to simulate stop and go city driving, and the HFET is meant
                roughly to simulate steady flowing highway driving at about 50 mph.
                The 2-cycle dynamometer test results differ somewhat from what
                consumers will experience in the real world driving environment
                because of the lack of high speeds, rapid accelerations, and hot and
                cold temperatures evaluations with the A/C operation. These added
                conditions are more so reflected in the EPA 5-cycle test results
                listed on each vehicle's fuel economy label and on the
                fueleconomy.gov website.
                 \3009\ Technically, for the CAFE program, carbon-based tailpipe
                emissions (including CO2, CH4, and CO) are
                measured, and fuel economy is calculated using a carbon balance
                equation. EPA uses carbon-based emissions (CO2,
                CH4, and CO, the same as for CAFE) to calculate the
                tailpipe CO2 equivalent for the tailpipe portion of its
                standards.
                ---------------------------------------------------------------------------
                 Each vehicle model's performance on the test cycles is
                calculated;
                 The number of vehicles of that model that were produced is
                divided by the performance;
                 That number, in turn is summed for all the manufacturer's
                model types;
                 The manufacturer's total product volume is then divided by
                the summed value of all the model types; and
                 That number represents the manufacturer's fleet harmonic
                average performance.
                 That performance is then compared to the manufacturer's unique
                compliance obligation (standard). This compliance obligation is
                calculated using the same approach that is used to determine
                performance, except that the fuel economy or CO2 target
                value (based on the footprint of each vehicle model) is used instead of
                the model's measured performance value. The fuel economy or
                CO2 target values for each of the vehicle models in the
                manufacturer's fleet and production volumes are used to derive the
                manufacturer's fleet harmonic average standard. Using fuel economy
                targets to illustrate the concept, the following figure shows two
                vehicle models produced in a model year for which passenger cars are
                subject to a fuel economy target function that extends from about 30
                mpg for the largest cars to about 41 mpg for the smallest cars:
                [[Page 25188]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.745
                 If these are the only two vehicle models the manufacturer produces,
                the manufacturer's required CAFE obligation is determined by
                calculating the production-weighted harmonic average of the fuel
                economy target values applicable at the hatchback and sedan footprints
                (from the curve, about 41 mpg for the hatchback and about 33 mpg for
                the sedan). The manufacturer's achieved CAFE level is determined by
                calculating the production-weighted harmonic average of the hatchback
                and sedan fuel economy levels (in this example the values shown in the
                boxes in Figure IX-1, 48 mpg for the hatchback and 25 mpg for the
                sedan). Depending on the relative mix of hatchbacks and sedans the
                manufacturer produces, the manufacturer's fleet may meet the standard,
                or perform better than the standard (if required CAFE is less than
                achieved CAFE) and thereby earn credits or perform worse than the
                standard (if required CAFE is greater than achieved CAFE) and thereby
                have a shortfall that may be made up, in whole or in part, using CAFE
                credits, discussed below, or be subject to civil penalties. Although
                the arithmetic is different for CO2 standards (which do not
                involve harmonic averaging), the underlying concept is the same.
                 There are thus two parts to the foundation of compliance with CAFE
                and CO2 emissions standards: First, how well any given
                vehicle model performs relative to its target, and second, how many of
                each vehicle model a manufacturer produces. While no given model need
                precisely meet its target (and virtually no model exactly meets its
                target in the real-world), if a manufacturer finds itself producing
                large numbers of vehicles that fall well short of their targets, it
                will have to find a way of offsetting that shortfall, either by
                increasing production of vehicles that exceed their targets, or by
                taking advantage of compliance flexibilities and incentives, or the
                manufacturer will be subject to civil penalties. Given that
                manufacturers typically need to produce for sale vehicles that
                consumers want to buy, and not all consumers value fuel economy, their
                options for pursuing the former approach can often be limited.
                 The CAFE and CO2 programs both offer a number of
                compliance flexibilities and incentives, discussed in more detail
                below. For example, starting in model year 2017, manufactures have
                flexibility to account for efficiency improvements in air conditioning
                (A/C) systems and/or for the application fuel economy improving
                technologies that increase fuel economy in the real-world, but that
                are, in whole or in part, not accounted for (e.g., stop-start
                technology, or high efficiency alternators) using the 1975-based 2-
                cycle compliance dynamometer test procedures.\3010\ These fuel economy
                improvements are added to the 2-cycle performance results and are
                included in the calculation of a manufacturer's fuel economy in
                determining compliance relative to standards. In addition, for MYs
                2017--2021, there are also two levels of compliance incentives for
                full-size pickup trucks with mild-HEV or strong-HEV technology or that
                overperform standards by 15 percent or more, or by 20 percent or
                more.\3011\ This final rule removes this incentive starting in MY 2022,
                as discussed in more detail below. These fuel economy improvements are
                also included, for those model years and as earned, in the
                [[Page 25189]]
                calculation of a manufacturer's fuel economy.\3012\
                ---------------------------------------------------------------------------
                 \3010\ EPA regulations provided an equivalent program beginning
                in MY 2012.
                 \3011\ Manufacturers also must apply the technology to a minimum
                percentage of their full-size pickup truck production.
                 \3012\ NHTSA characterizes any programmatic benefit
                manufacturers can use to comply with CAFE standards that fully
                accounts for fuel use as a ``flexibility'' (e.g., credit trading)
                and any benefit that counts less than the full fuel use as an
                ``incentive'' (e.g., adjustment of alternative fuel vehicle fuel
                economy). NHTSA flexibilities and incentives are discussed further
                in Section IX.D.
                ---------------------------------------------------------------------------
                 Some flexibilities and incentives are expressly provided for by
                statute, and some have been implemented by the agencies through
                regulations, consistent with the statutory scheme. Compliance
                flexibilities and incentives for the CAFE and CO2 programs
                have a great deal of theoretical attractiveness: If designed properly,
                they can help to reduce overall regulatory costs, while maintaining or
                improving programmatic benefits. If designed poorly, they may create
                significant potential for market distortion (for instance, when
                manufacturers--in response to an incentive to deploy a particular type
                of technology--produce vehicles for which there is no natural market,
                such vehicles must be discounted in order to sell).\3013\
                Manufacturers' use of compliance flexibilities and incentives requires
                proper governmental and industry collaboration for manufacturers to
                achieve the most effective pathways to compliance.\3014\ Overly-
                complicated flexibility and incentive programs can result in greater
                expenditure of both private sector and government resources to track,
                account for, and manage. Moreover, flexibilities or incentives that
                tend to favor specific technologies could distort the market. By these
                means, compliance flexibilities or incentives could create an
                environment in which entities are encouraged to invest in such favored
                technologies and, unless those technologies are independently supported
                by market forces, encourage rent seeking in order to protect, preserve,
                and enhance profits of companies that seek to take advantage of the
                distortions created by government mandate. Further, to the extent that
                there is a market demand for vehicles with lower CO2
                emissions and higher fuel economy, compliance flexibilities and
                incentives may cause some manufacturers to fall behind the industry's
                pace if they become overly reliant on them rather than simply improving
                the efficiency of their vehicles to meet that market demand.
                ---------------------------------------------------------------------------
                 \3013\ While many manufacturers publicly discuss their
                commitment to certain technologies that reduce CO2
                emissions, consumer interest in them thus far remains low, despite
                often-large financial incentives from both manufacturers and the
                Federal and State governments in the form of tax credits (i.e.,
                natural gas or fuel-cell vehicles). It is questionable whether
                continuing to provide significant compliance incentives for
                technologies that consumers appear not to want is an efficient means
                to achieve either compliance or national goals (see, e.g., Congress'
                phase-out of the AMFA dual-fueled vehicle incentive in EISA, 49
                U.S.C. 32906).
                 \3014\ For these reasons, in this final rule, NHTSA is asking
                manufacturers to provide more detailed information on the new
                incentives allowed for A/C and off-cycle technologies and on credit
                trades for better collaboration in understanding the economic impact
                of these flexibilities and incentives and for the government to
                provide better oversight of the CAFE program.
                ---------------------------------------------------------------------------
                 If standards are maximum feasible levels, as required by statute,
                then the need for extensive compliance flexibilities and incentives
                should be low. The agencies sought comments in the NPRM on whether and
                how each agency's existing flexibilities and incentives might be
                amended, revised, or deleted to avoid the inefficiencies and market
                distortions discussed above. Specifically, comments were sought on the
                appropriate level of compliance flexibility, including credit trading,
                in a program that is correctly designed to be maximum feasible, in
                accordance with the statute. Comments were also sought on whether to
                allow all incentive-based adjustments, except those that are mandated
                by statute, to expire, in addition to other possible simplifications to
                reduce market distortion, improve program transparency and
                accountability, and improve overall performance of the compliance
                programs. The agencies considered comments on those issues in preparing
                the final rule. A summary of all the flexibilities for the CAFE and
                CO2 programs finalized as a part of this final rule is
                provided in Table IX-1 though Table IX-4.
                BILLING CODE 4910-59-P
                [[Page 25190]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.746
                [[Page 25191]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.747
                [[Page 25192]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.748
                [[Page 25193]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.749
                [[Page 25194]]
                BILLING CODE 4910-59-C
                2. Light-Duty CAFE Compliance Data for MYs 2011-2019
                 To understand manufacturers' potential approaches to using
                compliance flexibilities and incentives, CAFE compliance data for MYs
                2011 through 2019 is discussed in this section. NHTSA believes that
                providing these data is important because it gives the public a better
                understanding of current compliance trends and the potential impacts
                that increasing CAFE standards have had on those model years and future
                model years addressed by this rulemaking.
                 NHTSA uses data from CAFE reports submitted by manufacturers to EPA
                or directly to NHTSA to evaluate compliance with the CAFE program. The
                data for MYs 2011 through 2017 include manufacturers' final compliance
                data that have been verified by EPA.\3015\ The data for MYs 2018 and
                2019 include the most recent projections from manufacturers' mid-model
                year and final-model year reports submitted to EPA and NHTSA, as
                required by 49 CFR part 537 and 40 CFR 600.512-12.\3016\ Because the
                projections do not reflect final vehicle production levels, the EPA
                verified final CAFE values may be slightly different than the
                manufacturers' projections. MY 2011 was selected as the start of the
                data because it represents the first compliance model year for which
                manufacturers were permitted to trade and transfer credits.\3017\ MY
                2019 is also important because it shows the projected performance of
                the fleet two years after manufacturers were allowed to use new
                flexibilities and incentives starting in MY 2017 to address increasing
                CAFE standards.
                ---------------------------------------------------------------------------
                 \3015\ The data contain the latest information available from
                manufacturers except certain low volume manufacturers complying with
                standards under 49 CFR part 525.
                 \3016\ MY 2018 data come from information received in
                manufacturers' final reports submitted to EPA according to 40 CFR
                600.512-12 and MY 2019 data come from information received in
                manufacturers' mid-model year CAFE reports submitted to NHTSA
                according to 49 CFR part 537.
                 \3017\ 49 CFR 535.6(c).
                ---------------------------------------------------------------------------
                 Figure IX-2 through Figure IX-5 provide a graphical overview of
                fuel economy performance and standards. Fuel economy performance
                includes three parts: (1) Measured performance, on the 2-cycle
                dynamometer test; (2) performance increases for alternative fueled
                vehicles, under the Alternative Motor Fuels Act of 1988 (AMFA); and (3)
                performance adjustments for improved A/C systems and off-cycle
                technologies.3018 3019 3020 These Figures do not account for
                credits earned or expected to be earned from overcompliance in prior or
                future model years that were used or are available for complying with
                CAFE standards. Graphs are included for the total fuel economy
                performance (the combination of all passenger cars and light trucks
                produced for sale during the model year) as a single fleet, and for
                each of the three CAFE compliance fleets: Domestic passenger car,
                import passenger car, and light truck fleets.
                ---------------------------------------------------------------------------
                 \3018\ In the Figures, the label ``CAFE with Capped AMFA''
                represents the maximum increase each year in the average fuel
                economy set to the limitation ``cap'' for manufacturers attributable
                to dual-fueled automobiles as prescribed in 49 U.S.C. 32906. The
                labels ``A/C'' and ``off-cycle'' represents the increase in the
                average fuel economy adjusted for A/C and off-cycle fuel consumption
                improvement values as prescribed by 40 CFR 600.510-12.
                 \3019\ The Alternative Motor Fuels Act (AMFA) allows
                manufacturers to increase their fleet fuel economy performance
                values by producing dual-fueled vehicles. Incentives are available
                for building advanced technology vehicles such as hybrids and
                electric vehicles, compressed natural gas vehicles and for building
                vehicles able to run on dual-fuels such as E85 and gasoline. For MYs
                1993 through 2014, the maximum possible increase in CAFE performance
                is ``capped'' for a manufacturer attributable to dual-fueled
                vehicles at 1.2 miles per gallon for each model year and thereafter
                decreases by 0.2 miles per gallon each model year through MY 2019.
                49 U.S.C. 32906.
                 \3020\ Consistent with applicable law, NHTSA established
                provisions starting in MY 2017 allowing manufacturers to increase
                fuel economy performance-based on fuel consumption benefits gained
                by technologies not accounted for during normal 2-cycle EPA
                compliance testing (called ``off-cycle technologies'' for
                technologies such as stop-start systems) as well as for A/C systems
                with improved efficiencies and for hybrid or electric full-size
                pickup trucks.
                ---------------------------------------------------------------------------
                BILLING CODE 4910-59-P
                [[Page 25195]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.750
                [[Page 25196]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.751
                 As shown in Figure IX-2, manufacturers' fuel economy performance
                for the total fleet was better than the overall CAFE standard through
                MY 2015. On average, the total fleet exceeded the overall CAFE
                standards by approximately 0.9 mpg for MYs 2011 to 2015. Comparatively,
                as shown in Figure IX-3 through Figure IX-5, for these same model
                years, domestic and import passenger cars exceeded standards on average
                by 2.1 mpg and 2.3 mpg, respectively. By contrast, for light trucks,
                manufacturers on average fell below standards by 0.3 mpg.
                 For MYs 2016 through 2019, as shown in the Figures, NHTSA has
                determined that the combined CAFE performance, including all
                flexibilities and incentives, of the total fleet has or is expected to
                be worse than the applicable CAFE standards, and increasingly so. The
                domestic passenger car fleet is the only compliance category expected
                to continue to be better than CAFE standards through MY 2018. But even
                the overall domestic passenger car fleet
                [[Page 25197]]
                is expected to be worse than standards in MY 2019. The data show MYs
                2016 through 2019 standards involve significant compliance challenges
                for many vehicle manufacturers. This is evident in the fact that the
                total fleet falls below the applicable CAFE standards on average by 0.6
                mpg for these model years. Compliance challenges become even more
                substantial when observing individual compliance fleets. The largest
                individual performance shortfalls (i.e. the difference between CAFE
                performance values and standards) exist for import passenger car
                manufacturers, with an expected shortfall of 2.5 mpg in MY 2019,
                followed by light truck manufacturers, with a shortfall of 1.4 mpg in
                MY 2016.
                 Table IX-5 provides the numerical final CAFE performance values and
                standards for MYs 2004 to 2017. Notably, there was an increase in total
                fleet fuel economy of only 0.1 mpg for MY 2014, and no increase for MY
                2016. In MY 2016, the total fleet's performance fell below the CAFE
                standard by 0.5 mpg. An increase in the total fleet's CAFE performance
                for MY 2017 was largely due to manufacturers gaining benefits from A/C
                and off-cycle technologies. For MY 2017, the total fleet's CAFE
                performance without A/C and off-cycle allowances increased by 0.1 mpg
                compared to MY 2016. However, even combined with new flexibilities, the
                total fleet's CAFE performance, for MY 2017, still falls below the CAFE
                standard by 0.4 mpg.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.752
                BILLING CODE 4910-59-C
                 Figure IX-6 provides a historical overview of the industry's use of
                CAFE compliance flexibilities for addressing performance
                shortfalls.\3021\ MY 2016 is the latest model year for which CAFE
                compliance determinations are complete, and credit application and
                civil penalty payment determinations made by the manufacturer.
                Historically, manufacturers have generally resolved credit shortfalls
                first by carrying forward any earned credits and then applying traded
                credits. In MYs 2014 and 2015, the amount of credit shortfalls is
                almost the same as the amount of carry-forward and traded credits.
                Manufacturers occasionally carryback credits or opt to transfer earned
                credits between their fleets to resolve performance shortfalls. Trading
                credits from another manufacturer and transferring them across fleets
                occurs far more frequently. Also, credit trading has generally taken
                the place of civil penalty payments for resolving performance
                shortfalls. Only a handful of manufacturers have made civil penalty
                payments since the implementation of the credit trading program.\3022\
                NHTSA expects there may be sufficient credits in manufacturers' credit
                accounts to resolve all import passenger car and light truck
                performance shortfalls expected through MY 2019. By statute,
                manufacturers cannot use traded or transferred credits to address
                performance shortfalls for failing to meet the minimum domestic
                passenger car standards.\3023\ One domestic passenger car manufacturer
                paid civil penalties for failing to comply with the minimum domestic
                passenger car standards for MYs 2016 and 2017.\3024\ Additional
                manufacturers are
                [[Page 25198]]
                expected to pay civil penalty payments for failing to comply with the
                minimum domestic passenger cars standards for MYs 2018 through 2019.
                ---------------------------------------------------------------------------
                 \3021\ The Figure includes all credits manufacturers have used
                in credit transactions to date. Credits contained in carryback plans
                yet to be executed or in pending enforcement actions are not
                included in the Figure.
                 \3022\ Six manufacturers have paid CAFE civil penalties since
                credit trading began in 2011. Fiat Chrysler paid the largest civil
                penalty total over the period, followed by Jaguar Land Rover and
                then Volvo. See Summary of CAFE Civil Penalties Collected, CAFE
                Public Information Center, https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Fines_LIVE.html.
                 \3023\ Congress prescribed minimum domestic passenger car
                standards for domestic passenger car manufacturers and unique
                compliance requirements for these standards in 49 U.S.C. 32902(b)(4)
                and 32903(f)(2).
                 \3024\ Fiat Chrysler paid $77,268,702.50 in civil penalties for
                MY 2016 and $79,376,643.50 for MY 2017 for failing to comply with
                the minimum domestic passenger car standards for those MYs.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.753
                 The compliance data show that the rate at which industry has been
                increasing fuel economy, as shown by the actual fuel economy of the
                overall fleet, has not kept pace with the year-over-year increases in
                the stringency of the standards since MY 2010. The margin of CAFE
                overcompliance diminished steadily through MY 2015. In MY 2016, the
                fuel economy of the fleet was worse than standards, and the margin of
                the shortfall has or is projected to become worse through MY 2019.
                Manufacturers have increasingly used CAFE compliance flexibilities and
                paid more in civil penalties to address the growing CAFE shortfalls.
                The data show use of these flexibilities is likely to increase at least
                through 2019.
                3. Shift in Sales Production From Passenger Cars to Light Trucks
                 The notable trend in the stagnant growth in the automotive
                industry's CAFE performance is likely related to an increase in the
                purchase of light trucks beginning with MY 2013. Light trucks had a
                sharp spike in sales, increasing by a total of 5 percent from MYs 2013
                to 2014. In MY 2014, light trucks comprised approximately 41 percent of
                the total sales production volume of automobiles and has continued to
                grow ever since. In comparison, for model year 2014, domestic passenger
                cars represented 36 percent of the total fleet and import passenger
                cars represented 23 percent. Both domestic and import passenger car
                sales have continued to fall every year since MY 2013. Figure IX-7
                shows the sales production volumes of light trucks and domestic and
                import passenger cars for MYs 2004 to 2017. The proportion of light
                trucks in the fleet, being driven by consumer demand and lower fuel
                prices, raises some concern for the ability of that fleet to comply
                with future CAFE standards. Historically, light truck fleets have
                fallen below their associated CAFE standards and have had larger
                performance shortages than either import and domestic passenger car
                fleets. This trend is expected to continue, even with allowance for A/C
                and off-cycle flexibilities. For MY 2019, NHTSA expects even greater
                CAFE performance shortages in the light truck and import passenger car
                fleets than in prior model years, based upon manufacturer's MMY
                reports. The combined effect of these fuel economy shortages will
                require manufacturers to rely heavily on compliance flexibilities or
                pay civil penalties.
                 Another important factor in automobile sales production impacting
                CAFE performance values involves increasing trends in the volume of
                small SUVs and pickup trucks. These vehicles as a percentage of total
                fleet increased from approximately 52 percent in MY 2012 to 63 percent
                in MY 2017. As shown in Figure IX-8, small SUVs, with 4WD and 2WD
                drivetrains, in particular have surpassed the sales production volumes
                of all other vehicle classes over these the given model years. The
                number of small and standard SUVs sold in the U.S. for MY 2017 nearly
                doubled compared to sales in the U.S. for MY 2012. During that same
                period, passenger car sales production as a total of vehicle sales
                production decreased by approximately 11 percent. The combination of
                low gas prices and the increased utility that SUVs provide may explain
                the shift in sales production. Nonetheless, if the sales of these small
                SUVs and pickup trucks continue to increase, NHTSA expects there will
                be continued stagnation in the CAFE performance of the overall fleet.
                BILLING CODE 4910-59-P
                [[Page 25199]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.754
                [[Page 25200]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.755
                BILLING CODE 4910-59-C
                4. Vehicle Classification
                 Before manufactures can comply with CAFE and CO2
                standards, they must first determine how a vehicle is classified in
                accordance with 49 CFR part 523, ``Vehicle Classification.'' In EPCA,
                Congress designated some vehicles as passenger automobiles and some as
                non-passenger automobiles. Vehicle classification, for purposes of the
                light-duty CAFE and CO2 programs, refers to whether a
                vehicle is classified as a passenger automobile (car) or a non-
                passenger automobile (light truck).3025 3026 As discussed
                previously, passenger cars and light trucks are subject to different
                fuel economy and CO2 standards, and light trucks have less
                stringent standards to accommodate their utility usage.
                ---------------------------------------------------------------------------
                 \3025\ See 40 CFR 86.1803-01. For the MYs 2012-2016 standards,
                the MYs 2017-2025 standards, and this rule, EPA uses NHTSA's
                regulatory definitions for determining which vehicles would be
                subject to which CO2 standards.
                 \3026\ EPCA uses the terms ``passenger automobile'' and ``non-
                passenger automobile;'' NHTSA's regulation on vehicle
                classification, 49 CFR part 523, further clarifies the EPCA
                definitions and introduces the term ``light truck'' as a plainer
                language alternative for ``non-passenger automobile.''
                ---------------------------------------------------------------------------
                 Under EPCA and NHTSA's current regulations, vehicles are classified
                as light trucks either on the basis of off-highway capability or on the
                basis of having truck-like (utility)
                characteristics.3027 3028 3029 Determining whether a vehicle
                is capable of ``off-highway operation'' is a two-part determination:
                First, does the vehicle either have 4-wheel drive or a gross vehicle
                weight rating (GVWR) over 6,000 pounds, and second, does the vehicle
                (that has either 4-wheel drive or over 6,000 pounds GVWR) also have ``a
                significant feature . . . designed for off-highway operation.'' \3030\
                NHTSA's current regulations specify that this ``significant feature''
                requires the vehicle to meet at least four out of five ground clearance
                dimensions.\3031\ Further, to be classified as a light truck on the
                basis of having truck-like characteristics instead, NHTSA regulations
                also require the vehicle to perform at least one of the following
                [[Page 25201]]
                functions: Carry more than 10 persons, provide temporary living
                quarters, have an open bed (i.e., a pickup truck), provide more cargo-
                carrying volume than passenger-carrying volume, or permit expanded
                cargo volume capacity by the removal or stowing of rear seats.\3032\
                ---------------------------------------------------------------------------
                 \3027\ 49 U.S.C. 32901(a)(18); 49 CFR part 523.
                 \3028\ 49 CFR 523.5(b).
                 \3029\ 49 CFR 523.5(a).
                 \3030\ 49 U.S.C. 32901(a)(18).
                 \3031\ The ground clearance dimensions are: (i) Approach angle
                of not less than 28 degrees; (ii) breakover angle of not less than
                14 degrees; (iii) departure angle of not less than 20 degrees; (iv)
                running clearance of not less than 20 centimeters; and/or (v) front
                and rear axle clearances of not less than 18 centimeters each.
                 \3032\ By statute, vehicles that NHTSA, on behalf of the
                Secretary of DOT, ``decides by regulation [are] manufactured
                primarily for transporting not more than 10 individuals'' are
                passenger automobiles. 49 U.S.C. 32901(a)(18).
                ---------------------------------------------------------------------------
                 Over time, NHTSA has revised its light truck vehicle classification
                regulations and issued legal interpretations to address changes in
                vehicle designs. Based upon agency observations of current vehicle
                design trends, compliance testing and evaluation, and discussions with
                stakeholders, NHTSA has become aware of certain additional design
                changes that further complicate light truck classification
                determinations for the CAFE and CO2 programs. NHTSA
                discussed several classification issues in the NPRM and sought comments
                on potential resolutions. Only a few comments were received, primarily
                from vehicle manufacturers, and they were aimed generally at requesting
                flexibility in how NHTSA applies the existing classification criteria.
                A summary of the comments received and NHTSA's responses for the final
                rule are explained in the following sections.
                a) Classification Based on ``Truck-Like Characteristics''
                 One of the ``truck-like characteristics'' that allows manufacturers
                to classify vehicles as light trucks is having at least three rows of
                seats as standard equipment, as long as the design also ``permit[s]
                expanded use of the automobile for cargo-carrying purposes or other
                non-passenger-carrying purposes through the removal or stowing of
                foldable or pivoting seats so as to create a flat, leveled cargo
                surface extending from the forwardmost point of installation of those
                seats to the rear of the automobile's interior.'' \3033\ Typically,
                most minivans qualify under the provision by expanding the cargo area
                through removable or stowable seats, and a small percentage of sports
                utility vehicles qualify through folding seats that use the seat backs
                to form a secondary ``raised'' cargo floor.\3034\ NHTSA identified two
                issues with this criterion that various manufacturers appear to be
                approaching differently. Both relate to how expanded cargo area is
                provided when seats are removed or stowed in the vehicle.
                ---------------------------------------------------------------------------
                 \3033\ 49 CFR 523.5(a)(5)(ii).
                 \3034\ All minivans and a small percentage of sports utility
                vehicles that qualify as light trucks do so by meeting the
                characteristic for third row seats. As more advanced seating designs
                are introduced in minivans, manufacturers that wish to retain this
                status will need to avoid losing the expanded cargo characteristics
                that are the basis for the allowing minivans to be qualified as
                light trucks.
                ---------------------------------------------------------------------------
                 The first issue is how to identify the ``forwardmost point of
                installation'' and how the location impacts the available cargo floor
                area and volume behind the seats. Seating configurations have evolved
                considerably over the last twenty years, as minivan seats are now very
                complex in design, providing far more ergonomic functionality. For
                example, the market demand for increased rear seat leg room has
                resulted in adjustable second row seats mounted to sliding tracks.
                Earlier seating designs had fixed attachment points on the vehicle
                floor, and it was easy to identify the ``forwardmost point of
                installation'' because it was readily observable and did not change.
                When seats move forward and backward on sliding tracks, however, the
                ``forwardmost point of installation'' is less readily identifiable. To
                avoid this complication, most manufacturers maintain light truck
                qualification by using adjustable seats that can be removed from the
                vehicle and having a flat floor rearward of the front seats.\3035\ For
                others, the qualification is not as apparent because new adjustable
                seats have been introduced that remain within vehicle to accommodate
                side airbags. Manufacturers designate various positions for the
                forwardmost point of installation in vehicles where the seat in the
                sliding track can be moved far enough forward to allow the entire seat
                to compress against the back of the front seat where it can be stowed
                beyond the forwardmost point of installation, while the seat cushion
                bottom folds towards the seatback. In some cases, manufacturers
                designate the forwardmost point of installation at a location in the
                sliding track where the seat is positioned at its rearmost position in
                the track. In others, the initial point of installation is designated
                at a location in the sliding track accommodating the seating position
                of a 75-percentile male test dummy. The amount of the flat floor
                surface area and cargo volume behind the seats can vary depending on
                which approach a manufacturer adopts.
                ---------------------------------------------------------------------------
                 \3035\ NHTSA notes that to qualify as a light truck, a vehicle
                still requires a flat floor from the forwardmost point of
                installation of removable second row seats to the rear of the
                vehicle.
                ---------------------------------------------------------------------------
                 NHTSA sought public comments in the NPRM to explore potential
                options for establishing the forwardmost point of installation for
                adjustable second row seats and to evaluate whether an additional
                classification criteria could be required, specifying a minimum amount
                of cargo volume behind the seats. Comments were received from the Auto
                Alliance and Fiat Chrysler.\3036\ Both the Auto Alliance and Fiat
                Chrysler commented that some flexibility is needed in determining the
                forwardmost point of installation that allows manufacturers to set the
                location of the seat attachment point to the sliding track in any
                manufacturer-designated position that allows for customer-ergonomics
                and safety, while still meeting the spirit of the expanded cargo-
                carrying requirement.\3037\ The Auto Alliance further commented that
                the forwardmost attachment point of the seat structure to the floor is
                still a viable method of measurement, even when there is a sliding
                track between the floor attachment point and the seat.\3038\
                ---------------------------------------------------------------------------
                 \3036\ The National Automobile Dealers Association commented
                generally that it does not support any substantial modifications to
                the existing passenger car and light truck fleet definitions.
                 \3037\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3038\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 NHTSA did not propose any vehicle reclassifications and is not
                adopting a regulatory change at this time. Based on its review of the
                comments, NHTSA agrees that flexibility is warranted to accommodate
                safety and customer demand but clarifies that the regulation requires
                seats that are not removed to be stowed--that is, moved so as to form a
                cargo area behind the seats. Manufacturers can freely designate the
                seating location in the sliding track to establish the forwardmost
                point of installation. At that seat location, the forwardmost point of
                installation is the forwardmost attachment point of the seat structure
                (including any carriage structures) to the floor in the sliding track.
                Vehicles will be considered to meet the characteristic provided the
                rear of the seats can be moved forward beyond that point and the seats
                articulate to an unusable stowed position either in the floor of the
                vehicle or at the front perimeter of expanded cargo area.\3039\
                ---------------------------------------------------------------------------
                 \3039\ The front perimeter of the cargo area is the plane formed
                behind the front seats and extending from one side of the vehicle to
                the other.
                ---------------------------------------------------------------------------
                 The second issue concerns the ``flatness'' and ``levelness'' of
                folded rear seats that use the seat backs to form a raised cargo
                surface and whether the seats must form a continuous flat, leveled
                surface. Many SUVs have three rows of designated seating positions,
                where the second row has ``captain's seats'' (i.e., two independent
                bucket
                [[Page 25202]]
                seats), rather than the traditional bench-style seating more common
                when the provision was added to NHTSA's regulation. When captain's
                seats are folded down, the seatback can form a flat surface for
                expanded cargo-carrying purposes, but the surface of the seatbacks may
                be angled (i.e., at some angle slightly greater than 0[deg]), or may be
                at a different level with the rest of the cargo area (i.e., horizontal
                surface of folded seats is 0[deg] at a different height from horizontal
                surface of cargo area behind the seats). Captain's seats, when folded
                flat, may also leave significant gaps around and between the seats.
                Some manufacturers have opted to use plastic panels to level the
                surface and to covers the gaps between seats, while others have left
                the space open and the surface angled or at different levels. NHTSA
                sought comments in the NPRM on the following questions related to the
                requirement for a flat, leveled cargo surface:
                 Does the cargo surface need to be flat and level in
                exactly the same plane, or does it fulfill the intent of the criterion
                and provide appropriate cargo-carrying functionality for the cargo
                surface to be other than flat and level in the same plane?
                 Does the cargo surface need to be flat and level across
                the entire surface, or are (potentially large) gaps in that surface
                consistent with the intent of the criterion and providing appropriate
                cargo-carrying functionality? Should panels to fill gaps be required?
                 Certain third row seats are located on top the rear axle
                causing them to sit higher and closer to the vehicle roof. When these
                seats fold flat the available cargo-carrying volume is reduced. Is
                cargo-carrying functionality better ensured by setting a minimum amount
                of useable cargo-carrying volume in a vehicle when seats fold flat?
                 The Auto Alliance, Fiat Chrysler, Hyundai, Kia, and one individual,
                Walter Kreucher, commented on these seating issues. The Auto Alliance,
                Fiat Chrysler, and Walter Kreucher believed that the criteria for a
                ``flat, leveled cargo surface'' should not be interpreted to mean that
                a cargo surface must be flat and level in exactly the same plane.\3040\
                The comments noted that a surface that is not exactly flat and level in
                the same plane can still provide substantial cargo-carrying capacity,
                while allowing manufacturers to provide ergonomically comfortable seats
                that meet safety requirements.\3041\ The comments stated that NHTSA
                should not establish a minimum amount of cargo surface area for seats
                that remain within the vehicle.\3042\ Instead, they preferred that
                manufacturers should be allowed to determine the methodology for
                providing appropriate cargo-carrying functionality without NHTSA
                stipulating additional requirements for flat and level surfaces or gaps
                and gap-filling panels.\3043\
                ---------------------------------------------------------------------------
                 \3040\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943; Walter
                Kreucher, Detailed Comments, NHTSA-2018-0067-0444.
                 \3041\ See, e.g., Fiat Chrysler, Detailed Comments, NHTSA-2018-
                0067-11943.
                 \3042\ See, e.g., Auto Alliance, Detailed Comments, NHTSA-2018-
                0067-12073.
                 \3043\ See, e.g., Fiat Chrysler, Detailed Comments, NHTSA-2018-
                0067-11943.
                ---------------------------------------------------------------------------
                 The Auto Alliance and Fiat Chrysler argued that area or volume
                requirements are not needed, as those attributes speak to overall
                vehicle size and shape, which should remain a consumer choice.\3044\
                The requirements for expanded cargo- or other non-passenger-carrying
                purposes are fully met in the existing regulation, which requires a
                flat, leveled cargo surface with two rows of seats that are folded or
                stowed. Fiat Chrysler also commented that potential new requirements
                would likely be interpreted and executed differently across
                manufacturers and could narrow the choice of engineering solutions and
                negatively affect other important vehicle attributes.\3045\
                ---------------------------------------------------------------------------
                 \3044\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3045\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                ---------------------------------------------------------------------------
                 Hyundai and Kia commented that instead of requiring panels, NHTSA
                could limit the size of the gaps around and between folded seats.\3046\
                In that case, manufacturers would have flexibility to use panels if
                they wish but could take other measures to narrow gaps. On the other
                hand, Walter Kreucher stated that NHTSA should allow gaps of any size
                and not require the use of panels to cover them.\3047\
                ---------------------------------------------------------------------------
                 \3046\ Hyundai, Detailed Comments, EPA-HQ-OAR-2018-0283-4411;
                Kia, Detailed Comments, EPA-HQ-OAR-2018-0283-4195.
                 \3047\ Kreucher, Detailed Comments, NHTSA-2018-0067-0444.
                ---------------------------------------------------------------------------
                 NHTSA is not adopting a regulatory change at this time. NHTSA
                agrees with commenters that it should not require a minimum amount of
                cargo surface area or volume for seats that remain within the vehicle,
                which could be difficult to meet for certain vehicle sizes and shapes
                that would otherwise be considered non-passenger vehicles. NHTSA agrees
                that the amount of cargo volume should be a consumer choice. Setting a
                minimum amount of cargo area or volume could have an adverse effect on
                potential new car buyers.
                 NHTSA notes that there may also be safety considerations involved
                with the requirement to have a flat, leveled cargo surface area formed
                by seat backs. A flat, leveled cargo surface area could prevent objects
                from having a ramp-like surface to gain momentum in rolling backwards
                into the tailgate's interior surface, potentially causing stress or
                damage on the tailgate's latching mechanism. For these reasons, several
                standards exist in the industry for preventing objects from sliding,
                such as standards from the American Disability Act (ADA) that specify
                floor and ground design requirements for protecting wheelchair seated
                occupants. In addition, objects resting on the tailgate could become a
                hazard or source of injury for individuals opening the tailgate. At
                this time, NHTSA accepts the commenters' position that having a cargo
                surface area that is exactly flat and level in the same plane may not
                be necessary. Comments did not provide enough information for NHTSA to
                identify any changes to the existing requirements. Therefore, at this
                time, NHTSA will retain its existing provisions for the stowing of
                foldable or pivoting seats to create a flat, leveled cargo surface, but
                NHTSA may consider conducting research in the future regarding these
                issues. NHTSA has also determined that it should set not a limit on the
                size of the gaps between folded seats at this time, although it may
                consider adopting such limits in the future. NHTSA continues to
                encourage manufacturers to consider the safety implications of all
                aspects of their vehicle designs, including any angling of the seat
                back cargo surface and whether it is appropriate to offer panels as
                optional equipment for covering any large gap openings.
                b) Issues That NHTSA Has Observed Regarding Classification Based on
                ``Off-Road Capability''
                (1) Measuring Vehicle Characteristics for Off-Highway Capability
                 For a vehicle to qualify as off-highway capable, in addition to
                either having 4WD or a GVWR more than 6,000 pounds, the vehicle must
                have four out of five characteristics indicative of off-highway
                operation.\3048\ These characteristics are:
                ---------------------------------------------------------------------------
                 \3048\ 49 CFR 523.5(b)(2).
                ---------------------------------------------------------------------------
                 An approach angle of not less than 28 degrees
                 A breakover angle of not less than 14 degrees
                 A departure angle of not less than 20 degrees
                [[Page 25203]]
                 A running clearance of not less than 20 centimeters
                 Front and rear axle clearances of not less than 18
                centimeters each
                 NHTSA's regulations require manufacturers to measure these
                characteristics when a vehicle is at its curb weight, on a level
                surface, with the front wheels parallel to the automobile's
                longitudinal centerline, and the tires inflated to the manufacturer's
                recommended cold inflation pressure.\3049\ Given that the regulations
                describe the vehicle's physical position and characteristics at time of
                measurement, NHTSA previously assumed that manufacturers would use
                physical measurements of vehicles. In practice, NHTSA has instead
                received from manufacturers a mixture of angles and dimensions from
                design models (i.e., the vehicle as designed, not as actually produced)
                and/or physical vehicle measurements.\3050\ When appropriate, the
                agency will verify reported values by measuring production vehicles in
                the field. NHTSA currently requires that manufacturers use physical
                vehicle measurements as the basis for values reported to the agency for
                purposes of vehicle classification. NHTSA sought comment on whether
                regulatory changes are needed with respect to this issue.
                ---------------------------------------------------------------------------
                 \3049\ Id.
                 \3050\ NHTSA previously encountered a similar issue when
                manufacturers reported CAFE footprint information. In the October
                2012 final rule, NHTSA clarified manufacturers must submit footprint
                measurements based upon production values. 77 FR 63138 (October 15,
                2012).
                ---------------------------------------------------------------------------
                (2) Approach, Breakover, and Departure Angles
                 Approach angle, breakover angle, and departure angle are relevant
                to determining off-highway capability. Large approach and departure
                angles ensure the front and rear bumpers and valance panels have
                sufficient clearance for obstacle avoidance while driving off-road. The
                breakover angle ensures sufficient body clearance from rocks and other
                objects located between the front and rear wheels while traversing
                rough terrain. Both the approach and departure angles are derived from
                a line tangent to the front (or rear) tire static loaded radius arc
                extending from the ground near the center of the tire patch to the
                lowest contact point on the front or rear of the vehicle. The term
                ``static loaded radius arc'' is based upon the definitions in SAE J1100
                and J1544. The term is defined as the distance from wheel axis of
                rotation to the supporting surface (ground) at a given load of the
                vehicle and stated inflation pressure of the tire (manufacturer's
                recommended cold inflation pressure).\3051\
                ---------------------------------------------------------------------------
                 \3051\ 49 CFR 523.2.
                ---------------------------------------------------------------------------
                 The static loaded radius arc is easy to measure, but the imaginary
                line tangent to the static loaded radius arc is difficult to ascertain
                in the field. The approach and departure angles are the angles between
                the line tangent to the static loaded radius arc and the level ground
                on which the test vehicle rests. Simpler measurements that provide good
                approximations for the approach and departure angles involve using
                either a line tangent to the outside diameter or perimeter of the tire
                or a line that originates at the geometric center of the tire contact
                patch and extends to the lowest contact point on the front or rear of
                the vehicle. The first method provides an angle slightly greater than,
                and the second method provides an angle slightly less than, the angle
                derived from the true static loaded radius arc. Both approaches can be
                used to measure angles in the field to verify data submitted by the
                manufacturers used to determine light truck classification decisions.
                 NHTSA sought comment on what the effect would be if it replaced
                reference to the ``static loaded arc radius'' with a different term
                like ``outside perimeter of the tire'' or ``geometric center of the
                tire contact patch.'' The Auto Alliance and Fiat Chrysler offered
                comments. The Auto Alliance and Fiat Chrysler commented that only a
                measurement using the static loaded arc radius reasonably reflects the
                tire condition during off-road events that approach, breakover, and
                departure angles are quantifying. They also stated the static loaded
                arc radius best reflects the actual condition that exists versus the
                outside tire diameter.\3052\ Finally, the Auto Alliance commented the
                static loaded arc radius is easy to measure; therefore, the off-road
                criteria should remain tied to the static loaded arc radius.\3053\
                ---------------------------------------------------------------------------
                 \3052\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3053\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 After reviewing the comments, NHTSA agrees that the static loaded
                arc radius is the most accurate way to account for the condition of the
                tire and the vehicle-to-ground interaction during off-road events.
                NHTSA has decided to accept the Auto Alliance's and Fiat Chrysler's
                views and will retain the existing definitions for off-road angles
                based upon the static loaded arc radius.
                (3) Running Clearance
                 NHTSA regulations define ``running clearance'' as ``the distance
                from the surface on which an automobile is standing to the lowest point
                on the automobile, excluding unsprung weight.'' \3054\ Unsprung weight
                includes the components (e.g., suspension, wheels, axles, and other
                components directly connected to the wheels and axles) that are
                connected and translate with the wheels. Sprung weight, on the other
                hand, includes all components fixed underneath the vehicle and
                translate with the vehicle body (e.g., mufflers and subframes). To
                clarify these requirements, NHTSA previously issued a letter of
                interpretation stating that certain parts of a vehicle--such as tire
                aero deflectors that are made of flexible plastic, bend without
                breaking, and return to their original position--would not count
                against the 20-centimeter running clearance requirement.\3055\ The
                agency explained that this does not mean a vehicle with less than 20-
                centimeters running clearance could be elevated by an upward force that
                bends the deflectors and still be considered compliant with the running
                clearance criterion, as it would be inconsistent with the conditions
                listed in the introductory paragraph of 49 CFR 523.5(b)(2). Further,
                NHTSA explained that without a flexible component installed, the
                vehicle must meet the 20-centimeter running clearance along its entire
                underside. This 20-centimeter clearance is required for all sprung
                weight components.
                ---------------------------------------------------------------------------
                 \3054\ Id.
                 \3055\ See letter to Mark D. Edie, Ford Motor Company, July 30,
                2012, available at https://isearch.nhtsa.gov/files/11-000612%20M.Edie%20(Part%20523).htm.
                ---------------------------------------------------------------------------
                 The agency is aware of vehicle designs that incorporate rigid
                (i.e., inflexible) air dams, valance panels, exhaust pipes, and other
                components, equipped as manufacturers' standard or optional equipment
                (e.g., running boards and towing hitches), that likely do not meet the
                20-centimeter running clearance requirement. Despite these rigid
                features, it appears manufacturers are not taking these components into
                consideration when making measurements. Additionally, NHTSA believes
                some manufacturers may provide dimensions for their base vehicles
                without considering optional or various trim level components that may
                reduce the vehicle's ground clearance. Consistent with our approach to
                other measurements, NHTSA believes that ground clearance, as well as
                all the other off-highway criteria for a light truck determination,
                should use the measurements from vehicles with all standard and
                optional equipment
                [[Page 25204]]
                installed, at the time of the first retail sale.\3056\ The agency
                reiterates that the characteristics listed in 49 CFR 523.5(b)(2) are
                characteristics indicative of off-highway capability. A fixed feature--
                such as an air dam that does not flex and return to its original state
                or an exhaust that could detach--inherently interferes with the off-
                highway capability of these vehicles. If manufacturers seek to classify
                these vehicles as light trucks under 49 CFR 523.5(b)(2) and the
                vehicles do not meet the four remaining characteristics to demonstrate
                off-highway capability, they must be classified as passenger cars.
                ---------------------------------------------------------------------------
                 \3056\ See NHTSA's footprint test procedure for verifying CAFE
                standards uses vehicles equipped at time of first retail sale. See
                TP-537-01 located at https://www.nhtsa.gov/vehicle-manufacturers/test-procedures.
                ---------------------------------------------------------------------------
                 In the NPRM, NHTSA sought public comments on how to consider
                components such as air dams, exhaust pipes, and other hanging component
                features--especially those that are inflexible--as relates to running
                clearance and whether the agency should consider amending its
                definition in Part 523 to account for these components. The Auto
                Alliance and three automobile manufacturers--Fiat Chrysler, Hyundai,
                and Kia--commented on the questions. The Auto Alliance and Fiat
                Chrysler commented that no change is needed for the 20-centimeter
                running clearance requirement for fixed features of the vehicle; all
                fixed components must have 20-centimeter of running clearance.\3057\
                They agreed that flexible components that bend without breaking and
                return to their original position do not count against the 20-
                centimeter running clearance requirement.\3058\ They disagreed with
                NHTSA's position that these requirements should apply to all vehicles
                with standard and optional equipment installed at the time of the first
                retail sale and proposed instead that the requirement should be ``as
                shipped to the dealer.'' \3059\ Additionally, the Auto Alliance asked
                NHTSA to make a specific allowance for vehicles that have adjustable
                ride height, such as air suspension, and permit the running clearance
                and other off-road clearance measurements to be made in the lifted or
                off- road mode.\3060\ Hyundai and Kia urged NHTSA not to modify the
                definition of ``running clearance,'' which currently is defined as
                ``the distance from the surface on which an automobile is standing to
                the lowest point on the automobile, excluding unsprung weight.'' \3061\
                ---------------------------------------------------------------------------
                 \3057\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3058\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3059\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3060\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3061\ Hyundai, Detailed Comments, EPA-HQ-OAR-2018-0283-4411;
                Kia, Detailed Comments, EPA-HQ-OAR-2018-0283-4195.
                ---------------------------------------------------------------------------
                 Based upon the comments above, NHTSA has decided to retain its
                running clearance requirements for qualifying light trucks without
                change. First, running clearance means the distance from the surface on
                which an automobile is standing to all fixed components under the
                vehicle, excluding unsprung components, axle clearance components and
                flexible components that bend without breaking and returning to their
                original position as explained in NHTSA's previous interpretation.
                Second, NHTSA acknowledges that at this time, during validation testing
                for running clearance, a vehicle with optional equipment installed will
                only be tested ``as shipped to the dealer.'' NHTSA has found that
                optional equipment can impact a vehicle's ability to comply with
                running clearance requirements, while optional equipment must be
                considered for other light truck agency validation tests unless the
                equipment has no impact on the outcome of the test.
                (4) Front and Rear Axle Clearance
                 NHTSA regulations state that front and rear axle clearances of not
                less than 18 centimeters are another criterion that can be used for
                designating a vehicle as off-highway capable.\3062\ The agency defines
                ``axle clearance'' as the vertical distance from the level surface on
                which an automobile is standing to the lowest point on the axle
                differential of the automobile.\3063\
                ---------------------------------------------------------------------------
                 \3062\ 49 CFR 523.5(b)(2)(v).
                 \3063\ 49 CFR 523.2.
                ---------------------------------------------------------------------------
                 The agency believes this definition may be outdated because of
                vehicle design changes, including axle system components and
                independent front and rear suspension components. In the past,
                traditional light trucks with and without 4WD systems had solid rear
                axles with center- mounted differentials on the axle. For these trucks,
                the rear axle differential was closer to the ground than any other axle
                or suspension system component. This traditional axle design still
                exists today for some trucks with a solid chassis (also known as body-
                on-frame configuration). Today, however, many SUVs and CUVs that
                qualify as light trucks are constructed with a unibody frame and have
                unsprung (e.g., control arms, tie rods, ball joints, struts, shocks,
                etc.) and sprung components (e.g., the axle subframes) connected
                together as a part of the axle assembly.\3064\ These unsprung and
                sprung components are located under the axles, making them lower to the
                ground than the axles and the differential, and were not contemplated
                when NHTSA established the definition and the allowable clearance for
                axles. The definition also did not originally account for 2WD vehicles
                with GVWRs greater than 6,000 pounds that had one axle without a
                differential, such as the model year 2018 Ford Expedition. Vehicles
                with axle components that are low enough to interfere with the
                vehicle's ability to perform off-road would seem inconsistent with the
                regulation's intent of ensuring off-highway capability, as Congress
                required.\3065\
                ---------------------------------------------------------------------------
                 \3064\ Unibody frames integrate the frame and body components
                into a combined structure.
                 \3065\ 49 U.S.C. 32901(a)(18)(A).
                ---------------------------------------------------------------------------
                 In light of these issues, comments were sought in the NPRM on
                whether (and if so, how) to revise the definition of axle clearance.
                NHTSA sought comments on what unsprung axle components should be
                considered when determining a vehicle's axle clearance. The agency
                questioned whether the definition for axle clearance should be modified
                to account for axles without differentials. NHTSA also sought comment
                on whether the axle subframes surrounding the axle components but
                affixed directly to the vehicle unibody as sprung mass (lower to the
                ground than the axles) should be considered in the allowable running
                clearance discussed above. Finally, NHTSA sought comments on whether it
                should consider replacing both the running and axle clearance criteria
                with a single ground clearance criterion that considers all components
                underneath the vehicle that impact a vehicle's off-road capability.
                 Comments were received from the Auto Alliance, Fiat Chrysler,
                Hyundai, and Kia. All the manufacturers that commented claimed no
                change is needed to the current definition, regardless of whether the
                axle components are sprung or unsprung masses, as the bottom of the
                differential is the vulnerable component.\3066\ The Auto Alliance also
                stated there is no
                [[Page 25205]]
                need to further modify the definition to account for axles without
                differentials. Further, the Auto Alliance does not think a single
                criterion that considers all components under the axle is needed and
                prefers to keep the existing regulation.\3067\ Fiat Chrysler and the
                Auto Alliance also recommended that 2WD SUVs and CUVs be reclassified
                back into the truck fleet, where they had been placed prior to the 2011
                MY. Their position is that 2WD SUVs are designed to meet the ``off-
                road-capable'' definition in NHTSA's rules by having the required
                running and/or axle clearances as well as meeting other off-road
                dimensional criteria.\3068\ Hyundai stated that changing the point of
                measurement now would have significant development and economic
                impacts.\3069\ Kia stated that it has designed its vehicles and
                developed product plans in reliance on the current definitions, and
                those designs and product plans cannot be modified cheaply or
                quickly.\3070\
                ---------------------------------------------------------------------------
                 \3066\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943; Hyundai,
                Detailed Comments, EPA-HQ-OAR-2018-0283-4411; Kia, Detailed
                Comments, EPA-HQ-OAR-2018-0283-4195.
                 \3067\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3068\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943;
                Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3069\ Hyundai, Detailed Comments, EPA-HQ-OAR-2018-0283-4411.
                 \3070\ Kia, Detailed Comments, EPA-HQ-OAR-2018-0283-4195.
                ---------------------------------------------------------------------------
                 NHTSA already addressed the comments on 2WD SUVs in a previous
                rulemaking, and NHTSA has no additional response at this time.\3071\
                Upon review of other comments, manufacturers did not clearly
                distinguish which parts of the axle sub-frames should be considered as
                sprung masses in order for NHTSA to understand if modifications are
                needed to its axle clearance requirements. Therefore, at this time,
                NHTSA is retaining its axle clearance requirements as currently
                specified. However, NHTSA still believes it is beneficial to continue
                efforts at defining those axle components that are sprung or unsprung
                masses before considering any changes to its regulatory provisions. In
                addition, NHTSA needs to understand any significant developmental and
                economic impacts that might be associated with any possible changes to
                its requirements. Therefore, NHTSA will consider collecting further
                information on these issues and may take further action related to this
                issue in the future.
                ---------------------------------------------------------------------------
                 \3071\ No new arguments have been raised beyond those already
                considered in the April 6, 2006, final rule (see 71 FR 17566).
                ---------------------------------------------------------------------------
                B. EPA Compliance and Enforcement
                1. Overview of the EPA Compliance Process
                 EPA established comprehensive vehicle certification, compliance,
                and enforcement provisions for the GHG standards as part of the
                rulemaking establishing the initial GHG standards for MY 2012-2016
                vehicles.\3072\ Manufacturers have been following these provisions
                since MY 2012 and EPA did not propose or seek comments on changing its
                compliance and enforcement program.
                ---------------------------------------------------------------------------
                 \3072\ See 75 FR 25468-25488 and 77 FR 62884-62887 for a
                description of these provisions. See also ``The 2018 EPA Automotive
                Trends Report, Greenhouse Gas Emissions, Fuel Economy, and
                Technology since 1975,'' EPA-420-R-19-002 March 2019 for additional
                information regarding EPA compliance determinations.
                ---------------------------------------------------------------------------
                a) What Compliance Flexibilities and Incentives are Currently Available
                Under the CO2 Program and How Do Manufacturers Use Them?
                 Under EPA's regulations, manufacturers can use credit flexibilities
                to comply with CO2 standards for passenger car or light
                truck compliance fleets. Similar to the CAFE program, manufacturers
                gain credits when the performance of a fleet exceeds its required
                CO2 fleet average standard which can be carried forward for
                five years. EPA also allows a one-time credit carry-forward exceeding 5
                years, allowing MY 2010-2015 to be carried forward through MY2021. A
                manufacturer's fleet performance that does not meet the fleet average
                standard generates a credit deficit. Manufacturers can carry credit
                deficits forward up to three model years before having to resolve the
                shortfall.
                 NHTSA's program continues the 5-year carry-forward and 3-year
                carryback, as required by statute. Credit ``transfer'' means the
                ability of manufacturers to move credits from their passenger car fleet
                to their light truck fleet, or vice versa. As part of the EISA
                amendments to EPCA, NHTSA was required to establish by regulation a
                CAFE credit transferring program, now codified at 49 CFR part 536, to
                allow a manufacturer to transfer credits between its car and truck
                fleets to achieve compliance with the standards. For example, credits
                earned by over-compliance with a manufacturer's car fleet average
                standard could be used to offset debits incurred because the
                manufacturer did not meet the truck fleet average standard in a given
                year.
                 Under Section 202(a) of the CAA, there is no statutory limitation
                on car/truck credit transfers, and EPA's CO2 program allows
                unlimited credit transfers across a manufacturer's car and light truck
                fleets to meet CO2 standards.
                 EPA requested comment on a variety of ``enhanced flexibilities''
                whereby EPA could make adjustments to current incentives and credit
                provisions and potentially add new flexibility opportunities to expand
                the means by which manufacturers may satisfy standards. Some of these
                additional flexibilities would not result in a reduction in program
                stringency, while others would incentivize technologies that could
                realize greater CO2 emissions reductions over a longer term,
                but would result in a loss of emission benefits in the short-term, as
                discussed below. EPA requested comments on these topics to support the
                increased application of technologies that the automotive industry is
                developing and deploying that could potentially lead to further long-
                term emissions reductions and allow manufacturers to comply with
                standards while reducing costs.
                 EPA explained that one category of flexibilities, such as off-cycle
                credits and credit banking, involve credits that are based on real
                world emissions reductions and do not represent a loss of overall
                emissions benefits or a reduction in program stringency, yet offer
                manufacturers potentially lower-cost or more efficient path to
                compliance. Another category of flexibilities, such as incentives for
                battery electric vehicles, hybrid technologies, and alternative fuels,
                do result in a loss of emissions benefit and represent a reduction in
                the effective stringency of the standards to the extent the incentives
                are used by manufacturers. These incentives would help manufacturers
                meet a numerically more stringent standard, but would not reduce real-
                world CO2 emissions in the short term compared to a lower
                stringency option with fewer such incentives. EPA's policy rationale
                for providing such incentives, as articulated in the 2012 rulemaking,
                was that such programs could incentivize the development and deployment
                of advanced technologies with the potential to lead to greater
                CO2 emissions reductions in the longer-term, where such
                technologies today are limited by higher costs, market barriers,
                infrastructure, and consumer awareness.\3073\ Such incentive approaches
                would also result in rewarding automakers who invest in certain
                technological pathways, rather than being technology neutral.
                ---------------------------------------------------------------------------
                 \3073\ See 77 FR 62810-62826 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 Prior to the proposal, automakers and other stakeholders expressed
                support for
                [[Page 25206]]
                this type of compliance flexibility. For example, in March 2018, Ford
                stated, ``We support increasing clean car standards through 2025 and
                are not asking for a rollback. We want one set of standards nationally,
                along with additional flexibility to help us provide more affordable
                options for our customers.'' \3074\ Honda, in April 2018, also
                expressed its support for an approach that retained the existing
                standards while extending the advanced technology multipliers for
                electrified vehicles, eliminated automakers' responsibility for the
                impact of upstream emissions from the electric grid, and accommodated
                more off-cycle technologies.\3075\
                ---------------------------------------------------------------------------
                 \3074\ ``A Measure of Progress'' Bill Ford, Executive Chairman,
                Ford Motor Company, and Jim Hackett, President and CEO, Ford Motor
                Company, March 27, 2018, https://medium.com/cityoftomorrow/a-measure-of-progress-bc34ad2b0ed.
                 \3075\ Honda Release ``Our Perspective--Vehicle Greenhouse Gas
                and Fuel Economy Standards,'' April 20, 2018, http://news.honda.com/newsandviews/pov.aspx?id=10275-en.
                ---------------------------------------------------------------------------
                 EPA's request for comments was largely based on its consideration
                of input from automakers and other stakeholders, including suppliers
                and alternative fuels industries, supporting a variety of program
                flexibilities.\3076\ The following provides an overview of EPA's
                request for comments on several flexibility concepts, the comments EPA
                received, and the agency's response to those comments. After
                considering comments, EPA is not adopting new incentives in the areas
                of credit multipliers (with the exception of multipliers for natural
                gas vehicles), new incentives for hybrid vehicles, incentives for
                autonomous or connected vehicles, or alternative fueled vehicles other
                than natural gas, as part of this final rule. EPA is finalizing program
                changes for the treatment of upstream emissions for electric vehicles,
                the treatment of natural gas vehicles, the treatment of hybrid and
                target-beating full-size pickup trucks, and off-cycle credits, as
                discussed below.
                ---------------------------------------------------------------------------
                 \3076\ Memorandum to docket EPA-HQ-OAR-2018-0283 regarding
                meetings with the Alliance of Automobile Manufacturers on April 16,
                2018 and Global Automakers on April 17, 2018. EPA-HQ-OAR-2018-0283-
                0022.
                ---------------------------------------------------------------------------
                (1) Credit Flexibilities
                 Under the EPA program, CO2 credits may be carried
                forward, or banked, for a period of five years, with the exception that
                MY 2010-2015 credits may be carried forward and used through MY 2021.
                CO2 credits may also be traded between manufacturers and
                transferred between passenger car and light truck fleets similar to the
                CAFE program, but without any adjustment for fuel savings. Under
                Section 202(a) of the CAA, there is no statutory limitation on credit
                transfers between a manufacturer's passenger car and light truck
                fleets, and EPA's CO2 program allows unlimited credit
                transfers across a manufacturer's passenger car and light truck fleets
                to comply with CO2 standards. This flexibility is based on
                the expectation that it will help facilitate manufacturer compliance
                with CO2 standards in the lead time provided, and allow
                CO2 emissions reductions to be achieved in the most cost
                effective way.
                 Automakers suggested, prior to the NPRM proposal, a variety of ways
                in which CO2 credit life could be extended under the CAA,
                like allowing automakers to carry-forward MY 2010 and later banked
                credits to MY 2025, extending the life of credits beyond five years, or
                even unlimited credit life where credits would not expire. EPA
                requested comments in the NPRM on extending credit carry-forward under
                the CO2 program beyond the current five years, including
                unlimited credit life.
                 General comments were received in response to the NPRM from the
                National Automobile Dealers Association and Volkswagen. They commented
                that credit carry-forward and carryback options help with annual
                compliance with the CO2 program.\3077\ They stated that
                these mechanisms allow manufacturers to become compliant over the
                course of the time a credit is usable in the market.\3078\ Toyota,
                General Motors, Fiat Chrysler, the Auto Alliance, and the Global
                Automakers each commented that CO2 credits earned by
                manufacturers need a longer life so they may be carried forward further
                than the current five-year limitation.\3079\ They asked for an
                unlimited period for using CO2 credits without restrictions,
                since they argue that automakers have earned those credits and should
                be allowed to use them however they see fit.\3080\ They also stated
                that this would incentivize manufacturers to make early reductions in
                CO2 emissions.\3081\ Furthermore, it was noted that credits
                are earned when manufacturers achieve lower CO2 fleet
                average emissions than otherwise required by regulation in any given
                model year. They stated that this typically results from actions taken
                by a manufacturer to deploy specific models or more efficient
                technology than required, often at a higher cost. Such technologies
                reduce the amount of CO2 emissions released into the
                atmosphere over the life of the vehicle, which could be over several
                decades. Therefore, the resulting credit earned by a manufacturer for
                having made the product or technology investment that resulted in the
                reduced emissions should not be limited to five years.
                ---------------------------------------------------------------------------
                 \3077\ National Automobile Dealers Association, Detailed
                Comments, NHTSA-2018-0067-12064; Volkswagen, Detailed Comments,
                NHTSA-2017-0069-0583.
                 \3078\ See, e.g., National Automobile Dealers Association,
                NHTSA-2018-0067-12064.
                 \3079\ Toyota, Detailed Comments, NHTSA-2018-0067-12150; General
                Motors, Detailed Comments, NHTSA-2018-0067-11858; Fiat Chrysler,
                Detailed Comments, NHTSA-2018-0067-11943; Auto Alliance, Detailed
                Comments, NHTSA-2018-0067-12073; Global Automakers, Detailed
                Comments, NHTSA-2018-0067-12032.
                 \3080\ See, e.g., Global Automakers, Detailed Comments, NHTSA-
                2018-0067-12032.
                 \3081\ See, e.g., General Motors, Detailed Comments, NHTSA-2018-
                0067-11858.
                ---------------------------------------------------------------------------
                 Global Automakers, the Auto Alliance, Fiat Chrysler, and Toyota
                requested a one-time expiration date extension through 2026 for
                CO2 credits earned in MYs 2010-2015.\3082\ They asserted
                that earned credits represent actual CO2 reductions and
                increasing their lifespan will allow for better compliance. Conversely,
                Honda disagreed with the extension of MY 2010-2015 credits through 2026
                because they have been selling their credits under the assumption that
                they would expire.\3083\ Honda stated that shorter life (soon to
                expire) credits are worth less than longer life credits, leading to a
                disadvantage for manufacturers who have already sold these credits at a
                lower price. Honda asserted that the one-time extension would benefit
                only a few automakers.\3084\ However, Honda did agree that a one-time
                extension through 2026 for MYs 2016-2020 CO2 credits would
                assist with compliance because these credits have yet to be involved in
                trades.\3085\
                ---------------------------------------------------------------------------
                 \3082\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
                12032; Alliance, Detailed Comments, NHTSA-2018-0067-12073; Fiat
                Chrysler, Detailed Comments, NHTSA-2018-0067-11943; Toyota Detailed
                Comments, NHTSA-2018-0067-12150.
                 \3083\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                 \3084\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                 \3085\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                ---------------------------------------------------------------------------
                 In sum, commenters requested either unlimited allowances to carry-
                forward surplus credits without any expiration date, a one-time
                expiration date extension through 2026 for CO2 credits
                earned from MY 2010 and later, or consideration for extending credit
                life longer than the current five-year provision. After considering the
                comments received, EPA has decided not to change its credit carry-
                forward provisions at this time, and will retain the credit carry-
                forward period under the CO2 program at five years for
                credits
                [[Page 25207]]
                generated in MYs 2016 and later. EPA does not believe any changes to
                its credit carry-forward provisions are warranted. EPA notes that
                NHTSA's CAFE program is constrained by statute to a five-year carry-
                forward so if EPA adopted a longer carry-forward period, it might be of
                limited use since the level of stringency of the CO2 and
                CAFE standards is similar across the programs. Also, the analysis on
                which the tailpipe CO2 emissions standards finalized today
                are based, assumed a five-year carry-forward period for credits.
                 Another reason for denying manufacturers' requests is the potential
                inequitable advantage a longer credit life could have for manufacturers
                with surplus credits, especially those with significant amounts of
                credits currently banked for multiple model years. Manufacturers
                without credits, or manufacturers who have already sold their credits
                at current market values based on the present five-year carry-forward
                credit lifespan, as Honda discussed, will be significantly
                disadvantaged.\3086\ These manufacturers are unlikely to be able to
                renegotiate the price of credit trades already made. Manufacturers with
                large amounts of credits would clearly be advantaged and able to
                distort the market in ways unfavorable to the goal of reducing
                emissions. EPA is concerned that these manufacturers will be able to
                create uncertainties in the market by being able to infuse large
                volumes of credits into future model years where it may even be
                possible to delay some cost-effective technologies from entering
                production because manufacturers are relying upon these credits as an
                alternative pathway to compliance.
                ---------------------------------------------------------------------------
                 \3086\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                ---------------------------------------------------------------------------
                (2) Advanced Technology Incentives
                 The existing EPA CO2 program provides incentives for
                electric vehicles, fuel-cell vehicles, plug-in hybrid vehicles, and
                natural gas vehicles. The 2012 rulemaking allowed manufacturers to use
                a 0 grams/mile emissions factor for all electric powered vehicles
                rather than having to account for the CO2 emissions
                associated with upstream electricity generation, up to a per-
                manufacturer cumulative production cap for MYs 2022-2025. The program
                also includes multiplier incentives that allow manufacturers to count
                advanced technology vehicles as more than one vehicle in the compliance
                calculations. The multipliers began with MY 2017 and end after MY
                2021.\3087\ Prior to the proposal, stakeholders suggested that these
                incentives should be expanded to support further the production of
                advanced technologies by allowing manufacturers to continue to use the
                0 grams/mile emissions factor for electric powered vehicles rather than
                having to account for upstream electricity generation emissions and by
                extending and potentially increasing the multiplier incentives.
                ---------------------------------------------------------------------------
                 \3087\ The multipliers are for EV/FCVs: 2017-2019--2.0, 2020--
                1.75, 2021--1.5; for PHEVs and dedicated and dual-fuel CNG vehicles:
                2017-2019--1.6, 2020--1.45, 2021--1.3.
                ---------------------------------------------------------------------------
                 First, EPA requested comments on extending the use of 0 grams/mile
                emissions factor for electric powered vehicles.
                 The Auto Alliance, Global Automakers, and several manufacturers
                commented that upstream utility emissions come from power plants, not
                vehicle tailpipes, and manufacturers have no control over the feedstock
                used by those power plants and should not be held responsible for their
                upstream electricity emissions.\3088\ The Auto Alliance further
                commented that removing upstream accounting is not an incentive for
                advanced technology vehicles; rather, it should be seen as a correction
                to remove responsibility for emissions over which the automakers have
                no control.\3089\ Fiat Chrysler commented that ``requiring upstream
                accounting could impede development of BEVs or PHEVs, as accounting of
                upstream emissions degrades the CO2 performance of BEVs to
                the level of PHEVs, and PHEVs to the level of a conventional hybrid
                electric vehicle. This, in effect, disincentivizes the technology.''
                \3090\
                ---------------------------------------------------------------------------
                 \3088\ See, e.g., Volkswagen, Detailed Comments, NHTSA-2017-
                0069-0583.
                 \3089\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3090\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                ---------------------------------------------------------------------------
                 Several other commenters also supported not counting upstream
                emissions and instead only counting electric powered vehicle tailpipe
                emissions of 0 grams/mile.\3091\ These commenters included NCAT, SAFE,
                BorgWarner, CALSTART, Eaton, and Edison Electric Institute.
                ---------------------------------------------------------------------------
                 \3091\ See, e.g., NCAT, NHTSA-2018-0067-11969.
                ---------------------------------------------------------------------------
                 API did not support continuing the 0 grams/mile emission factor for
                electricity use, commenting that by failing to factor the real
                contribution of upstream CO2 emissions from electric
                generation, the regulatory agencies would distort the market for
                developing transportation fuel alternatives.\3092\ API commented that
                EPA should not ignore the environmental burden of upstream emissions in
                granting production incentives to automakers.
                ---------------------------------------------------------------------------
                 \3092\ API, Detailed Comments, EPA-HQ-OAR-2018-0283-5458.
                ---------------------------------------------------------------------------
                 Manufacturers of Emission Controls Association (MECA) commented
                that ``with the growing emphasis on real-world emission reductions, it
                becomes increasingly important to consider all emissions to the
                environment, including upstream emissions. Numerous studies have shown
                that in many parts of the country, the temporary 0 grams/mile upstream
                emissions factor is not delivered in the real-world . . . MECA believes
                that EPA should continue to set performance-based standards that assess
                technology pathways based on delivering the intended emission
                reductions over the full well-to-wheels vehicle life cycle in the real-
                world.'' \3093\ Motor & Equipment Manufacturers Association (MEMA) also
                supported a well-to-wheel fuel lifecycle approach, commenting that
                without this type of comprehensive assessment on the fuel impacts and
                comprehensive CO2 costs, policies improperly ``slant toward
                preferred technologies.'' \3094\ Nonetheless, MEMA commented that it is
                not opposed to continuing to allow 0 grams/mile emissions factor for
                electric powered vehicles through 2026.
                ---------------------------------------------------------------------------
                 \3093\ MECA, Detailed Comments, NHTSA-2018-0067-11994.
                 \3094\ MEMA, EPA-HQ-OAR-2018-0283-5692. See https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf.
                ---------------------------------------------------------------------------
                 The Union of Concerned Scientists (UCS) commented that not
                accounting for upstream emissions combined with the multipliers has a
                significant impact on the efficacy of the standard, and extending these
                regulatory incentives is more likely to result in a credit giveaway
                than to drive additional deployment of electric vehicles.\3095\ UCS
                further commented that, to date, more than half of the electric
                vehicles sold have been in California and the states that have adopted
                California's ZEV standards; however, UCS asserted, federal standards
                ignore the upstream emissions for all vehicles sold.
                ---------------------------------------------------------------------------
                 \3095\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
                ---------------------------------------------------------------------------
                 After carefully considering the wide range of comments on whether
                to include upstream emissions associated with electricity use in the
                compliance calculations for electrified vehicles, EPA has decided to
                allow the continued use of the 0 grams/mile emissions factor with no
                per-manufacturer production caps or other limitations. EPA is revising
                its regulations to remove the production caps and related provisions.
                When EPA initially adopted a production cap for manufacturers that
                [[Page 25208]]
                use the 0 grams/mile emissions factor, in the rulemaking to establish
                CO2 standards for MY 2012-2016 vehicles, there were no
                controls in place for CO2 emissions from electricity
                production.\3096\ This was also the case when EPA extended the 0 grams/
                mile upstream provision and revised the production caps in the rule
                establishing MY 2017-2025 standards.\3097\ However, since then, EPA has
                adopted a program to control CO2 emissions from power
                plants.\3098\ Emissions from the power sector have been declining and
                that trend is projected to continue.\3099\ For these reasons, EPA no
                longer views the upstream emissions factor as an incentive in the same
                way it views a multiplier incentive which provides bonus credits. EPA
                agrees that, at this time, manufacturers should not account for
                upstream utility emissions. Therefore, EPA is adopting regulatory
                changes consistent with its historical practice of basing compliance
                with vehicle emissions standards on tailpipe emissions through model
                year 2026. EPA may choose to reconsider this decision in a future
                CO2 rulemaking, and will reexamine the issue when
                establishing standards commencing with the 2027 model year.\3100\
                ---------------------------------------------------------------------------
                 \3096\ 75 FR 25341, May 7, 2010.
                 \3097\ 77 FR 62816, October 15, 2012.
                 \3098\ 84 FR 32520, July 8, 2019.
                 \3099\ 84 FR 32561.
                 \3100\ By comparison, the CAFE program uses an energy efficiency
                metric instead of an emissions metric, and standards that are
                expressed in miles per gallon. For PHEVs and BEVs, to determine
                gasoline the equivalent fuel economy for operation on electricity, a
                Petroleum Equivalency Factor (PEF) is applied to the measured
                electrical consumption. The PEF for electricity was established by
                the Department of Energy, as required by statute, and includes an
                accounting for upstream energy associated with the production and
                distribution for electricity relative to gasoline. Therefore, the
                CAFE program includes upstream accounting based on the metric that
                is consistent with the fuel economy metric. The PEF for electricity
                also includes an incentive that effectively counts only 15 percent
                of the electrical energy consumed.
                ---------------------------------------------------------------------------
                 Second, EPA requested comments on extending or increasing advanced
                technology incentives, including multiplier incentives, with
                multipliers in the range of 2.0-4.5. EPA received a wide range of
                comments both for and against increasing the multiplier incentives. The
                MY 2017-2025 CO2 program finalized in 2012 included
                incentive multipliers for certain advanced technologies for MY 2017-
                2021 vehicles.
                 The Auto Alliance, Global Automakers, and several individual
                manufacturers commented in support of continued and increased
                multipliers. The Auto Alliance commented that EPA should extend and
                significantly expand multipliers ``to encourage a transition to these
                technologies while cost, range, and infrastructure challenges are
                addressed to encourage ongoing investments in advanced technologies.''
                \3101\ Global Automakers commented that multipliers should be included
                through MY 2026, set at values that encourage ongoing investment in
                advanced technologies, without diluting overall efficiency improvements
                in the program.\3102\ NCAT, Eaton, Plug-in America, Alliance to Save
                Energy, SAFE, and MEMA also supported additional multiplier incentives
                to encourage further the production and sale of advanced technology
                vehicles.\3103\
                ---------------------------------------------------------------------------
                 \3101\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3102\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
                12032.
                 \3103\ NCAT, Detailed Comments, NHTSA-2018-0067-11969; Eaton,
                Detailed Comments, EPA-HQ-OAR-2018-0283-5068; Plug-In America,
                Detailed Comments, NHTSA-2018-0067-12028; Alliance to Save Energy,
                Detailed Comments, NHTSA-2018-0067-11837; SAFE, Detailed Comments,
                NHTSA-2018-0067-11981; see https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf.
                ---------------------------------------------------------------------------
                 EPA also received comments against extending the multiplier
                credits. UCS commented that reducing the stringency of the standards
                lessens the need for the adoption of these vehicles and undermines the
                initial rationale for these credits, resulting in a significant bank of
                credits which would further erode the benefits of these
                standards.\3104\ American Council for an Energy-Efficient Economy
                (ACEEE) commented that providing multiplier incentives for any longer
                period, or at a greater rate than those currently in place, would
                create windfall credits for manufacturers given the industry's current
                product plans.\3105\ Fiat Chrysler commented generally in support of a
                multiplier incentive, but noted that since multipliers are a
                CO2--only flexibility not present in the CAFE program,
                greater use of multipliers would result in further disharmonizing the
                programs.\3106\ API commented against multipliers, stating that the
                program should be technology neutral and that regulatory agencies
                should not incentivize either producer or consumer investments in
                government-selected technologies applied to government-selected vehicle
                categories.\3107\
                ---------------------------------------------------------------------------
                 \3104\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
                 \3105\ ACEEE, Detailed Comments, NHTSA-2018-0067-12122.
                 \3106\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3107\ API, Detailed Comments, EPA-HQ-OAR-2018-0283-5458.
                ---------------------------------------------------------------------------
                 In this final rule, EPA is neither adopting any additional EV or
                FCV multipliers nor extending the existing multipliers scheduled to
                phase out after MY 2021 for EVs, PHEVs, and FCVs. EPA is concerned that
                additional multiplier incentives beyond those already in place for
                these vehicles which are currently available to consumers would reduce
                the emissions benefits associated with the program. As discussed below
                in section IX.B.1.a.(3)(b), EPA is providing an additional multiplier
                for dedicated and dual-fuel NGVs, which are not currently produced by
                auto manufacturers, for MYs 2022-2026. The CO2 program
                already provides a significant incentive for PHEVs, EVs, and FCVs by
                only counting tailpipe emissions (not accounting for upstream
                emissions).
                (3) Special Considerations
                (a) Incentives for Connected or Automated Vehicles
                 Connected and automated (including autonomous) vehicles have the
                potential to impact significantly vehicle emissions in the future, with
                their aggregate impact being either positive or negative, depending on
                a large number of vehicle-specific and system-wide factors. EPA noted
                in the proposal that connected or automated vehicles would be eligible
                for credits under the off-cycle program if a manufacturer provides data
                sufficient to demonstrate the real-world emissions benefits of such
                technology applied to its vehicles. However, demonstrating the
                incremental real-world benefits of these emerging technologies will be
                challenging. Prior to the proposal, stakeholders suggested that EPA
                should consider an incentive for these technologies without requiring
                individual manufacturers to demonstrate real-world emissions benefits
                of the technologies. A number of stakeholders also requested that EPA
                consider credits for automated and connected vehicles that are placed
                in ridesharing or other high mileage applications, where any potential
                environmental benefits could be multiplied due to the high utilization
                of these vehicles. EPA requested comment on such incentives as a way to
                facilitate increased use of these technologies, including some level of
                assurance that they will lead to future additional emissions
                reductions. For example, EPA stated in the proposal that any near-term
                incentive program should include some demonstration that the
                technologies will be both truly new and have some connection to overall
                environmental benefits. EPA further outlined and sought comment on
                several approaches
                [[Page 25209]]
                to incentivize automated and connected vehicle technologies.
                 EPA received comments supporting and opposing incentives for
                automated and connected vehicles. The Auto Alliance commented that the
                agencies should incentivize the adoption of these technologies and
                provide for possibly additional credit once the benefits beyond the
                credit values have been confirmed.\3108\ It further commented that a
                growing body of modeling results, as well as real-world driving
                statistics, show that current automated driving technologies improve
                real-world fuel efficiency and reduce CO2 emissions. SAFE
                commented that connected automated vehicles have tremendous potential
                to save lives, and when combined with ride-sharing and electric
                powertrains, they can also increase efficiencies and save fuel.\3109\
                SAFE argued that an initial review of the literature shows the
                potential for these technologies to improve fuel economy by up to 25
                percent when they are optimized and aggregated alongside other
                traditional efficiency technologies. Toyota commented that automated
                vehicles, and possibly new mobility models such as ridesharing, can
                help attain societal goals concerning climate change, energy security,
                traffic congestion, and safety.\3110\ Ford commented that it is
                supportive of credits for future connected and automated vehicles and
                that autonomous vehicles are considered the future of personal
                mobility, with many manufacturers announcing plans to release
                autonomous-capable vehicles in the near term.\3111\ Ford added that
                these vehicles have the potential to not only provide meaningful real-
                world CO2 and fuel economy benefits, but also add true
                societal benefit for the public good by providing transportation to
                those who would otherwise not have access. General Motors and Jaguar
                Land Rover commented in favor of additional credits for vehicles placed
                in ride-sharing or high mileage applications.\3112\
                ---------------------------------------------------------------------------
                 \3108\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3109\ SAFE, Detailed Comments, NHTSA-2018-0067-11981.
                 \3110\ Toyota, Detailed Comments, NHTSA-2018-0067-12150.
                 \3111\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
                 \3112\ General Motors, Detailed Comments, NHTSA-2018-0067-11858;
                Jaguar Land Rover, Detailed Comments, NHTSA-2018-0067-11916.
                ---------------------------------------------------------------------------
                 SAFE commented that autonomous vehicles will lead to new jobs and
                better worker productivity. It stated that these vehicles will also
                reduce congestion and lead to safer travel.\3113\
                ---------------------------------------------------------------------------
                 \3113\ SAFE, Detailed Comments, NHTSA-2018-0067-11981.
                ---------------------------------------------------------------------------
                 Other commenters opposed incentives for automated and connected
                vehicles, generally commenting that while the technologies are
                promising, the impacts of the technologies remain highly uncertain and
                therefore incentives are not appropriate. ACEEE commented that EPA
                should not incentivize technologies such as automated vehicle
                technology or ridesharing services, unless and until it can be
                demonstrated that such an incentive will result in emissions reduction
                benefits and will not undermine the existing standards.\3114\ ACEEE
                believes that there currently exists no real-world data to justify
                granting of off-cycle credits for automated vehicle technologies, and
                that providing automakers credits for deploying technologies which are
                driven by demands other than fuel savings and emissions reduction only
                allows them to make fewer real-world emissions reductions elsewhere.
                ACEEE further stated that while automated vehicles promise all-new
                possibilities and efficiencies in transportation and the use of
                infrastructure, the net impact on transportation sector energy use and
                emissions is unknown.
                ---------------------------------------------------------------------------
                 \3114\ ACEEE, Detailed Comments, NHTSA-2018-0067-12122.
                ---------------------------------------------------------------------------
                 UCS commented that the ``evidence to-date does not warrant
                incentivizing such technologies--there is no provable environmental
                benefit of such technologies, and the agencies have previously
                correctly acknowledged that any such potential impacts would be related
                to indirect benefits, which raise serious concerns about compliance and
                enforcement to ensure the integrity of the program.'' \3115\ Honda
                commented that there remains considerable uncertainty in the literature
                regarding the energy and environmental benefits (or negative benefits)
                of connected/automated vehicle technology.\3116\ Honda commented that
                if technology benefits can be verified under robust, repeatable
                conditions, they should warrant off-cycle credits under the existing
                off-cycle program. Honda does not believe credits should be granted for
                application of technology alone.
                ---------------------------------------------------------------------------
                 \3115\ U.S.C., Detailed Comments, NHTSA-2018-0067-12039.
                 \3116\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                ---------------------------------------------------------------------------
                 CARB commented that new compliance flexibilities (or off-cycle
                credit categories) for automated vehicles are not appropriate at this
                time.\3117\ CARB believes that, although the technology is widely
                expected to provide safety and mobility benefits, automakers are
                expected to bring the technology to market regardless, so incentives
                are unnecessary, and it is not established that these technologies will
                reduce emissions given their potential for high annual mileage.
                Resources for the Future commented they do not see a rationale for
                providing special credits to automated vehicles since such vehicles
                could increase or decrease emissions.\3118\ Competitive Enterprise
                Institute (CEI) commented that some connected and/or automated vehicle
                technology applications--namely platooning--may improve fuel efficiency
                through improved aerodynamics and thus reduce CO2 emissions;
                however, such applications to date are limited to heavy-vehicle
                prototypes beyond the scope of this rulemaking and in any event should
                be subject to verification prior to any award of off-cycle
                credits.\3119\ CEI commented further: ``We urge EPA to preserve the
                existing off-cycle program requirement that manufacturers demonstrate
                CO2 emissions reductions prior to the award of credits,
                rather than picking technology winners and losers that have nothing to
                do with fuel economy or emissions.'' National Association of Truck Stop
                Operators (NATSO) commented against incentives, stating that although
                automated vehicles have the potential positively to transform
                transportation (and indeed day-to-day life) in the U.S., there are also
                a number of complexities and potential costs associated with
                them.\3120\
                ---------------------------------------------------------------------------
                 \3117\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
                 \3118\ Resources for the Future, Detailed Comments, NHTSA-2018-
                0067-11789.
                 \3119\ CEI, Detailed Comments, EPA-HQ-OAR-2018-0283-4166.
                 \3120\ NATSO, Detailed Comments, EPA-HQ-OAR-2018-0283-5484.
                ---------------------------------------------------------------------------
                 EPA is not adopting new incentives for automated and connected
                vehicles. While EPA agrees there may be potential for such technologies
                to reduce emissions long-term, depending on how the technologies are
                developed, implemented, and used, EPA remains concerned about the high
                degree of uncertainty regarding the impacts of the technologies and
                potential loss of emissions reductions associated with such incentives.
                EPA agrees with the comments that, at this time, it is more appropriate
                for manufacturers to seek credits through the existing off-cycle
                credits program where manufacturers would be required to provide data
                demonstrating direct emissions improvements for the technologies.
                [[Page 25210]]
                (b) Natural Gas Vehicle (NGV) Credits
                 Vehicles that are able to run on compressed natural gas (CNG) are
                eligible for an advanced technology multiplier credit for MYs 2017-
                2021, as discussed in the Advanced Technology Incentives section above.
                Dual-fueled natural gas vehicles, which can run either on natural gas
                or on gasoline, also may use utility factors higher than 0.5 when
                weighting tailpipe emissions measured over the test procedures while
                operating on natural gas and gasoline test fuels if the vehicles meet
                minimum design criteria, including minimum CNG range requirements.
                Prior to the proposal, EPA received input from several industry
                stakeholders that supported expanding these incentives to stimulate
                production of vehicles capable of operating on natural gas, including
                treating incentives for natural gas vehicles on par with those for
                electric vehicles and other advanced technologies, and adjusting or
                removing the minimum range requirements for dual-fueled CNG vehicles.
                EPA requested comments on these potential additional incentives for
                natural gas fueled vehicles.
                 Among comments received regarding incentives for NGVs, Ariel
                Corporation and VNG together commented that NGVs can be effectively
                promoted by providing a level playing field and regulatory parity with
                EVs.\3121\ They stated, ``an effective alternative compliance pathway
                for NGVs can be established with a few simple changes to the
                regulations including applying the '0.15 divisor' to emissions
                calculations, which would harmonize EPA's regulations with the
                statutory CAFE program, and recognize the real-world emissions benefits
                of RNG [renewable natural gas], and provide NGVs with reasonable parity
                with EVs.'' Ariel and VNG commented also that EPA should offer advanced
                technology production multipliers for NGVs on par with EVs and FCVs,
                with NGVs receiving these incentives at the same level and for the same
                duration as electric and fuel-cell vehicles. These commenters believe
                that while NGVs have lower technology hurdles than these vehicles, they
                face similar infrastructure challenges and offer similar or superior
                emissions benefits through the use of RNG.
                ---------------------------------------------------------------------------
                 \3121\ Joint Submission from Ariel Corp. and VNG.co, Detailed
                Comments, NHTSA-2018-0067-7573.
                ---------------------------------------------------------------------------
                 Coalition for Renewable Natural Gas, NGVAmerica, the American Gas
                Association, and the American Public Gas Association commented in a
                joint submission that NHTSA and EPA should use this rulemaking
                opportunity to expand incentives for NGVs and thereby increase the
                availability of NGVs in the light-duty sector, particularly for pickup
                trucks, work vans, and sport utility vehicles.\3122\ These commenters
                also submitted comments supporting additional incentives for full-size
                pickup NGVs and incentives for vehicles equipped to be converted to
                operate on natural gas. Coalition for Renewable Natural Gas, et al.,
                commented that allowing 0 grams/mile accounting for electricity use is
                favorable to electric vehicles because it allows electric vehicle
                manufacturers to take credit for anticipated improvements in emissions
                associated with the electric grid resulting from increased use of
                natural gas and renewable energy.\3123\ It further commented that given
                the significant amount of renewable natural gas currently being used
                and projected to be used in future years, using a factor of 0.15 or
                even greater to offset NGV emissions is warranted because RNG use
                reduces carbon dioxide emissions by 85 percent or more in most cases.
                Ingevity similarly commented in support of EPA including a 0.15
                multiplier incentive for purposes of CO2 compliance parity
                between natural gas and electric dual-fuel vehicles as necessary and
                critical to promote the commercialization of light-duty natural gas
                vehicles and stimulate the increased utilization of RNG. Ingevity added
                that growth in the natural gas vehicle market is necessary to meet
                future RFS obligations.\3124\
                ---------------------------------------------------------------------------
                 \3122\ Joint Submission from the Coalition for Renewable Natural
                Gas, NGVAmerica, the American Gas Association, and the American
                Public Gas Association, Detailed Comments, NHTSA-2018-0067-11967.
                 \3123\ Joint Submission from the Coalition for Renewable Natural
                Gas, NGVAmerica, the American Gas Association, and the American
                Public Gas Association, Detailed Comments, NHTSA-2018-0067-11967.
                 \3124\ Ingevity, Detailed Comments, NHTSA-2018-0067-8666.
                ---------------------------------------------------------------------------
                 United States Senator James M. Inhofe commented that ``even if all
                current incentives for EVs are eliminated, EVs still have a compliance
                advantage going forward. This is because the policy and technical
                approaches underlying the [CO2] regulations embedded
                preferential treatment for the previous administration's favored
                technology. I respectfully ask you not to give NGVs preferential
                treatment, but to level the playing field to allow the marketplace to
                determine the future of NGV adoption and not the federal bureaucracy.
                To achieve this parity, reinstating the 0.15 [CO2]
                multiplier is essential.'' \3125\
                ---------------------------------------------------------------------------
                 \3125\ James M. Inhofe, Detailed Comments, EPA-HQ-OAR-2018-0283-
                7456.
                ---------------------------------------------------------------------------
                 In addition to supporting the application of a 0.15 factor, some in
                the natual gas industry also commented in support of production
                multipliers for NGVs. Ariel and VNG commented that EPA should offer
                advanced technology production multipliers for NGVs on par with EVs and
                FCVs, with NGVs receiving these incentives at the same level and for
                the same duration as electric and fuel cell vehicles. Ingevity
                commented that dual-fuel and dedicated NGV multipliers should be
                extended through 2025 as an effective way to promote the
                commercialization of these kinds of vehicles by the automakers. NGV
                America et al. commented that ``NGVs, both dedicated and dual-fuel,
                should be provided with the same vehicle production multiplier credits
                as have previously been, and continue to be, provided to EVs and FCVs.
                Given that the expected and likely range capabilities of NGVs will
                generally exceed EV ranges (including natural gas dual-fuel vehicles
                that significantly outperform the range capabilities of PHEVs which
                justifiably enjoy a lower multiplier as compared to EVs), the vehicle
                production multipliers that are used for EVs should be applied to NGVs,
                including dual fuel NGVs. Specifically, dedicated and dual-fuel NGVs
                (or all covered advanced technology vehicles) should receive a base
                multiplier of 2.0 (or any such higher multiplier afforded to EVs/FCVs)
                for at least model years 2019 through 2021 and the same multipliers
                afforded to EVs/FCVs thereafter through 2025.''
                 National Association of Convenience Stores (NACS) and the Society
                of Independent Gasoline Marketers of America (SIGMA) commented, ``the
                Associations urge you to treat all fuels and technologies equally,
                including NGVs, EVs, and petroleum-based motor fuels. It is the role of
                the Agencies to set performance specifications via notice-and-comment
                rulemaking to ensure that they are appropriate. Once the specifications
                are set, however, it should be up to the market to determine how best
                to meet them.'' \3126\
                ---------------------------------------------------------------------------
                 \3126\ Joint submission on behalf of NACS and SIGMA, Detailed
                Comments, EPA-HQ-OAR-2018-0283-5824.
                ---------------------------------------------------------------------------
                 UCS commented that natural gas is a potent greenhouse gas, and any
                direct emissions of methane pose a significant threat to any effort to
                limit climate change.\3127\ UCS stated, ``these direct emissions
                upstream significantly
                [[Page 25211]]
                undermine any potential benefit that could come from the pump-to-wheel
                benefits of displacing gasoline or diesel with natural gas.'' UCS also
                commented, ``furthermore, the technology underpinning any natural gas-
                powered vehicle is exceptionally mundane--natural gas has been deployed
                previously in vehicles like the Honda Civic, and aftermarket CNG
                conversions have long been available on the market. Again, there is no
                critical hurdle to overcome with CNG powered vehicles, and there is
                little if any benefit to any such incentives. We strongly recommend
                that EPA eliminate all incentives for natural gas vehicles and instead
                ensure such vehicles are credited commensurate with their impact on the
                environment.'' CARB also commented that new compliance flexibilities
                for NGVs are not appropriate at this time.\3128\
                ---------------------------------------------------------------------------
                 \3127\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
                 \3128\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 The Natural Gas Vehicles of America (NGVAmerica) commented that
                there is no incentive under existing EPA and NHTSA regulations for an
                automaker to sell vehicles equipped to be converted to operate on
                natural gas (so-called ``gaseous-prep vehicles''), even though selling
                such vehicles often results in the increased availability of
                alternative fuel vehicles. Today, most alternative fuel conversions are
                performed on newly manufactured gaseous-prep vehicles or vehicles that
                have been equipped by the original equipment manufacturers with
                hardened valves, valve seats, pistons, and piston rings. As an example,
                most of Ford's commercial truck line-up is available as gaseous-prep,
                and many such vehicles are converted to natural gas or propane by
                qualified vehicle manufacturers. Converting these vehicles, producing
                an assembly-line gaseous-prep vehicle, and sharing diagnostic
                information are critical to ensuring that aftermarket conversions
                perform well in-use and do not degrade the vehicle's emission control
                equipment. Given the complexity of today's automobiles, it is virtually
                impossible to legally convert new vehicles without this level of
                cooperation from vehicle manufacturers.
                 NGVAmerica further commented that providing a regulatory incentive
                for automakers to sell these vehicles would expand the availability of
                gaseous-prep vehicles and increase consumer choice for alternative fuel
                vehicles. EPA, therefore, should provide a credit for selling such
                vehicles if the automaker can verify that the vehicles were
                subsequently upfitted or converted using an EPA certified alternative
                fuel system. Given the significant cost associated with certifying
                vehicles and installing natural gas tanks, there is very little
                likelihood that such an incentive would be abused by automakers. As
                with credits for original equipment manufactured vehicles, the utility
                factor for these vehicles would be based on the range of the vehicle
                when operating on natural gas. In this way, vehicles with larger range
                would earn more credit and vehicles with reduced range would earn less
                credit.
                 Regarding comments that EPA should provide additional credits to
                auto manufacturers for the potential use of RNG due to upstream
                benefits associated with the production of RNG by applying a 0.15
                factor, EPA disagrees because auto manufacturers would not be required
                to ensure such fuels are used in the vehicles they produce over the
                life of those vehicles. Commenters provided a rationale for why they
                believe all NGVs produced in the future will be fueled with RNG, but
                EPA believes there is no assurance that this would be the case. If
                fossil fuel-based natural gas is used in the vehicles, the
                environmental benefits asserted by the commenters would not exist and
                the substantial vehicle incentives recommended by the commenters would
                result in a loss of environmental benefits. EPA does not believe it is
                appropriate to attribute most or all of the potential benefits of the
                production and use of RNG to the vehicle manufacturer. EPA's Renewable
                Fuel Standards (RFS) already appropriately credit RNG use as compared
                to fossil fuel-based natural gas. The RFS program provides a
                substantial incentive for RNG production, and those incentives may lead
                to even lower fuel pricing and greater demand for RNG as vehicle fuel,
                and for NGVs in the future. The RFS program also can provide incentives
                for liquid cellulosic fuels, advanced bio-diesel, and other types of
                renewable transportation fuels. Consistent with EPA's decision not to
                include upstream emissions associated with electricity use for EVs and
                PHEVs discussed above, EPA believes it is appropriate at this time to
                maintain the focus of the light-duty vehicle GHG standards on the
                capabilities of the vehicle to control emissions, and not rely on
                lifecycle fuel characteristics as a basis for developing specific
                vehicle incentives, particularly where those fuels are already
                incentivized by the RFS program.
                 After considering comments regarding incentive multipliers for NGVs
                and the current lack of light-duty NGV offerings by OEMs in the market,
                EPA has decided to include a multiplier incentive of 2.0 for MY 2022-
                2026 dedicated and dual-fuel NGVs. This multiplier will go into effect
                when the previously established multipliers expire, thus extending the
                mulipler for NGVs for 5 years beyond those previously established for
                NGVs. While other alternative fuel vehicles that were provided
                multiplier incentives are increasingly available in the light-duty
                marketplace, no OEM is currently offering light-duty NGVs. Since Honda
                ended production of the CNG version of the Honda Civic at the end of MY
                2015, there have been no OEM NGV offerings available to consumers. EPA
                continues to believe that NGVs could be an important part of the
                overall light-duty vehicle fleet mix, and such offerings would enhance
                the diversity of potentially cleaner alternative fueled vehicles
                available to consumers.\3129\ EPA believes it is appropriate to extend
                the availability of a production multiplier through MY 2026 for both
                dual-fuel and dedicated NGVs to potentially help spur their re-
                introduction by OEMs in the light-duty vehicle market.
                ---------------------------------------------------------------------------
                 \3129\ The CNG Honda Civic had approximately 20 percent lower
                CO2 than the gasoline Civic in MY 2015.
                ---------------------------------------------------------------------------
                 EPA also received comments on the application of the regulatory
                utility factor. For dual-fuel vehicles, emissions are measured on both
                fuels (e.g., gasoline and natural gas) and weighted using a factor
                referred to in the regulations as a utility factor. To use a utility
                factor for natural gas greater than 0.5, a dual-fuel NGV must meet
                design criteria requiring the vehicle to have a natural gas to gasoline
                driving range of 2:1. The vehicle must also preferentially operate on
                natural gas until the natural gas tank is empty. EPA adopted these
                design criteria as part of the 2012 final rule to help ensure vehicles
                using a utility factor of higher than 0.5 would likely be fueled with
                and use natural gas most of the time on the road. At that time, EPA was
                concerned that natural gas refueling may be much more inconvenient for
                drivers relative to electric charging for PHEVs due to a lack of CNG
                refueling stations (or home refueling, compared to the availability of
                home chargers for many PHEVs) and, therefore, dual-fuel vehicles with
                limited driving range on natural gas would likely frequently operate on
                gasoline.
                 EPA received comments regarding the design criteria. Ingevity
                commented that it has developed a low-pressure (900 psi) adsorbed
                natural gas (ANG) fuel storage technology that allows vehicles to be
                refueled using an affordable and reliable low-pressure natural gas
                fueling
                [[Page 25212]]
                appliance.\3130\ Ingevity commented that ANG will allow for a
                distributed refueling network at users' homes and businesses, just like
                electrical recharging equipment has been installed for PHEVs over the
                last several years. Ingevity commented that the design criteria for
                dual-fuel NGVs that were established in the MYs 2017-2025 final rule
                ``make it impossible to reasonably and affordably manufacture a dual-
                fuel NGV that can fully utilize the utility factor (UF) approach for
                determining fuel economy and [CO2] emissions.'' Ingevity
                recommended that the design criteria for dual-fuel NGVs be removed and
                that the utility factor be based only on the range of the NGV on
                natural gas, equivalent to the treatment of PHEVs. MECA submitted
                similar comments regarding ANG technology.\3131\
                ---------------------------------------------------------------------------
                 \3130\ Ingevity, Detailed Comments, NHTSA-2018-0067-8666.
                 \3131\ See MECA, Detailed Comments, NHTSA-2018-0067-11999.
                ---------------------------------------------------------------------------
                 Ariel and VNG also commented that design criteria imposed on dual-
                fuel NGVs add unnecessary costs and complexity, and currently are
                arbitrarily applied only to dual-fuel NGVs, and not to their dual-fuel
                hybrid counterparts.\3132\ NACS, SIGMA, and NATSO also recommended that
                EPA remove eligibility requirements associated with the utility
                factor.\3133\
                ---------------------------------------------------------------------------
                 \3132\ Joint Submission from Ariel Corp. and VNG, Detailed
                Comments, NHTSA-2018-0067-7573.
                 \3133\ Joint submission on behalf of NACS and SIGMA, Detailed
                Comments, EPA-HQ-OAR-2018-0283-5824; NATSO, Detailed Comment, EPA-
                HQ-OAR-2018-0283-5484.
                ---------------------------------------------------------------------------
                 After considering the comments, EPA is removing the design criteria
                from the regulations and thereby allowing higher utility factors to be
                used for dual-fuel natural gas vehicles based solely on driving range
                on natural gas, as is the case for PHEVs. The utility factor represents
                a reasonable way of weighting the emissions of a dual-fuel vehicle on
                each fuel to derive a single emissions value when including the dual-
                fuel vehicles in a manufacturer's fleet average compliance
                determination. Ideally, the utility factor would match the use of each
                fuel in real-world vehicle operation. The utility factor is not meant
                to incentivize the adoption of a particular technology, so it differs
                fundamentally from incentives such as multipliers. With the development
                of low-pressure natural gas vehicle fueling system technology since the
                2012 final rule, EPA's concerns regarding limited fueling
                infrastructure that led the agency to adopt the design criteria in the
                2012 rule are significantly diminished. EPA believes that low-pressure
                fueling is a new advancement that offers the potential for more
                convenient refueling for individuals or businesses similar to that for
                PHEVs. EPA expects owners of dual-fuel CNG vehicles preferentially to
                seek to refuel and operate on CNG fuel as much as possible, both
                because the owner would have to pay a higher vehicle price for the
                dual-fuel capability, and because CNG fuel is considerably cheaper than
                gasoline. With the opportunity for relatively low-cost on-site
                refueling at homes or businesses, EPA expects such vehicles to be
                refueled with natural gas similar to how people refuel PHEVs. Vehicle
                purchasers that choose high pressure vehicle systems over low pressure
                systems would likely do so only if they have ready access to a high
                pressure refueling system, for example, at a fleet's central fueling
                location. Removing the design criteria for dual-fuel natural gas
                vehicles also addresses the concerns of some commenters regarding the
                differing treatment of PHEVs and dual-fuel NGVs.
                 EPA believes that with the advancement of technology offering the
                potential for more flexible refueling of NGVs, removing the design
                criteria is a reasonable change to the regulations. This regulatory
                change will apply starting with MY 2021. MY 2021 will provide
                sufficient time for orderly implementation and EPA is not aware of any
                dual-fuel NGVs emissions certified for MYs 2019-2020 that would
                otherwise be affected if this change were to be implemented sooner.
                 EPA received comments that vehicle conversions and ``gaseous-prep''
                vehicles should be eligible for credits. In response to comments on
                vehicle conversions, alternative fuel converters are not required to
                meet fleet average standards but instead may comply with 40 CFR part 85
                subpart F regulations providing a tampering exemption. Fleet average
                standards are generally not appropriate for fuel conversion
                manufacturers because the ``fleet'' of vehicles to which a conversion
                system may be applied has already been accounted for under the OEM's
                fleet average standard. Alternative fuel converters are not
                manufacturing new vehicles, but are converting existing vehicles that
                have already been certified by the OEM. CO2 credits are
                available to OEMs based on fleet emissions performance compared to the
                fleet average standards and therefore conversions are not eligible for
                these credits. EPA did not propose to change and is not changing the
                exemption process promulgated in 40 CFR part 85 subpart F. Because fuel
                conversions are not required to meet the fleet average standards,
                credits generated under those standards are not available. Regarding
                gaseous-prep vehicles, these vehicles are not NGVs at initial sale and
                therefore are not eligible for NGV incentives. Instead, they are
                included in the OEM's fleet as gasoline-only vehicles. EPA disagrees
                with the commenters that such vehicles should be eligible for NGV
                incentives at time of initial sale if the vehicle is later converted to
                natural gas since the OEM does not measure the emissions of the vehicle
                on natural gas at time of certification and is not responsible for the
                emissions performance of the vehicle on natural gas over the life of
                the vehicle.
                C. NHTSA Compliance and Enforcement
                1. Overview of the NHTSA Compliance Process
                 Consumer choice drives the mixture of automobiles on the road.
                Manufacturers largely produce a mixture of vehicles to meet consumer
                demand and address compliance with CAFE standards though the
                application of fuel economy improving technologies to those vehicles,
                and by using compliance flexibilities and incentives that are available
                in the CAFE program. As discussed earlier in this notice, each vehicle
                manufacturer is subject to separate CAFE standards for passenger cars
                and light trucks, and for the passenger car standards, a manufacturer's
                domestically-manufactured and imported passenger car fleets are
                required to comply separately.\3134\ Additionally, domestically-
                manufactured passenger cars are subject to a statutory minimum
                standard.\3135\ CAFE program flexibilities are largely provided for in
                statute. Credits for air conditioning efficiency, off-cycle, and pickup
                truck advanced technologies are not expressly specified by CAFE
                statute, but are ``implemented consistent with EPCA's provisions
                regarding calculation of fuel economy'' as discussed in section C.2
                below.
                ---------------------------------------------------------------------------
                 \3134\ 49 U.S.C. 32904(b).
                 \3135\ 49 U.S.C. 32902(b)(4).
                ---------------------------------------------------------------------------
                 Compliance with the CAFE program begins with manufacturers
                submitting required reports to NHTSA in advance and during the model
                year that contain information, specifications, data, and projections
                about their fleets.\3136\ Manufacturers report early product
                projections to NHTSA describing their efforts to comply with CAFE
                standards per EPCA's reporting requirements.\3137\ Manufacturers' early
                projections are required to identify any of the
                [[Page 25213]]
                flexibilities and incentives manufacturers plan to use for air-
                conditioning (A/C) efficiency, off-cycle and, through MY 2021, full-
                size pickup truck advanced technologies. EPA consults with NHTSA when
                reviewing and considering manufacturers' requests for fuel consumption
                improvement values for A/C and off-cycle technologies that improve fuel
                economy. NHTSA evaluates and monitors the performance of the industry
                using the information provided. NHTSA also audits manufacturers'
                projected data for conformance and verifies vehicle design data through
                testing to ensure manufacturers are complying as projected. After the
                model year ends, manufacturers submit final reports to EPA, including
                final information on all the flexibilities and incentives allowed or
                approved for the given model year.\3138\ EPA then calculates the fuel
                economy level of each fleet produced by each manufacturer, and
                transmits that information to NHTSA.\3139\
                ---------------------------------------------------------------------------
                 \3136\ 49 U.S.C. 32907(a); 49 CFR 537.7.
                 \3137\ 49 U.S.C. 32907(a).
                 \3138\ For example, alternative fueled vehicles get special
                calculations under EPCA (49 U.S.C. 32905-06), and fuel economy
                levels can also be adjusted to reflect air conditioning efficiency
                and ``off-cycle'' improvements, as discussed below.
                 \3139\ 49 U.S.C. 32904(c)-(e). EPCA granted EPA authority to
                establish fuel economy testing and calculation procedures; EPA uses
                a two-year early certification process to qualify manufacturers to
                start selling vehicles, coordinates manufacturer testing throughout
                the model year, and validates manufacturer-submitted final test
                results after the close of the model year.
                ---------------------------------------------------------------------------
                 NHTSA notes that some manufacturers have submitted and/or
                resubmitted requests for A/C and off-cycle benefits after EPA final
                reports are completed or nearly completed and, in those cases, such
                submissions are causing considerable delays in EPA's ability to
                finalize CAFE reports. Late and revised submissions can place
                significant burdens on the government in order to reassess a
                manufacturer's CAFE performances and standards and can also cause
                significant impacts on previous compliance model years. In the
                following sections, EPA and NHTSA are incorporating regulatory
                modifications or providing guidance to help manufacturers expedite
                approvals and to facilitate the governments processing of the
                flexibilities and incentives.
                 NHTSA determines each manufacturer's obligation to comply with
                applicable model year's CAFE standards and notifies the manufacturer if
                any of its fleet performances fall below standards. Manufacturers must
                submit plans detailing the compliance flexibilities to be used to
                resolve any possible noncompliances or may pay civil penalties to
                address any deficits for falling below standards. NHTSA periodically
                releases data and reports to the public through its CAFE Public
                Information Center (PIC) based on information in the EPA final reports
                for the given compliance model year, and based on the projections
                manufacturers provide to NHTSA for the next two model years.\3140\
                ---------------------------------------------------------------------------
                 \3140\ NHTSA CAFE Public Information Center, https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Home.htm.
                ---------------------------------------------------------------------------
                2. NHTSA's CAFE Program Compliance
                 EPCA and EISA specify several flexibilities and incentives that are
                available to help manufacturers comply with CAFE standards. Some
                flexibilities are defined, and sometimes limited by statute--for
                example, while Congress allowed manufacturers to transfer credits
                earned for over-compliance from their car fleet to their truck fleet
                and vice versa, Congress also limited the amount by which manufacturers
                could increase their CAFE levels using those transfers.\3141\
                Consistent with the limits Congress placed on certain statutory
                flexibilities and incentives, NHTSA crafted and implements the credit
                transfer and trading regulations authorized by EISA to help ensure that
                total fuel savings are preserved when manufacturers exercise statutory
                compliance flexibilities.
                ---------------------------------------------------------------------------
                 \3141\ See 49 U.S.C. 32903(g).
                ---------------------------------------------------------------------------
                 NHTSA and EPA have previously developed other compliance
                flexibilities and incentives for the CAFE program consistent with the
                statutory provisions regarding EPA's calculation of manufacturers' fuel
                economy levels. As discussed previously, NHTSA finalized in the 2012
                final rule, for MYs 2017 and later, an approach for manufacturers'
                ``credits'' under EPA's program to be applied as fuel economy
                ``adjustments'' or ``improvement values'' under NHTSA's program for:
                (1) Technologies that cannot be measured or cannot be fully measured on
                the 2-cycle test procedure, i.e., ``off-cycle'' technologies; and (2)
                A/C efficiency improvements that also improve fuel economy but cannot
                be measured on the 2-cycle test procedure. Additionally, both agencies'
                programs give manufacturers compliance incentives through MY 2021 for
                utilizing specified technologies on pickup trucks, such as pickup truck
                hybridization.
                 The following sections outline how NHTSA determines whether
                manufacturers are in compliance with the CAFE standards for each model
                year, and how manufacturers may use compliance flexibilities, or
                address noncompliance by paying civil penalties. As addressed above,
                some compliance flexibilities are expressly prescribed in statute and
                some are implemented consistent with EPCA's provisions regarding
                calculation of fuel economy. NHTSA proposed new language updating and
                clarifying existing regulatory text in this area as part of the NPRM.
                NHTSA also sought comments in the NPRM on these changes, as well as on
                the general efficacy of these flexibilities in the fuel economy and
                CO2 programs.
                 Moreover, the following sections explain how manufacturers submit
                data and information to the agency. As part of the NPRM, NHTSA proposed
                to implement a new standardized template for manufacturers to use to
                submit CAFE data to the agency, as well as a standardized template for
                reporting credit transactions. Additionally, NHTSA proposed adding
                requirements that specify the precision of the fuel savings adjustment
                factor in 49 CFR 536.4. These new requirements are intended to
                streamline reporting and data collection from manufacturers, in
                addition to helping the agency use the best available data to inform
                CAFE program decision makers. The comments received to these proposals
                are included in Section IX.C.2.a)(2)(d) along with NHTSA's responses to
                the comments and final resolutions established in the final rule.
                 NHTSA also sought comments on removing or modifying certain CAFE
                program flexibilities. The comments received and NHTSA's responses to
                those comments are discussed below.
                a) How does NHTSA determine compliance?
                (1) Manufacturers Submit Data to NHTSA and EPA and the Agencies
                Validate Results
                 EPCA, as amended by EISA, requires a manufacturer to submit reports
                to the Secretary of Transportation explaining whether the manufacturer
                will comply with an applicable CAFE standard for the model year for
                which the report is made; the actions a manufacturer has taken or
                intends to take to comply with the standard; and other information the
                Secretary requires by regulation.\3142\ A manufacturer must submit a
                report containing the above information during the 30-day period before
                the beginning of each model year, and during the 30-day period
                beginning the 180th day of the model year.\3143\ When a manufacturer
                determines it is unlikely to comply with a CAFE standard, the
                manufacturer must report additional
                [[Page 25214]]
                actions it intends to take to comply and include a statement about
                whether those actions are sufficient to ensure compliance.\3144\
                ---------------------------------------------------------------------------
                 \3142\ 49 U.S.C. 32907(a).
                 \3143\ Id.
                 \3144\ Id.
                ---------------------------------------------------------------------------
                 To implement these reporting requirements, NHTSA issued 49 CFR part
                537, ``Automotive Fuel Economy Reports,'' which specifies three types
                of CAFE reports that manufacturers must submit. A manufacturer must
                first submit a pre-model year (PMY) report containing the
                manufacturer's projected compliance information for that upcoming model
                year. By regulation, the PMY report must be submitted in December of
                the calendar year prior to the corresponding model year.\3145\
                Manufacturers must then submit a mid-model year (MMY) report containing
                updated information from manufacturers based upon actual and projected
                information known midway through the model year. By regulation, the MMY
                report must be submitted by the end of July for the applicable model
                year.\3146\ Finally, manufacturers must submit a supplementary report
                to supplement or correct previously submitted information, as specified
                in NHTSA's regulation.\3147\
                ---------------------------------------------------------------------------
                 \3145\ 49 CFR 537.5(b).
                 \3146\ Id.
                 \3147\ 49 CFR 537.8.
                ---------------------------------------------------------------------------
                 If a manufacturer wishes to request confidential treatment for a
                CAFE report, it must submit both a confidential and redacted version of
                the report to NHTSA. CAFE reports submitted to NHTSA contain estimated
                sales production information, which may be protected as confidential
                until the termination of the production period for that model
                year.\3148\ NHTSA temporarily protects each manufacturer's competitive
                sales production strategies, but does not permanently exclude sales
                production information from public disclosure. Sales production volumes
                are part of the information NHTSA routinely makes publicly available
                through the CAFE PIC.
                ---------------------------------------------------------------------------
                 \3148\ 49 CFR part 512, appx. B(2).
                ---------------------------------------------------------------------------
                 The manufacturer reports provide information on light-duty
                automobiles such as projected and actual fuel economy standards, fuel
                economy performance values, and production volumes, as well as
                information on vehicle design features (e.g., engine displacement and
                transmission class) and other vehicle attribute characteristics (e.g.,
                track width, wheelbase, and other off-road features for light trucks).
                Beginning with MY 2017, to obtain credit for fuel economy improvement
                values attributable to additional technologies, manufacturers must also
                provide information regarding A/C systems with improved efficiency,
                off-cycle technologies (e.g., stop-start systems, high-efficiency
                lighting, active engine warm-up), and full-size pickup trucks with
                hybrid technologies or with emissions/fuel economy performance that is
                better than footprint-based targets by specified amounts. This includes
                identifying the makes and model types equipped with each technology,
                the compliance category those vehicles belong to, and the associated
                fuel economy improvement value for each technology.\3149\ In some
                cases, NHTSA may require manufacturers to provide supplementary
                information to justify or explain the benefits of these technologies
                and their impact on fuel consumption or to evaluate the safety
                implication of the technologies. These details are necessary to
                facilitate NHTSA's technical analyses and to ensure the agency can
                perform enforcement audits as appropriate.
                ---------------------------------------------------------------------------
                 \3149\ NHTSA collects model type information based upon the EPA
                definition for ``model type'' in 40 CFR 600.002.
                ---------------------------------------------------------------------------
                 NHTSA uses manufacturer-submitted PMY, MMY, and supplementary
                reports to assist in auditing manufacturer compliance data and
                identifying potential compliance issues as early as possible.
                Additionally, as part of its footprint validation program, NHTSA
                conducts vehicle testing throughout the model year to confirm the
                accuracy of the track width and wheelbase measurements submitted in the
                reports.\3150\ These tests help the agency better understand how
                manufacturers may adjust vehicle characteristics to change a vehicle's
                footprint measurement, and ultimately its fuel economy target. NHTSA
                also includes a summary of manufacturers' PMY and MMY data in an annual
                fuel economy performance report made publicly available on its PIC.
                ---------------------------------------------------------------------------
                 \3150\ U.S. Department of Transportation, NHTSA, Laboratory Test
                Procedure for 49 CFR part 537, Automobile Fuel Economy Attribute
                Measurements (Mar. 30, 2009), available at http://www.nhtsa.gov/DOT/NHTSA/Vehicle%20Safety/Test%20Procedures/Associated%20Files/TP-537-01.pdf.
                ---------------------------------------------------------------------------
                 NHTSA uses EPA-verified final-model year (FMY) data to evaluate
                manufacturers' compliance with CAFE program requirements, and draws
                conclusions about the performance of the industry. After manufacturers
                submit their FMY data, EPA verifies the information, accounting for
                NHTSA and EPA testing, and subsequently forwards the final verified
                data to NHTSA.
                (2) Changes to CAFE Reporting Requirements Made by This Final Rule
                 NHTSA proposed changes to its CAFE reporting requirements with the
                intent of streamlining data collection and reporting for manufacturers
                while helping the agency obtain the best available data to inform CAFE
                program decision-makers. The agency developed two new standardized
                reporting templates for manufacturers and proposed to start using the
                templates beginning in the 2019 compliance model year. In the NPRM,
                NHTSA sought comments on the templates. NHTSA's responses to the
                comments received and the changes to the templates for the final rule
                are presented below.
                (a) Standardized CAFE Reporting Template
                 When NHTSA received and reviewed manufacturers' projection reports
                for MYs 2013--2015, the agency observed that most did not conform to
                the requirements specified in 49 CFR part 537. For example, NHTSA
                identified several instances where manufacturers' CAFE reports included
                a ``yes'' or ``no'' response to a request for a vehicle's numerical
                ground clearance values. In a 2015 notice of proposed rulemaking, NHTSA
                proposed to amend 49 CFR part 537 to require a new data format for
                manufacturers' light-duty vehicle CAFE projection reports.\3151\ In
                response to the proposal, some manufacturers commented that the
                previous changes in reporting requirements generated confusion and led
                to reporting errors. NHTSA recognized that the modification to the base
                tire definition in the 2012 final rule for MYs 2017 and later seemed to
                make some manufacturers uncertain about what footprint data was
                required in the reports.\3152\ Specifically, certain manufacturers did
                not understand that the modified base tire definition required them to
                provide estimated attribute-based target standards for each unique
                model type/footprint combination beginning with MY 2013. NHTSA
                discovered cases where manufacturers only provided target or vehicle
                data for certified vehicle configurations, and did not report
                information for each of the unique model type/footprint combinations
                for their available production vehicles in the market. However, NHTSA
                did not adopt the proposed data format from the 2015 proposed rule
                after receiving
                [[Page 25215]]
                adverse comments from manufacturers.\3153\
                ---------------------------------------------------------------------------
                 \3151\ 80 FR 40540 (Jul. 13, 2015).
                 \3152\ 49 CFR 523.2.
                 \3153\ 81 FR 73958 (Oct. 25, 2016).
                ---------------------------------------------------------------------------
                 Since the issuance of the final rule in 2016, NHTSA has continued
                to receive projection reports that contain inaccurate and/or missing
                data. These noncompliant reports impede NHTSA's ability to audit
                manufacturer compliance data, identify potential compliance issues, and
                analyze industry trends. Problems with inaccurate or missing data has
                become an even greater issue for manufacturers reporting on the new MY
                2017 incentives for efficient A/C systems, off-cycle technologies, and
                full-size pickup trucks with hybrid technologies/improved exhaust
                emission performance.\3154\ These incentives are explained in Section
                IX.C.2.c). Manufacturers seeking to take advantage of these new
                benefits must provide information at the model-type level; however,
                many manufacturers did not submit the required information in their PMY
                reports for MYs 2017, 2018, and 2019. This caused NHTSA's Office of
                Enforcement to send letters reminding manufacturers of their obligation
                to submit accurate and complete CAFE reports. NHTSA will continue to
                monitor the accuracy, completeness, and timeliness of manufacturers'
                CAFE reports and may take additional action as appropriate.
                ---------------------------------------------------------------------------
                 \3154\ NHTSA allows manufacturers to use these flexibilities and
                incentives for complying with standards starting in MY 2017; the
                FCIV for full-size pickup trucks with hybrid technologies/improved
                exhaust emission performance applies only through MY 2021, as
                discussed further below.
                ---------------------------------------------------------------------------
                 In the NPRM, NHTSA proposed a new standardized template for
                reporting PMY and MMY information, as specified in 49 CFR 537.7(b) and
                (c), as well as supplementary information required by 49 CFR 537.8. The
                template allows manufacturers to build out the required confidential
                versions of CAFE reports specified in 49 CFR part 537 and to produce
                automatically the required non-confidential versions by clicking a
                button within the template. While NHTSA recognizes that modifications
                to the reporting requirements may initially be a slight inconvenience
                to manufacturers, the number of noncompliant reports the agency
                continues to receive justifies development of a uniform reporting
                method to help ensure compliance with CAFE regulations. Adopting a
                standardized template will assist manufacturers in providing the agency
                with all necessary data, thereby helping manufacturers to ensure they
                are complying with CAFE regulations. The template organizes the
                required data in a manner consistent with NHTSA and EPA regulations and
                simplifies the reporting process by incorporating standardized
                responses consistent with those provided to EPA. The template collects
                the relevant data, calculates intermediate and final values in
                accordance with EPA and NHTSA methodologies, and aggregates all the
                final values required by NHTSA regulations in a single summary
                worksheet. Thus, NHTSA believes that the standardized templates will
                benefit both the agency and manufacturers by helping to avoid reporting
                errors, such as data omissions and miscalculations, and will ultimately
                simplify and streamline reporting.
                 NHTSA proposed to require that manufacturers use the standardized
                template for all PMY, MMY, and supplementary CAFE reports. NHTSA
                observed that a significant number of manufacturers submit their MMY
                reports as updated PMY reports--using the same amount of information,
                despite fewer data requirements. To conform with this method, NHTSA
                designed the template based on one standardized format that uses the
                same data requirements for all CAFE reports. This approach will further
                simplify CAFE projection reporting for manufacturers. The template
                contains a few additional data fields for certain vehicle
                characteristics; however, the inclusion of model type indexes will
                limit the number of required entries by populating a number of pre-
                entered data fields based on one value.
                 The standardized template will also allow NHTSA to modify its
                existing compliance database to accept and import uniform data and
                automatically aggregate manufacturers' data. This will allow NHTSA to
                execute its regulatory obligations more efficiently and effectively.
                Overall, the template will help to ensure compliance with data
                requirements under EPCA/EISA and drastically reduce the industry and
                government's burden for reporting in accordance with the Paperwork
                Reduction Act.\3155\ NHTSA made the template available through its
                docket as well as its PIC, and sought comment on the regulatory changes
                to the reporting process.
                ---------------------------------------------------------------------------
                 \3155\ 44 U.S.C. 3501 et seq.
                ---------------------------------------------------------------------------
                 Comments on the template were received from the Auto Alliance,
                Global Automakers, Ford, Mercedes-Benz, Toyota, Volvo and Volkswagen.
                The Auto Alliance, Toyota, and Volkswagen opposed adopting the proposed
                template; however, Global Automakers agreed with the appropriateness of
                a standardized template that combines credit trading information with a
                data reporting template.\3156\ Global Automakers also made two
                recommendations: (1) Combine EPA's AB&T template with NHTSA's CAFE
                Projections Reporting Template to streamline reporting and reduce
                burden; and (2) add an FMY report requirement as an update to the MMY
                report submission.\3157\
                ---------------------------------------------------------------------------
                 \3156\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Toyota, Detailed Comments, NHTSA-2018-0067-12150; Volkswagen,
                Detailed Comments, NHTSA-2017-0069-0583.
                 \3157\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
                12032.
                ---------------------------------------------------------------------------
                 Mercedes-Benz, Ford, and Volkswagen commented about data fields
                they believed were outdated, or not relevant to fuel economy testing or
                projecting fuel economy performance.\3158\ Mercedes-Benz stated that
                some required data fields are not currently collected as a part of the
                fuel economy testing process, and their capture would require
                additional burden.\3159\ Mercedes-Benz believes those data fields
                should be an optional requirement. Additionally, Mercedes-Benz
                recommended that NHTSA omit certain data fields, and stated that it
                would be helpful if NHTSA clarified its intention for the information
                in others.\3160\ The specific data fields mentioned by Mercedes-Benz
                are in Table IX-6. Ford stated that many of the data fields are
                outdated, have no bearing on compliance assessments, and are misaligned
                with the current reporting structure, which is dictated by model type
                index.\3161\ Similarly, Volkswagen stated that the proposed reporting
                template is populated with many fields that do not immediately appear
                relevant to projecting CAFE performance, align with the existing
                requirements in 49 CFR 537.7, or seem relevant in the space of
                automotive technology.\3162\
                ---------------------------------------------------------------------------
                 \3158\ Daimler Mercedes, Detailed Comments, EPA-HQ-OAR-2018-
                0283-4182; Ford, Detailed Comments, NHTSA-2018-0067-11928;
                Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
                 \3159\ Daimler Mercedes, Detailed Comments, EPA-HQ-OAR-2018-
                0283-4182.
                 \3160\ Daimler Mercedes, Detailed Comments, EPA-HQ-OAR-2018-
                0283-4182.
                 \3161\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
                 \3162\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
                ---------------------------------------------------------------------------
                [[Page 25216]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.756
                 The Auto Alliance and Mercedes-Benz noted the differences in how
                NHTSA and EPA request data on A/C efficiency and off-cycle
                technologies. Mercedes-Benz highlighted the difficulty in predicting
                the projected sales production of the technologies, and the Auto
                Alliance cautioned that the number of reporting entries would increase
                by a factor of ten or more.\3163\ The Auto Alliance stated its belief
                that the change in reporting requirements would cost its members more
                than $1 million in information technology changes and that the changes
                could not be completed prior to MY 2021.\3164\ Likewise, Ford contended
                that an implementation date for MY 2019 is aggressive and does not
                provide manufacturers with adequate lead time.\3165\
                ---------------------------------------------------------------------------
                 \3163\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Daimler Mercedes, Detailed Comments, EPA-HQ-OAR-2018-0283-4182.
                 \3164\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3165\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
                ---------------------------------------------------------------------------
                 The Auto Alliance emphasized that the templates lack common
                reporting standardization with submissions to EPA.\3166\ The Auto
                Alliance, Global Automakers, Toyota, and Volvo all requested that NHTSA
                and EPA accept a single, common reporting format to satisfy reporting
                for both agencies.\3167\ Mercedes-Benz and Volkswagen requested
                stakeholder workshops to review the template with agency staff, with
                the former recommending that NHTSA host the workshops in partnership
                with EPA.\3168\
                ---------------------------------------------------------------------------
                 \3166\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3167\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Global Automakers, Detailed Comments, NHTSA-2018-0067-12032; Toyota,
                Detailed Comments, NHTSA-2018-0067-12150; Volvo, Detailed Comments,
                NHTSA-2018-0067-12036.
                 \3168\ Daimler Mercedes, Detailed Comments, EPA-HQ-OAR-2018-
                0283-4182; Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
                ---------------------------------------------------------------------------
                 Ford requests that NHTSA re-examine the proposed required
                submission methods and reconsider current electronic submission
                methods.\3169\ Ford expressed concern about the efficiency and security
                issues involved in submitting data on a CD through the mail containing
                confidential business information.\3170\ Ford identified what it
                believes are better available avenues for submission, such as secured
                email or online portals like EPA's Central Data Exchange.\3171\
                ---------------------------------------------------------------------------
                 \3169\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
                 \3170\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
                 \3171\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
                ---------------------------------------------------------------------------
                 NHTSA disagrees with many of the manufacturers' assertions.
                Differences in EPA and NHTSA regulations prevent establishing a single
                reporting format for CAFE purposes. For example, EPA only needs early
                model year information for manufacturers' applications for
                certification required under 40 CFR 86.1843-01. Manufacturers submit a
                single application with extensive details for each certified vehicle
                within a test group (i.e., the certified vehicle represents all the
                vehicles within the test group with similar technologies and
                performance characteristics). In comparison, NHTSA's required early
                model year information is far less detailed and is aggregated for model
                types and compliance categories. However, NHTSA and EPA already share
                all the relevant CAFE FMY information pursuant to an interagency
                agreement. This arrangement not only benefits manufacturers but also
                reduces the burden on the Federal government. Since much of the
                required data in NHTSA's projections template is already contained in
                EPA final reports, manufacturers would not be required to generate
                additional information but simply to provide estimates along the way to
                finalizing the data. NHTSA plans to release a data matrix that maps
                data elements between the CAFE template and the EPA final CAFE reports.
                NHTSA will notify the public when the matrix will be available on its
                website. Consequently, there is no need to create an additional final
                report as an updated version of NHTSA's MMY report, as suggested by
                Global Automakers. Once NHTSA configures its CAFE database to accept
                the reporting template via file upload, the agency will be able to use
                the model type index data field to connect data values from the
                template to corresponding values in EPA's final CAFE report.
                Manufacturers should note that CAFE reports are estimated projections
                of the EPA final CAFE compliance data. Contrary to Mercedes concerns
                about the difficulty in predicting the projected sales production of
                the technologies, NHTSA only expects manufacturers to provide the most
                up-to-date information available 30 days before a report is required to
                be submitted to the
                [[Page 25217]]
                Administrator as specified in 49 CFR part 537.5(d). While manufacturer
                PMY reports may be limited in certain instances (excluding vehicles
                already in sales distribution), the MMY reports should be more
                inclusive and closer to the final values reported to EPA. Manufacturers
                should also be submitting supplementary reports to NHTSA if they
                believe there will be significant differences between CAFE MMY reports
                and the EPA final reports.
                 Commenters also stated that the A/C and off-cycle information
                reported in the NHTSA template is inconsistent with the EPA EV-
                CIS.\3172\ NHTSA notes that the inconsistency between the agencies is
                intentional and necessary. NHTSA's off-cycle and A/C information must
                be collected in greater detail than that reported to the EPA EV-CIS.
                NHTSA collects detailed information on A/C and off-cycle technologies
                for determining penetration rates of specific technologies in the
                market, as well as analyzing the types of technologies as equipped on
                specific model types. In comparison, EPA aggregates the data for
                calculating credits, which allows for combining the benefits for all
                the technologies equipped on a model type. NHTSA also will use the
                detailed information for public disclosure and for auditing purposes.
                However, NHTSA acknowledges the Auto Alliance's concerns about the
                burden placed on the industry for providing more detailed data and
                therefore will not require manufacturers to start using the templates
                for reporting until MY 2023. NHTSA also agrees with Ford that it is
                important to consider the issues of security and efficiency with
                respect to the submission of confidential information to the agency,
                and the agency will consider possible changes to its procedures
                relating to the receipt and handling of confidential information to
                ensure streamlined, secure, and efficient submission of confidential
                information, including CAFE reports.\3173\
                ---------------------------------------------------------------------------
                 \3172\ See, e.g., Auto Alliance, Detailed Comments, NHTSA-2018-
                0067-12073.
                 \3173\ See 49 CFR part 512, 537.5.
                ---------------------------------------------------------------------------
                 Secondly, NHTSA agrees with Mercedes-Benz and Volkswagen that
                workshops will aid in implementing the templates by providing
                instruction on how to complete them. NHTSA plans to host a workshop for
                manufacturers to discuss the implementation process. NHTSA believes
                finalizing the template in this rulemaking is important to address
                continuing concerns with reporting noncompliance (i.e., missing,
                incomplete, or inaccurate submissions) with the existing provisions in
                Part 537. Ultimately, establishing the new templates and holding
                educational workshops will be more effective in achieving industry
                compliance than imposing penalties on a case-by-case basis for failure
                to comply with reporting provisions.
                 Finally, NHTSA is also adopting changes to the proposed template in
                response to comments from Mercedes-Benz, Ford, and Volkswagen. NHTSA
                made changes to several of the data fields discussed by Mercedes-Benz.
                NHTSA does not agree with Mercedes-Benz's recommendation to omit the
                ``Type of Overdrive'' or ``Type of Torque Converter'' data fields;
                however, the agency does believe the proposed data to be inserted into
                those fields may be too specific for CAFE purposes. Therefore, the
                agency is finalizing a requirement that manufacturers identify whether
                vehicles are equipped with overdrive or a torque converter by selecting
                ``Yes'' or ``No'' from a dropdown list. The agency has also changed the
                ``Calibration'' field to ``Other Calibration'' to clarify the data
                being requested, and changed the ``Auxiliary Emission Control Device''
                in the ``Fuel Economy'' worksheets to a dropdown that allows users to
                select multiple emission control systems. NHTSA believes that adding
                dropdown lists in the template creates uniformity in the reported
                information and makes the information more relevant to current
                vehicles.
                 The agency agrees with the essence of Volkswagen's assertion that
                some of the required data fields may no longer be as common on
                contemporary vehicles, and therefore, may not apply to all
                manufacturers. As suggested by Mercedes-Benz, NHTSA has decided to make
                the ``Catalyst Usage,'' ``Distributor Calibration,'' ``Choke
                Calibration,'' and ``Other Calibration'' data fields optional with a
                default value of ``N/A.'' NHTSA does not agree with Mercedes-Benz's
                recommendation that NHTSA provide a better understanding of its
                intention for the information in certain data fields. ``Electric
                Traction Motor, Motor Controller,'' ``Battery Configuration,''
                ``Electrical Charging System,'' and ``Energy Storage Device'' are the
                data fields that characterize the basic powerplant for electric
                vehicles. Basic Engine, along with Carline and Transmission Class, make
                up a model type for light-duty vehicles. Therefore, those five fields
                are used to group vehicles by model type in accordance with EPA
                regulations. Fuel economy performance is calculated by
                Subconfiguration, which is a subset of a model type. As such, those
                five data fields are an integral part of grouping vehicles for fuel
                economy testing purposes in accordance with EPA regulations. NHTSA also
                does not agree with Volkswagen's assertion that the template is
                populated with many fields that do not appear relevant to projecting
                CAFE performance. As previously mentioned, many of the data fields are
                used to arrange vehicles into groups for calculating fuel economy
                performance in accordance with 49 CFR 537.7.
                 Furthermore, NHTSA has re-engineered the template in a few areas to
                include additional supporting data elements used in calculating other
                data fields required by Part 537. These fields may not directly align
                with the existing requirements in Part 537 but are necessary for
                validation purposes. For this reason, NHTSA is also finalizing its
                proposal in the NPRM to remove the optional provisions for reporting
                the data fields for determining the CAFE model type target standards,
                making the information mandatory in the template. Additional changes
                have been made to the template to improve fuel economy calculations.
                NHTSA edited the template to include the calculation procedure for
                alternative-fuel vehicles and corrected the test procedure adjustment
                (TPA) calculation to align the fleet average fuel economy calculation
                methodology with 40 CFR 600.510-12. Several expanded worksheets and
                functional features were also added to the template to improve the
                usability of the templates for manufacturers. These changes include
                modifications such as adding the estimated credits and a minimum
                domestic passenger shortfall calculator as the last fields to the
                ``Summary'' worksheet. Other functional changes include protecting
                users from changing the formatting or data validation in each cell and
                allowing columns to be widened by users.
                (b) Standardized Credit Documents
                 A credit ``[t]rade'' is defined in 49 CFR 536.3 as ``the receipt by
                NHTSA of an instruction from a credit holder to place its credits in
                the account of another credit holder.'' \3174\ ``Traded credits are
                moved from one credit holder to the recipient credit holder within the
                same compliance category for which the credits were originally earned.
                If a credit has been traded to another credit holder and is
                subsequently traded back to the originating manufacturer, it will be
                deemed not to have been traded for compliance purposes.'' \3175\ NHTSA
                does not administer trade negotiations between manufacturers and when a
                [[Page 25218]]
                trade document is received the agreement must be issued jointly by the
                current credit holder and the receiving party.\3176\ NHTSA does not
                settle contractual or payment issues between trading manufacturers.
                ---------------------------------------------------------------------------
                 \3174\ 49 CFR 536.3(b).
                 \3175\ Id.
                 \3176\ See 49 CFR 536.8(a).
                ---------------------------------------------------------------------------
                 NHTSA created its CAFE database to maintain credit accounts for
                manufacturers and to track all credit transactions. A credit account
                consists of a balance of credits in each compliance category and
                vintage held by the holder. While maintaining accurate credit records
                is essential, it has become a challenging task for the agency given the
                recent increase in credit transactions. Manufacturers have requested
                that NHTSA approve trade or transfer requests not only in response to
                end-of-model year shortfalls, but also, during the model year, when
                purchasing credits to bank.
                 To reduce the burden on all parties, encourage compliance, and
                facilitate quicker NHTSA credit transaction approval, the agency
                proposed in the NPRM to add a required template to standardize the
                information parties submit to NHTSA in reporting a credit transaction.
                Presently, manufacturers are inconsistent in submitting the information
                required by 49 CFR 536.8, creating difficulty for NHTSA in processing
                transactions. The template NHTSA proposed is a simple spreadsheet that
                trading parties fill out. When completed, parties will be able to click
                a button on the spreadsheet to generate a credit transaction summary
                and if applicable credit trade confirmation, the latter of which shall
                be signed by both trading entities. The credit trade confirmation
                serves as an acknowledgement that the parties have agreed to trade
                credits. The completed credit trade summary and a PDF copy of the
                signed trade confirmation must be submitted to NHTSA. Using the
                template simplifies CAFE compliance aspects of the credit trading
                process, and helps to ensure that trading parties follow the
                requirements for a credit transaction in 49 CFR 536.8(a).\3177\
                ---------------------------------------------------------------------------
                 \3177\ Submitting a properly completed template and accompanying
                transaction letter will satisfy the trading requirements in 49 CFR
                part 536.
                ---------------------------------------------------------------------------
                 Additionally, the credit trade confirmation includes an
                acknowledgement of the ``error or fraud'' provisions in 49 CFR
                536.8(f)-(g), and the finality provision of 49 CFR 536.8(g). NHTSA
                sought comment on this approach, as well as on any changes to the
                template that may be necessary to facilitate manufacturer credit
                transaction requests. The agency uploaded the proposed template to the
                NHTSA's docket and the CAFE PIC site for manufacturers to download and
                review.
                 Only Global Automakers commented on the proposed credit transaction
                template, and Global Automakers supported adopting a uniform template.
                Global Automakers stated that, in theory, it agrees that a standardized
                template with credit trading information is appropriate, and a similar
                template is already in use for these types of reporting requirements by
                its members that could be integrated into the end of the year EPA final
                report. Global Automakers believes the use of similar templates have
                been well-established, and such a template could be implemented across
                multiple agencies (i.e. NHTSA and EPA) with very little lag time in
                learning.\3178\ No comments were received on the transaction letter
                generated by the template.
                ---------------------------------------------------------------------------
                 \3178\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
                12032.
                ---------------------------------------------------------------------------
                 For the final rule, NHTSA is finalizing the proposed requirements
                for its credit templates to be incorporated into provisions for Part
                536. NHTSA understands that manufacturers may be using similar credit
                reporting templates as part of their current business processes but has
                decided to adopt the template proposed in the NPRM. The NHTSA credit
                templates are an integral part of a long-range technology deployment
                that is already underway and will automate the NHTSA's CAFE database
                and web portal systems. When complete, the systems and portals will
                receive information directly from manufacturers and enable
                manufacturers, independently, to confirm credit trades and receive
                real-time credit balances. For this reason, diverging from the proposed
                templates for the final rule would impose unnecessary costs upon NHTSA.
                In the interest of accommodating the transition by manufacturers from
                other standardized templates, the agency will delay mandatory use of
                the CAFE credit template until January 1, 2021. Manufacturers may
                deviate from the generated language in the NHTSA credit trade
                confirmation by adding additional qualifications but, at a minimum,
                must include the core information generated by the template.
                (c) Credit Transaction Information
                 Credit trading among entities commenced in the CAFE program
                starting in MY 2011.\3179\ To date, NHTSA has received numerous credit
                trades from manufacturers but has only made limited information
                publicly available.\3180\ As discussed earlier, NHTSA maintains an
                online CAFE database with manufacturer and fleetwide compliance
                information that includes year-by-year accounting of credit balances
                for each credit holder. While NHTSA maintains this database, the
                agency's regulations currently state that it does not publish
                information on individual transactions, and NHTSA has not previously
                required trading entities to submit information regarding the
                compensation (whether financial, or other items of value) manufacturers
                receive in exchange for credits.3181 3182 Thus, NHTSA's PIC
                offers sparse information to those looking to determine the value of a
                credit.
                ---------------------------------------------------------------------------
                 \3179\ 49 CFR 536.6(c).
                 \3180\ Manufacturers may generate credits, but non-manufacturers
                may also hold or trade credits. Thus, the word ``entities'' is used
                to refer to those that may be a party to a credit transaction.
                 \3181\ 49 CFR 536.5(e)(1).
                 \3182\ NHTSA understands that not all credits are exchanged for
                monetary compensation. The proposal that NHTSA is adopting in this
                final rule requires entities to report compensation exchanged for
                credits, and is not limited to reporting monetary compensation.
                ---------------------------------------------------------------------------
                 The lack of information regarding credit transactions means
                entities wishing to trade credits have little, if any, information to
                determine the value of the credits they seek to buy or sell. It is
                widely assumed that the civil penalty for noncompliance with CAFE
                standards largely determines the upper value of a credit, because it is
                logical to assume that manufacturers would not purchase credits if it
                cost less to pay civil penalties instead, but it is unknown how other
                factors affect the value. For example, a credit nearing the end of its
                five-model-year lifespan would theoretically be worth less than a
                credit within its full five-model-year lifespan. In the latter case,
                the credit holder would likely value the credit more, as it can be used
                for compliance purposes for a longer period of time.
                 In the interest of facilitating a transparent and efficient credit
                trading market, NHTSA stated in the NPRM that consideration is being
                given to modifying its regulations for credit trade information. NHTSA
                sought comment in the NPRM about the feasibility of requiring more
                information disclosure around trades, including price information,
                noting that neither the public, shareholders, competitors, nor even the
                agencies themselves know the price of credit transactions. More
                specifically, NHTSA proposed requiring trading parties to submit
                information disclosing the identities of the parties to credit trades,
                the number of credits traded, and the amount of compensation exchanged
                for credits. Furthermore, NHTSA proposed that regulations
                [[Page 25219]]
                would also permit the agency to publish information about specific
                transactions on the PIC.
                 NHTSA received comments from Volkswagen, Honda, Fiat Chrysler,
                Toyota, Global Automakers, the Auto Alliance, UCS, and from one private
                citizen, Mr. Jason Schwartz, regarding the scope of available credit
                information. All auto associations and manufacturers requested that
                NHTSA maintain the confidentiality of credit trades and transactions.
                The remaining commenters felt increased transparency would benefit the
                market.
                 Global Automakers, the Auto Alliance, Fiat Chrysler, and Volkswagen
                stated that credit trades are business-to-business, contain internal
                information and can involve both financial and non-financial
                compensation between parties.\3183\ They stated credit transactions
                should be viewed as being similar to other competitive purchase
                agreements, which include non-disclosure terms and strict
                confidentiality with regard to cost and compensation.\3184\ They
                contended that negotiations must remain confidential to protect the
                sensitive business practices for both the buyer and seller, and that
                revealing purchasing terms could result in a competitive disadvantage
                for both.\3185\ Further, it was stated that certain transactions may
                not happen if they are publicized for fear of public criticism, making
                the program less efficient.\3186\
                ---------------------------------------------------------------------------
                 \3183\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
                12032; Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073; Fiat
                Chrysler, Detailed Comments, NHTSA-2018-0067-11943; Volkswagen,
                Detailed Comments, NHTSA-2017-0069-0583.
                 \3184\ See, e.g., Auto Alliance, Detailed Comments, NHTSA-2018-
                0067-12073.
                 \3185\ See, e.g., Auto Alliance, Detailed Comments, NHTSA-2018-
                0067-12073.
                 \3186\ See, e.g., Fiat Chrysler, Detailed Comments, NHTSA-2018-
                0067-11943.
                ---------------------------------------------------------------------------
                 Honda added that disclosing trading terms may not be as simple as a
                spot purchase at a given price.\3187\ Honda explained that it has
                undertaken a number of transactions for both CAFE and CO2
                credits, and there has been a range of complexity in these transactions
                due to numerous factors that are reflective of the marketplace, such as
                the volume of credits, compliance category, credit expiration date, a
                seller's compliance strategy, and even the CAFE penalty rate in effect
                at that time.\3188\ In addition, Honda stated that automakers have a
                range of partnerships and cooperative agreements with their own
                competitors.\3189\ Honda commented that credit transactions can be an
                offshoot of these broader relationships, and difficult to price
                separately and independently.\3190\ Thus, Honda believes there may not
                be a reasonable, or even meaningful, presentation of ``market''
                information in a transaction ``price.'' \3191\ Finally, Honda concluded
                by stating that information on pricing terms and business partner
                pairings is highly competitive and, if made public, could divulge to
                competitors a buyer's and/or seller's future compliance strategy.\3192\
                For these reasons, Honda believes it is appropriate to maintain the
                confidentiality of trade terms, pricing information, and of trading
                partner identification.\3193\
                ---------------------------------------------------------------------------
                 \3187\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                 \3188\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                 \3189\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                 \3190\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                 \3191\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                 \3192\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                 \3193\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                ---------------------------------------------------------------------------
                 Fiat Chrysler stated that revealing credit transaction information
                would reveal highly confidential business information.\3194\ It stated
                that credit transaction information may reveal the technology that is
                most valued by a company and the value of putting certain technology
                into a vehicle.\3195\ It believed that credit trades are complex
                business transactions made at arm's length.\3196\ As such, they may
                include monetary and non-monetary compensation, non-disclosure
                provisions, and other sensitive terms.\3197\ Fiat Chrysler commented
                that publicizing such sensitive information could stifle the credit
                market and potentially result in uncompetitive outcomes, and could also
                decrease the efficiency in the credit trading marketplace.\3198\ Fiat
                Chrysler further stated that the NPRM's justifications for requiring
                the disclosure of credit transaction information is unfounded and the
                government has no need of this information in the regular course of
                doing business.\3199\
                ---------------------------------------------------------------------------
                 \3194\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3195\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3196\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3197\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3198\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3199\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                ---------------------------------------------------------------------------
                 The Auto Alliance, Honda, Toyota, and Volkswagen argued against
                NHTSA publishing credit movements each model year on its PIC. They
                stated that detailed credit banks by account holder are available to
                the public or entities wishing to engage in the credit market and that
                information is already sufficient.\3200\ Global Automakers further
                contended that the agencies know which companies are trading and how
                those credits are being used, which is all that should be required for
                administering the program.\3201\ The Auto Alliance argued that in
                private markets, trades and prices often are not made public; this
                privacy does not mean that the markets operate any less effectively,
                nor that the public at large does not benefit from the transactions
                that lower costs for all parties.\3202\
                ---------------------------------------------------------------------------
                 \3200\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Honda, Detailed Comments, NHTSA-2018-0067-11818; Toyota, Detailed
                Comments, NHTSA-2018-0067-12150; Volkswagen, Detailed Comments,
                NHTSA-2017-0069-0583.
                 \3201\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
                12032.
                 \3202\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 Volkswagen further commented that revealing confidential purchase
                terms has no precedent in the automotive industry. Volkswagen's
                position is that it does not disclose contract pricing for purchasing
                fuel saving technologies from suppliers, such as for turbochargers or
                battery packs. Therefore, Volkswagen does not believe it is appropriate
                to disclose the purchase price for CAFE credits.\3203\
                ---------------------------------------------------------------------------
                 \3203\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
                ---------------------------------------------------------------------------
                 Opposite views from those expressed by automobile manufacturers
                were received in the comments from UCS and Jason Schwartz. Both
                commenters strongly supported an increase in information regarding
                credit trading in the CAFE program.\3204\ They argued that more
                information will allow manufacturers to make better informed decisions
                and lead to greater industry efficiency in general.\3205\ UCS added
                that while the PIC does have some information, it is difficult to
                discern how the manufacturers are dividing credits to offset
                shortfalls.\3206\ It requested NHTSA disclose at least as much
                information as EPA provides from its program, if not providing more
                information on transaction price and
                [[Page 25220]]
                compliance category.\3207\ Jason Schwartz had similar arguments for
                more transparency. Mr. Schwartz added that the agencies can assume that
                credits may be traded at prices similar to the civil penalty rate for
                noncompliance under the CAFE standards, but not knowing the actual
                prices greatly complicates the agencies' estimations of the costs of
                complying with the standards.\3208\ Schwartz used several examples to
                explain and justify the need for making data on credit transactions,
                prices, and holdings publicly available to help the agency and the
                public assess the efficacy of the program.\3209\ He also explained that
                such information will enable the smooth operation of the credit market
                by enabling credit buyers to better evaluate the value of credits and
                placing all players on equal informational footing which facilitates
                price discovery, and assists buyers and sellers in reaching
                terms.\3210\ He added that regulators should require greater
                transparency to facilitate oversight.\3211\ He asserted his belief that
                greater transparency in tracking transactions and credits helps
                regulators detect fraud, manipulation, market power, abuse, and to
                enforce compliance.\3212\
                ---------------------------------------------------------------------------
                 \3204\ UCS, Detailed Comments, NHTSA-2018-0067-12039; Jason
                Schwartz, Detailed Comments, NHTSA-2018-0067-12162.
                 \3205\ See, e.g., UCS, Detailed Comments, NHTSA-2018-0067-12039.
                 \3206\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
                 \3207\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
                 \3208\ Jason Schwartz, Detailed Comments, NHTSA-2018-0067-12162.
                 \3209\ Jason Schwartz, Detailed Comments, NHTSA-2018-0067-12162.
                 \3210\ Jason Schwartz, Detailed Comments, NHTSA-2018-0067-12162.
                 \3211\ Jason Schwartz, Detailed Comments, NHTSA-2018-0067-12162.
                 \3212\ Jason Schwartz, Detailed Comments, NHTSA-2018-0067-12162.
                ---------------------------------------------------------------------------
                 In response to these comments, NHTSA has decided not to share
                detailed information on credit transactions or the cost of individual
                credit transactions with the public. NHTSA agrees with manufacturers
                that revealing confidential purchase terms could result in a
                competitive disadvantage for both credit buyers and sellers, as well as
                harm to companies revealing highly confidential business materials.
                However, NHTSA believes that greater government oversight is needed
                over the CAFE credit market. NHTSA needs to understand more information
                surrounding trades, including costing information. As Honda recognized
                in its comments, NHTSA needs to understand the full range of complexity
                in transactions, monetary and non-monetary, in addition to the range of
                partnerships and cooperative agreements between credit account
                holders--which may impact the price of credit trades.\3213\ NHTSA also
                believes, as mentioned by commenters, that disclosure of information
                concerning credit trades is important for facilitating government
                oversight for protecting against fraud, manipulation, market power, and
                abuse which may occur in the credit market.
                ---------------------------------------------------------------------------
                 \3213\ Honda, Detailed Comments, NHTSA-2018-0067-11819.
                ---------------------------------------------------------------------------
                 NHTSA is adopting new reporting provisions in this final rule.
                Starting January 1, 2021, manufacturers will be required to submit all
                credit trade contracts, including costing and transactional
                information, to the agency. This information may be submitted
                confidentially, in accordance with 49 CFR part 512.\3214\ NHTSA will
                use this information to determine the true cost of compliance for all
                manufacturers. This information will allow NHTSA to assess better the
                impact of its regulations on the industry, and provide more insightful
                information to use in developing future rulemakings. This confidential
                information will be held by secure electronic means in NHTSA's database
                systems. As for public information, NHTSA will include more information
                on the PIC on aggregated credit transactions, such as the combined
                flexibilities all manufacturers used for compliance as shown in Figure
                IX-6, or information comparable to the credit information EPA makes
                available to the public. In the future, NHTSA will consider what
                information, if any, can be meaningfully shared with the public on
                credit transactional details or costs, while accounting for the
                concerns raised by the automotive industry.
                ---------------------------------------------------------------------------
                 \3214\ See also 49 U.S.C. 32910(c).
                ---------------------------------------------------------------------------
                (d) Precision of the CAFE Credit Adjustment Factor
                 EPCA, as amended by EISA, required the Secretary of Transportation
                to establish an adjustment factor to ensure total oil savings are
                preserved when manufacturers trade credits.\3215\ The adjustment factor
                applies to credits traded between manufacturers and to credits
                transferred across a manufacturer's compliance fleets.
                ---------------------------------------------------------------------------
                 \3215\ 49 U.S.C. 32903(f)(1).
                ---------------------------------------------------------------------------
                 In establishing the adjustment factor, NHTSA did not specify the
                exact precision of the output of the equation in 49 CFR 536.4(b).
                NHTSA's standard practice has been round to the nearest four decimal
                places (e.g., 0.0001) for the adjustment factor. However, in the
                absence of a regulatory requirement, many manufacturers have contacted
                NHTSA for guidance, and NHTSA has had to correct several credit
                transaction requests. In some instances, manufacturers have had to
                revise signed credit trade documents and submit additional trade
                agreements to properly address credit shortfalls.
                 NHTSA proposed in the NPRM to add requirements to 49 CFR 536.4
                specifying the precision of the adjustment factor by rounding to four
                decimal places (e.g., 0.0001). NHTSA has also included equations for
                the adjustment factor in its proposed credit transaction report
                template, mentioned above, with the same level of precision. NHTSA
                sought comment on this approach but received no comments, and therefore
                is finalizing this approach in this final rule.
                (3) NHTSA Then Analyzes EPA-Certified CAFE Values for Compliance
                 After manufacturers complete certification testing and submit their
                final compliance values to EPA, EPA verifies the data and issues final
                CAFE reports to manufacturers and NHTSA. NHTSA then evaluates whether
                the manufacturers' compliance categories (i.e., domestic passenger car,
                imported passenger car, and light truck fleets) meet the applicable
                CAFE standards. NHTSA uses EPA-verified data to compare fleet average
                standards with actual fleet performance values in each compliance
                category. Each vehicle a manufacturer produces has a fuel economy
                target based on its footprint (footprint curves are discussed above in
                Section II.C), and each compliance category has a CAFE standard
                measured in miles per gallon (mpg). The manufacturer's fleet average
                CAFE standard is calculated based on the fuel economy target value and
                production volume of each vehicle model. The CAFE performance is
                calculated based on the compliance value and production volume of each
                vehicle model. A manufacturer complies with the CAFE standard if its
                fleet average performance is greater than or equal to its required
                standard, or if it is able to use available compliance flexibilities,
                described below in Section IX.C.2.c. to resolve any shortfall.
                 If the average fuel economy level of the vehicles in a compliance
                category falls below the applicable fuel economy standard, NHTSA
                provides written notification to the manufacturer that it has not met
                that standard. The manufacturer is then required to confirm the
                shortfall and either submit a plan indicating how it will allocate
                existing credits, or if it does not have sufficient credits available
                in that fleet, how it will earn, transfer, and/or acquire credits, or
                pay the appropriate civil penalty. The manufacturer must submit a
                credit allocation plan or payment within 60 days of receiving agency
                notification.
                [[Page 25221]]
                 NHTSA approves a credit allocation plan unless it finds the
                proposed credits are unavailable or that it is unlikely that the plan
                will result in the manufacturer earning sufficient credits to offset
                the projected shortfall. If a plan is approved, NHTSA revises the
                manufacturer's credit account accordingly. If a plan is rejected, NHTSA
                notifies the manufacturer and requests a revised plan or payment of the
                appropriate civil penalty. Similarly, if the manufacturer is delinquent
                in submitting a response within 60 days, NHTSA takes action to collect
                a civil penalty. If NHTSA receives and approves a manufacturer's plan
                to carryback future earned credits within the following three years in
                order to comply with current regulatory obligations, NHTSA will defer
                levying civil penalties for noncompliance until the date(s) when the
                manufacturer's approved plan indicates that the credits will be earned
                or acquired to achieve compliance. If the manufacturer fails to acquire
                or earn sufficient credits by the plan dates, NHTSA will initiate
                noncompliance proceedings to collect civil penalties.\3216\
                ---------------------------------------------------------------------------
                 \3216\ See generally 49 CFR part 536.
                ---------------------------------------------------------------------------
                (4) Civil Penalties for Noncompliance
                 In the event that a manufacturer does not comply with a CAFE
                standard, EPCA provides that the manufacturer is potentially liable for
                a civil penalty.\3217\ The manufacturer determines whether to use
                available credits to reduce or offset its potential penalty.\3218\ This
                penalty rate is $5.50 for each tenth of a mpg that a manufacturer's
                average fuel economy falls short of the standard for a given model year
                multiplied by the total volume of those vehicles in the affected
                compliance category manufactured for that model year.\3219\ A person
                (or manufacturer) that violates 49 U.S.C. 32911(a), including general
                CAFE violations other than those for failing to comply with CAFE
                standards (i.e., fuel economy labeling violations), is also liable to
                the United States Government for a civil penalty of not more than
                $42,530 for each violation. A separate violation occurs for each day
                the violation continues. All penalties are paid to the U.S. Treasury
                and not to NHTSA.\3220\
                ---------------------------------------------------------------------------
                 \3217\ 49 U.S.C. 32911-12.
                 \3218\ See 49 U.S.C. 32912.
                 \3219\ NHTSA finalized a retaining the $5.50 civil penalty rate
                in an April 2018 NPRM. See 83 FR 13904 (Apr. 2, 2018).
                 \3220\ 49 U.S.C. 32912(e) allows for fiscal year 2008 and each
                fiscal year thereafter, the total amount deposited in the general
                fund of the Treasury during the preceding fiscal year from fines,
                penalties, and other funds obtained through enforcement actions
                conducted pursuant to EISA and EPCA (including funds obtained under
                consent decrees), the Secretary of the Treasury, subject to the
                availability of appropriations, shall: (1) transfer 50 percent of
                such total amount to the account providing appropriations to the
                Secretary of Transportation for the administration of this chapter,
                which shall be used by the Secretary to support rulemaking under
                this chapter; and (2) transfer 50 percent of such total amount to
                the account providing appropriations to the Secretary of
                Transportation for the administration of this chapter, which shall
                be used by the Secretary to carry out a program to make grants to
                manufacturers for retooling, reequipping, or expanding existing
                manufacturing facilities in the United States to produce advanced
                technology vehicles and components.
                Potential Civil Penalty = $5.50 x (Avg. FE Performance-Avg. FE
                ---------------------------------------------------------------------------
                Standard) x 10 x Total Production
                 Since the inception of the CAFE program, the U.S. Treasury has
                collected a total of $1,049,355,116 in CAFE civil penalty payments.
                Generally, import manufacturers have paid significantly more in civil
                penalties than domestic manufacturers, with the majority of payments
                made by import manufacturers for passenger cars and not light trucks.
                Over the total program lifetime, import manufacturers paid a total of
                $1,048,896,676 in CAFE penalties while domestic manufacturers paid a
                total of $458,440.\3221\
                ---------------------------------------------------------------------------
                 \3221\ These totals include penalties associated with all fleets
                for these manufacturers. For example, the total penalties paid by
                import manufacturers includes penalties associated with shortfalls
                in those manufacturers' domestic passenger car fleets.
                ---------------------------------------------------------------------------
                 Prior to the CAFE credit trade and transfer program, several
                manufacturers opted to pay civil penalties instead of complying with
                CAFE standards. Since NHTSA introduced trading and transferring,
                manufacturers have largely traded or transferred credits to achieve
                compliance, rather than paying civil penalties for noncompliance. NHTSA
                therefore assumes that buying and selling credits is a more cost-
                effective strategy for manufacturers than paying civil penalties, in
                part, because it seems logical that the price of a credit is directly
                related to the civil penalty rate and decreases as a credit's life
                diminishes.\3222\ Prior to trading and transferring, on average,
                manufacturers paid $28,073,281.93 in civil penalty payments annually (a
                total of $814,125,176 from MYs 1982 to 2010). Since trading and
                transferring began, manufacturers now pay an average of $26,136,660
                each model year. The agency notes that six manufacturers have paid
                civil penalties since 2011 totaling $235,229,940; Fiat Chrysler paid a
                civil penalty in MY 2016 equal to $77,268,720.50 and in MY 2017 equal
                to $79,376,643.50 for for failing to meet the minimum domestic
                passenger car standards for those MYs. NHTSA expects that, over the
                next several years, manufacturers will face challenges in avoiding
                paying further civil penalties as standards increase in stringency.
                Compared to the current $5.50 CAFE civil penalty rate, a rate of $14
                would cause manufacturers that do not comply with CAFE to pay
                significantly higher civil penalties, potentially in the magnitude of
                hundreds of millions of dollars annually beyond current projections.
                Additionally, although NHTSA has not historically been privy to the
                monetary terms of credit trades, NHTSA expects that the price of
                credits would increase in line with any increase in the CAFE civil
                penalty rate.
                ---------------------------------------------------------------------------
                 \3222\ See 49 CFR 536.4 for NHTSA's regulations regarding CAFE
                credits.
                ---------------------------------------------------------------------------
                b) What Exemptions and Exclusions Does NHTSA Allow?
                (1) Emergency and Law Enforcement Vehicles
                 Under EPCA, manufacturers are allowed to exclude emergency
                vehicles, which include law enforcement vehicles, from their CAFE
                fleet.\3223\ All manufacturers that produce emergency vehicles have
                historically done so. NHTSA did not propose any changes to this
                exclusion and therefore is retaining the provision without change for
                the final rule.
                ---------------------------------------------------------------------------
                 \3223\ 49 U.S.C. 32902(e).
                ---------------------------------------------------------------------------
                (2) Small Volume Manufacturers
                 Per 49 U.S.C. 32902(d), NHTSA established requirements for exempted
                small volume manufacturers in 49 CFR part 525, ``Exemptions from
                Average Fuel Economy Standards.'' The small volume manufacturer
                exemption is available for any manufacturer whose projected or actual
                combined sales (whether in the U.S. or not) are fewer than 10,000
                passenger automobiles in the model year two years before the model year
                for which the manufacturer seeks an exemption.\3224\ The manufacturer
                must submit a petition with information stating that the applicable
                CAFE standard is more stringent than the maximum feasible average fuel
                economy level that the manufacturer can achieve.\3225\ NHTSA must then
                issue by Federal Register notice, a proposed decision granting or
                denying the petition and inviting public comment.\3226\ If the agency
                proposed to grant the petition, the notice includes an alternative
                average fuel economy standard for the passenger automobiles
                manufactured by the manufacturer.\3227\ After conclusion of the public
                comment period, the agency publishes a final
                [[Page 25222]]
                decision in the Federal Register.\3228\ If the agency grants the
                petition, it establishes an alternative standard, which is the maximum
                feasible average fuel economy level for the manufacturers to which the
                alternative standard applies.\3229\ NHTSA did not propose and is not
                making any changes to the small volume manufacturer provision or
                alternative standards regulations in this rulemaking.
                ---------------------------------------------------------------------------
                 \3224\ 49 CFR 525.5.
                 \3225\ 49 CFR 525.7(h).
                 \3226\ 49 CFR 525.8(c).
                 \3227\ Id.
                 \3228\ 49 CFR 525.8(e).
                 \3229\ 49 U.S.C. 32902(d)(2); 49 CFR 525.8(e).
                ---------------------------------------------------------------------------
                c) What Compliance Flexibilities and Incentives Are Currently Available
                Under the CAFE Program and How Do Manufacturers Use Them?
                 There are several compliance flexibilities and incentives that
                manufacturers can use to achieve compliance with CAFE standards beyond
                applying fuel economy-improving technologies. Some compliance
                flexibilities and incentives are statutorily mandated by Congress
                through EPCA and EISA. These specifically include program credits
                generated from overcompliance, including the ability to carry-forward,
                carryback, trade and transfer credits, and special fuel economy
                calculations for dual- and alternative-fueled vehicles (discussed in
                turn, below). However, 49 U.S.C. 32902(h) expressly prohibits NHTSA
                from considering the availability of statutorily established credits
                (either for building dual- or alternative-fueled vehicles or from
                accumulated transfers or traders) in setting the level of the
                standards. Thus, NHTSA may not raise CAFE standards because
                manufacturers have enough credits to meet higher standards, or because
                alternative fuel vehicles (including electric vehicles) are available
                to help manufacturers achieve compliance. This is an important
                difference from EPA's authority under the CAA, which does not contain
                such a restriction, and which flexibility EPA has utilized in the past
                in determining appropriate levels of stringency for its program.
                 Generating, trading, transferring, and applying CAFE credits is
                governed by statute.\3230\ Program credits are generated when a vehicle
                manufacturer's fleet over-complies with its standard for a given model
                year, meaning its vehicle fleet achieved a higher corporate average
                fuel economy value than the amount required by the CAFE program for
                that fleet in that model year. Conversely, if the fleet average CAFE
                level does not meet the standard, the fleet would incur debits (also
                referred to as a shortfall). A manufacturer whose fleet generates a
                credit shortfall in a given model year can resolve its shortfall using
                any one or combination of several credits flexibilities, including
                credit carryback, credit carry-forward, credit transfers, and credit
                trades.
                ---------------------------------------------------------------------------
                 \3230\ 49 U.S.C. 32903.
                ---------------------------------------------------------------------------
                 NHTSA also has promulgated compliance flexibilities and incentives
                consistent with EPCA's provisions regarding calculation of fuel economy
                levels for individual vehicles and for fleets.\3231\ These compliance
                flexibilities and incentives, which were first adopted in the 2012 rule
                for MYs 2017 and later, include A/C efficiency improvement and off-
                cycle adjustments, and adjustments for advanced technologies in full-
                size pickup trucks, including adjustments for mild and strong hybrid
                electric full-size pickup trucks and performance-based incentives in
                full-size pickup trucks. The fuel consumption improvement benefits of
                these technologies measured by various testing methods can be used by
                manufacturers to increase the CAFE performance of their fleets. As
                discussed below, the adjustments for advanced technologies in full-size
                pickup trucks will no longer be available beginning in MY 2022.
                ---------------------------------------------------------------------------
                 \3231\ 49 U.S.C. 32904.
                ---------------------------------------------------------------------------
                 Under NHTSA regulations, credit holders (including, but not limited
                to manufacturers) have credit accounts with NHTSA where they can, as
                outlined above, hold credits, and use them to achieve compliance with
                CAFE standards, by carrying forward, carrying back, or transferring
                credits across compliance categories. Manufacturers with excess credits
                in their accounts can also trade credits to other manufacturers, who
                may use those credits to resolve a shortfall currently or in a future
                model year. A credit may also be cancelled before its expiration date
                if the credit holder so chooses. Traded and transferred credits are
                subject to an ``adjustment factor'' to ensure total oil savings are
                preserved.\3232\ Credits earned before MY 2011 may not be traded or
                transferred.\3233\
                ---------------------------------------------------------------------------
                 \3232\ 49 CFR 536.4(c).
                 \3233\ 49 CFR 536.6(c).
                ---------------------------------------------------------------------------
                 Credit ``carryback'' means that manufacturers are able to use
                credits to offset a deficit that had accrued in a prior model year,
                while credit ``carry-forward'' means that manufacturers can bank
                credits and use them towards compliance in future model years. EPCA, as
                amended by EISA allows manufacturers to carryback credits for up to
                three model years, and to carry-forward credits for up to five model
                years.\3234\ Credits expire the model year after which the credits may
                no longer be used to achieve compliance with fuel economy
                regulations.\3235\ Manufacturers seeking to use carryback credits must
                have an approved carryback plan from NHTSA demonstrating their ability
                to earn sufficient credits in future MYs that can be carried back to
                resolve the current MY's credit shortfall.
                ---------------------------------------------------------------------------
                 \3234\ 49 U.S.C. 32903(a).
                 \3235\ 49 CFR 536.3(b).
                ---------------------------------------------------------------------------
                 Credit ``trading'' refers to the ability of manufacturers or
                persons to sell credits to, or purchase credits from, one another. EISA
                gave NHTSA discretion to establish by regulation a CAFE credit trading
                program, to allow credits to be traded between vehicle manufacturers,
                now codified at 49 CFR part 536.\3236\ EISA prohibited manufacturers
                from using traded credits to meet the minimum domestic passenger car
                CAFE standard.\3237\
                ---------------------------------------------------------------------------
                 \3236\ 49 U.S.C. 32903(f).
                 \3237\ 49 U.S.C. 32903(f)(2).
                ---------------------------------------------------------------------------
                 As mentioned previously, the agencies sought comments in the NPRM
                on whether and how each agency's existing flexibilities and incentives
                might be amended, revised, or deleted to avoid the inefficiencies and
                market distortions as discussed earlier. NHTSA was concerned with the
                potential for unintended consequences. Specifically, comments were
                sought on the appropriate level of compliance flexibilities, including
                credit trading, in a program that is correctly designed to follow
                statutory direction to create maximum feasible fuel economy standards.
                Given that the credit trading program is discretionary under EISA,
                NHTSA also sought comments on whether the credit trading provisions in
                49 CFR part 536 should cease to apply beginning in MY 2022. Comments
                were sought on whether to allow all incentive-based adjustments, except
                those that are mandated by statute, to expire, in addition to other
                possible simplifications to reduce market distortion, improve program
                transparency and accountability, and improve overall performance of the
                compliance programs.
                 The comments received from the public and NHTSA's responses to
                those comments are discussed below. A summary of all the flexibilities
                and incentives, and information on whether they were either retained or
                modified for the final rule, is presented in Table IX-1 through Table
                IX-4.
                [[Page 25223]]
                (1) Credit Carry-Forward and Back
                 Under the CAFE program, when the average fuel economy of a
                compliance fleet manufactured in a particular model year exceeds its
                applicable average fuel economy standard, the manufacturer earns
                credits.\3238\ The credits may be applied to: (1) Any of the 3
                consecutive model years immediately before the model year for which the
                credits are earned; and (2) any of the 5 consecutive model years
                immediately after the model year for which the credits are earned. For
                example, a credit earned for exceeding model year 2017 standards will
                be usable for compliance purposes through and including the 2022
                compliance model year. NHTSA did not seek comment on or propose changes
                to any of the aspects of its lifespan for CAFE credits because of the
                existing statutory limitation set forth by Congress. The public offered
                no comments on such flexibilities under the CAFE program.
                ---------------------------------------------------------------------------
                 \3238\ 49 U.S.C. 32903 and 49 CFR 536.
                ---------------------------------------------------------------------------
                (2) Credit Trading
                 All commenters responding to the NPRM on this issue favored
                retaining the existing CAFE credit trading program. Comments on credit
                trading were received from Volkswagen, Honda, General Motors, CARB,
                BorgWarner, Jaguar Land Rover, Fiat Chrysler, Global Automakers, the
                Auto Alliance, the Institute for Policy Integrity, Toyota, and academic
                commenters, Jeremy Michalek and Jason Schwartz. No comments were
                received supporting the idea of changing the existing credit trading
                program.
                 In general, manufacturers' comments centered around problems in
                predicting whether consumers will purchase the fuel efficient vehicles
                necessary for manufacturers to meet their compliance obligations. They
                stated that continuing the credit trading program allows manufacturers
                to address uncertainty in the market better.\3239\ The Auto Alliance,
                Volkswagen, Fiat Chrysler, and Honda commented that credit
                flexibilities allow manufacturers to comply with the program even when
                faced with market uncertainties.\3240\ Honda stated that credit trading
                allows the government to set reasonable standards without fear of
                having to cater to the least-capable manufacturer.\3241\ Jaguar Land
                Rover stated the removal of NHTSA's credit trading programs would
                increase and intensify the dis-harmonization between the CO2
                and CAFE programs.\3242\
                ---------------------------------------------------------------------------
                 \3239\ See, e.g., Fiat Chrysler, Detailed Comments, NHTSA-2018-
                0067-11943.
                 \3240\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Volkswagen, Detailed Comments, NHTSA-2017-0069-0583-22; Fiat
                Chrysler, Detailed Comments, NHTSA-2018-0067-11943; Honda, Detailed
                Comments, NHTSA-2018-0067-11818.
                 \3241\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                 \3242\ Jaguar Land Rover, Detailed Comments, NHTSA-2018-0067-
                11916-9.
                ---------------------------------------------------------------------------
                 Global Automakers, Fiat Chrysler, Jason Schwartz, and Jeremy
                Michalek each commented that the credit trading program allows for a
                more efficient compliance process given that more fuel-efficient
                manufacturers can sell their credits to manufacturers who fall
                short.\3243\ These commenters and BorgWarner stated that the program
                lowers the overall cost of reducing fuel consumption.\3244\ Likewise,
                Jaguar Land Rover, Fiat Chrysler, and General Motors argued compliance
                flexibilities, like trading, increase the ability to achieve higher
                fuel economy and reduced CO2 emissions. They found that the
                credit trading flexibility allows them to invest more money in
                technologies that will lead to future increases in their fuel
                economy.\3245\ Similarly, CARB argued credit flexibilities have been
                shown to be successful in reducing emissions and spurring innovation.
                It saw no reason to remove a successful program.\3246\
                ---------------------------------------------------------------------------
                 \3243\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
                12032; Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943;
                Jason Schwartz, Detailed Comments, NHTSA-2018-0067-12162; Jeremy
                Michalek, Detailed Comments, NHTSA-2018-0067-11903.
                 \3244\ BorgWarner, Detailed Comments, NHTSA-2018-0067-11895.
                 \3245\ Jaguar Land Rover, Detailed Comments, NHTSA-2018-0067-
                11916; Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943;
                General Motors, Detailed Comments, NHTSA-2018-0067-11858.
                 \3246\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 Fiat Chrysler stated that credit trading allows manufacturers to
                provide more choices for consumers since manufacturers are not required
                to meet the standard exactly, but rather, they can purchase traded
                credits and then provide vehicles the public is demanding while still
                complying with fleet average standards.\3247\ They stated that this
                leads to the overall compliance of the U.S. fleet while allowing for
                more consumer choices. They further added that if the program is
                removed, manufacturers that currently generate credits from their fuel-
                efficient fleet may find it more profitable to begin producing less
                fuel-efficient vehicles, perhaps even halting the current improvements
                in fuel efficiency across the industry.\3248\
                ---------------------------------------------------------------------------
                 \3247\ General Motors, Detailed Comments, NHTSA-2018-0067-11943.
                 \3248\ General Motors, Detailed Comments, NHTSA-2018-0067-11943.
                ---------------------------------------------------------------------------
                 Honda commented that regulatory flexibilities, such as credit
                trading, built into the CO2 and CAFE programs have become
                critical elements to the programs' success, especially in the face of
                product cadences with uneven sales that do not always match compliance
                obligations.\3249\ General Motors stated its belief that program
                flexibilities will continue to play an increasingly important role in
                reducing CO2 emissions and increasing fuel economy through
                technologies and innovations.\3250\ CARB stated that existing
                flexibilities create consistency in compliance planning for automakers
                for model years in the existing program.\3251\ Fiat Chrysler added that
                each of the CAFE and CO2 programmatic tools and
                flexibilities should be retained, improved and strengthened. Fiat
                Chrysler opined that this is a chance for the agencies to make better
                policies that work more efficiently and as intended, and cautioned that
                eliminating them now could have the serious negative impact of making
                the standards more stringent and costlier for manufacturers.\3252\
                ---------------------------------------------------------------------------
                 \3249\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                 \3250\ General Motors, Detailed Comments, NHTSA-2018-0067-11858.
                 \3251\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
                 \3252\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                ---------------------------------------------------------------------------
                 NHTSA is not making changes to its credit trading provisions in the
                final rule. NHTSA sought comments on removing the optional credit
                trading program to explore public views on market distortions or
                windfalls that occur as a result of the credit trading program.
                However, commenters consistently opined that removing existing
                flexibilities might result in manufacturers not building certain types
                of vehicles. This could adversely impact compliance plans over multiple
                model years. NHTSA concurs with those views, and since this final rule
                adopts CAFE standards that continuously increase through MY 2026,
                understands the importance of allowing for credit trading to provide
                additional means of achieving compliance for manufacturers who face
                varying degrees of difficulty in achieving the standards the agencies
                are finalizing today. With increasing standards, credit trading
                flexibilities help to compensate for the possibility of an uneven sales
                mix of vehicle types and to aid with compliance planning.
                [[Page 25224]]
                Final sales volumes, as presented earlier, show a shift over the past
                several years in consumers purchasing more small SUVs subject to
                passenger car standards, and these vehicles are less fuel efficient
                than the compact and mid-sized passenger cars that previously dominated
                the market. The need to ensure consumer choice is adequately considered
                drives the need for NHTSA to provide credit trading flexibility to
                manufacturers. For example, even with increasing standards, a
                manufacturer could continue to sell certain types of vehicles with
                lower mpg performance over a longer period of time to satisfy its
                consumers by purchasing credits or carrying credits back from future
                model years to address the mpg fleet shortages caused by these
                vehicles, before ultimately having to introduce more fuel-efficient
                technologies. NHTSA believes that these types of scenarios are
                consistent with the purpose of the CAFE credit program, as adopted by
                Congress.
                (3) Credit Transferring
                 Credit ``transfer'' means the ability of manufacturers to move
                credits from their passenger car fleet to their light truck fleet, or
                vice versa. As part of the EISA amendments to EPCA, NHTSA was required
                to establish by regulation a CAFE credit transferring program, now
                codified at 49 CFR part 536, to allow a manufacturer to transfer
                credits between its car and truck fleets to achieve compliance with the
                standards.\3253\ For example, credits earned by overcompliance with a
                manufacturer's car fleet average standard may be used to offset debits
                incurred because of that manufacturer's failed to meet the truck fleet
                average standard in a given year. However, EISA imposed a cap on the
                amount by which a manufacturer could raise its CAFE performance through
                transferred credits: 1 mpg for MYs 2011-2013; 1.5 mpg for MYs 2014-
                2017; and 2 mpg for MYs 2018 and beyond.\3254\ These statutory limits
                will continue to apply to the determination of compliance with CAFE
                standards. EISA also prohibits the use of transferred credits to meet
                the minimum domestic passenger car fleet CAFE standard.\3255\
                ---------------------------------------------------------------------------
                 \3253\ See 49 U.S.C. 32903(g)(1).
                 \3254\ 49 U.S.C. 32903(g)(3).
                 \3255\ 49 U.S.C. 32903(g)(4).
                ---------------------------------------------------------------------------
                 In the NPRM, NHTSA responded to the 2016 petition for rulemaking
                from the Auto Alliance and Global Automakers (Alliance/Global or
                Petitioners) asking to amend the regulatory definition of ``transfer''
                as it pertains to compliance flexibilities.\3256\ In particular,
                Alliance/Global requested that NHTSA add text to the definition of
                ``transfer'' stating that the statutory transfer cap in 49 U.S.C.
                32903(g)(3) applies when the credits are transferred. Alliance/Global
                assert that adding this text to the definition is consistent with
                NHTSA's prior position on this issue in the MYs 2012-2016 final rule,
                in which NHTSA stated:
                ---------------------------------------------------------------------------
                 \3256\ Auto Alliance and Global Automakers Petition for Direct
                Final Rule with Regard to Various Aspects of the Corporate Average
                Fuel Economy Program and the Greenhouse Gas Program (June 20, 2016)
                at 13, available at https://www.epa.gov/sites/production/files/2016-09/documents/petition_to_epa_from_auto_alliance_and_global_automakers.pdf
                [hereinafter Alliance/Global Petition].
                 NHTSA interprets EISA not to prohibit the banking of transferred
                credits for use in later model years. Thus, NHTSA believes that the
                language of EISA may be read to allow manufacturers to transfer
                credits from one fleet that has an excess number of credits, within
                the limits specified, to another fleet that may also have excess
                credits instead of transferring only to a fleet that has a credit
                shortfall. This would mean that a manufacturer could transfer a
                certain number of credits each year and bank them, and then the
                credits could be carried forward or back `without limit' later if
                and when a shortfall ever occurred in that same fleet.\3257\
                ---------------------------------------------------------------------------
                 \3257\ 75 FR 25666 (May 7, 2010).
                 NHTSA clarified in the NPRM, based upon a previous interpretation,
                that the transfer cap from EISA does not limit how many credits may be
                transferred in a given model year, but it does limit the application of
                transferred credits to a compliance category in a model year.\3258\ The
                interpretation concludes by stating that, ``Thus, manufacturers may
                transfer as many credits into a compliance category as they wish, but
                transferred credits may not increase a manufacturer's CAFE level beyond
                the statutory limits.'' \3259\
                ---------------------------------------------------------------------------
                 \3258\ See, letter from O. Kevin Vincent, Chief Counsel, NHTSA
                to Tom Stricker, Toyota (July 5, 2011), available at https://isearch.nhtsa.gov/files/10-004142%20-%20Toyota%20CAFE%20credit%20transfer%20banking%20-%205%20Jul%2011%20final%20for%20signature.htm (last accessed Apr.
                18, 2018).
                 \3259\ Id.
                ---------------------------------------------------------------------------
                 NHTSA maintains its views that the transfer caps in 49 U.S.C.
                32903(g)(3) are properly read to apply to the application of credits.
                As NHTSA explained in the NPRM, it understands that the language in the
                MYs 2012-2016 final rule could be read to suggest that the transfer cap
                applies at the time credits are transferred. However, NHTSA believes
                its existing interpretation--that the transfer cap applies at the time
                the credits are used--is a more appropriate, plain language reading of
                the statute. While manufacturers have approached NHTSA with various
                interpretations that would essentially allow them to circumvent the
                EISA transfer cap, NHTSA believes such interpretations are improper
                because they would not give effect to the statutory transfer cap.
                Therefore, NHTSA proposed in the NPRM to deny Alliance/Global's
                petition to revise the definition of ``transfer'' in 49 CFR 536.3, and
                is now finalizing that denial.
                 In response to the tentative denial of the petition above in the
                NPRM, comments were received from the Global Automakers and Toyota
                asking NHTSA to reconsider applying the transfer cap of 2.0 mpg per
                year when credits are transferred rather than when they are
                applied.\3260\ They reiterated that imposing the cap when applying the
                credits is overly burdensome, but did not provide any new information
                that has persuaded NHTSA to change its view that the petition should be
                denied. The Auto Alliance also stated that NHTSA should revise its
                definition of ``transfer'' to be more consistent with EPA.\3261\
                ---------------------------------------------------------------------------
                 \3260\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
                12032; Toyota, Detailed Comments, NHTSA-2018-0067-12150.
                 \3261\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 Other more general comments to the NPRM were also received from
                Walter Kreucher, Jeremy Michalek, Global Automakers, the Auto Alliance,
                and Toyota, regarding the use of the credit transfer flexibility. These
                commenters generally appreciated the transfer flexibility for its
                ability to reduce compliance costs.\3262\ More specifically, Walter
                Kreucher commented that the ability to transfer credits between
                compliance categories was beneficial for manufacturers and allowed for
                efficiency in the markets and reduce compliance costs.\3263\
                ---------------------------------------------------------------------------
                 \3262\ See, e.g., Global Automakers, Detailed Comments, NHTSA-
                2018-0067-12032.
                 \3263\ Walter Kreucher, Detailed Comments, NHTSA-2018-0067-0444.
                ---------------------------------------------------------------------------
                 For the final rule, NHTSA is not making any changes to the existing
                provisions regarding transferring credits. NHTSA's position remains
                unchanged that the transfer cap in 49 U.S.C. 32903(g)(1) clearly limits
                the amount of performance increase for a manufacturer's fleet that
                fails to achieve the prescribed standards. The same statutory provision
                prevents NHTSA from changing its definition for transfer to be
                consistent with EPA. Consequently, NHTSA is not changing its definition
                or its previous interpretation that the application of transfer caps
                applies at the time the credits are used and not when
                [[Page 25225]]
                transferred. Therefore, NHTSA is finalizing its decision to deny the
                Auto Alliance and Global Automakers petition.
                (4) Minimum Domestic Passenger Car Standard
                 EPCA, as amended by EISA, addresses the minimum domestic passenger
                car standard (MDPCS), clearly stating that any manufacturer's
                domestically-manufactured passenger car fleet must meet the greater of
                either 27.5 mpg on average, or 92 percent of the average fuel economy
                projected by the Secretary for the combined domestic and non-domestic
                passenger automobile fleets manufactured for sale in the U.S. by all
                manufacturers in the model year, which projection shall be published in
                the Federal Register when the standard for that model year is
                promulgated in accordance with 49 U.S.C. 32902(b).\3264\ Since that
                requirement was added to the statute, NHTSA has always calculated the
                ``92 percent'' as greater than 27.5 mpg. NHTSA published the 92 percent
                MDPCS for MYs 2017-2025 at 49 CFR 531.5(d) as part of the 2012 final
                rule. 49 CFR 531.5(e) explains that the published MDPCS for MYs 2022-
                2025 are not final and may change when NHTSA sets standards for those
                model years. This is consistent with the statutory requirement that the
                92 percent standards must be determined at the time an overall
                passenger car standard is promulgated and published in the Federal
                Register.\3265\ Any time NHTSA establishes or changes a passenger car
                standard for a model year, the MDPCS for that model year must also be
                evaluated or re-evaluated and established accordingly. Thus, this final
                rule establishes the applicable MDPCS for MYs 2021-2026.
                ---------------------------------------------------------------------------
                 \3264\ 49 U.S.C. 32902(b)(4).
                 \3265\ 49 U.S.C. 32904(b)(4)(B).
                ---------------------------------------------------------------------------
                 NHTSA considered comments received about the MDPCS, and discusses
                the comments and the agency's assessment in Section VIII.B.1.b).
                 Table IX-7 lists the minimum domestic passenger car standards and
                compares them to standards that would correspond to each of the other
                regulatory alternatives considered. NHTSA has updated these to reflect
                its overall analysis and resultant projection for the CAFE standards
                finalized today, highlighted below as ``Preferred (Alternative 3),''
                and has calculated what those standards would be under the no action
                alternative (as issued in 2012, as updated for the NPRM, and as further
                updated by today's analysis) and under the other alternatives described
                and discussed further in Section V, above.
                [GRAPHIC] [TIFF OMITTED] TR30AP20.757
                (5) Fuel Savings Adjustment Factor
                 Under NHTSA's credit trading regulations, a fuel savings adjustment
                factor is applied when trading occurs between manufacturers or when a
                manufacturer transfers credits between its fleets, but not when a
                manufacturer carries credits forward or carries back credits within the
                same fleet.\3266\ The Alliance/Global requested in their 2016 petition
                that NHTSA require manufacturers to apply the fuel savings adjustment
                factor when credits are carried forward or carried back within the same
                fleet, including for existing, unused credits.
                ---------------------------------------------------------------------------
                 \3266\ See 49 CFR 536.4(c).
                ---------------------------------------------------------------------------
                 Per EISA, total oil savings must be preserved in NHTSA's credit
                trading program.\3267\ The statutory provisions for credit transferring
                within a manufacturer's fleet do not explicitly include the same
                requirement; however, NHTSA prescribed a fuel savings adjustment factor
                that applies to both credit trades between manufacturers and credit
                transfers between a manufacturer's compliance fleets.
                3268 3269
                ---------------------------------------------------------------------------
                 \3267\ 49 U.S.C. 32903(f)(1).
                 \3268\ 49 U.S.C. 32903(g).
                 \3269\ See 49 CFR 536.5; see also 74 FR 14430 (Mar. 30, 2009)
                (Per NHTSA's final rule for MY 2011 Average Fuel Economy Standards
                for Passenger Cars and Light Trucks, ``There is no other clear
                expression of congressional intent in the text of the statute
                suggesting that NHTSA would have authority to adjust transferred
                credits, even in the interest of preserving oil savings. However,
                the goal of the CAFE program is energy conservation; ultimately, the
                U.S. would reap a greater benefit from ensuring that fuel oil
                savings are preserved for both trades and transfers. Furthermore,
                accounting for traded credits differently than for transferred
                credits does add unnecessary burden on program enforcement. Thus,
                NHTSA will adjust credits both when they are traded and when they
                are transferred so that no loss in fuel savings occurs.'').
                ---------------------------------------------------------------------------
                 When NHTSA initially considered the preservation of oil savings,
                the agency
                [[Page 25226]]
                explained how one credit is not necessarily equal to another. For
                example, the fuel savings lost if the average fuel economy of a
                manufacturer falls one-tenth of an mpg below the level of a relatively
                low standard are greater than the average fuel savings gained by
                raising the average fuel economy of a manufacturer one-tenth of a mpg
                above the level of a relatively high CAFE standard.\3270\ The effect of
                applying the adjustment factor is to increase the numerical value of
                credits for compliance accounting that are earned for exceeding a CAFE
                standard, that are applied to a compliance category with a higher CAFE
                standard. Likewise, the adjustment factor has the effect of decreasing
                the numerical value of credits for compliance accounting that are
                earned for exceeding a CAFE standard, that are applied to a compliance
                category with a lower CAFE standard. While applying the adjustment
                factor impacts the compliance accounting value of credits which are
                denominated in miles per gallon, the adjustment maintains the real
                world value of credits from the perspective of the actual amount of
                fuel consumed or saved.
                ---------------------------------------------------------------------------
                 \3270\ 74 FR 14432 (Mar. 30, 2009).
                ---------------------------------------------------------------------------
                 Alliance/Global stated, in its 2016 petition, that while carry-
                forward and carryback credits have been used for many years, the CAFE
                standards did not change during the Congressional CAFE freeze, meaning
                credits earned during those years were associated with the same amount
                of fuel savings from year to year.\3271\ Alliance/Global suggest that
                because there is no longer a Congressional CAFE freeze, NHTSA should
                apply the adjustment factor when moving credits within a manufacturer's
                fleet (i.e. carry-forward or carryback) beginning retroactively in MY
                2011.\3272\
                ---------------------------------------------------------------------------
                 \3271\ Alliance/Global Petition at 10.
                 \3272\ Alliance/Global Petition at 4.
                ---------------------------------------------------------------------------
                 In the NPRM, NHTSA tentatively denied Alliance/Global's request to
                apply the fuel savings adjustment factor to credits that are carried
                forward or carried back within the same fleet to the extent that the
                request would impact credits carried forward or back retroactively
                within manufacturers' compliance fleets (i.e., credits that were
                generated prior to MY 2021 when the standards set by this rule first
                apply). NHTSA tentatively determined that applying the adjustment
                factor to credits earned in prior model years would be inequitable to
                apply retroactively. There would be an advantage for manufacturers
                carrying credits into future model years with higher CAFE standards.
                Manufacturers have historically planned compliance strategies based, at
                least in part, on the existing rules for how credits could be carried
                forward and back, including the lack of an adjustment factor when
                credits are carried forward or back within the same fleet. Thus,
                retroactively requiring an adjustment factor could disadvantage certain
                manufacturers without credits, and result in windfalls for other
                manufacturers.
                 To explore the impact on future model years, NHTSA sought
                additional comments in the NPRM on the feasibility of applying the fuel
                savings adjustment factor to credits carried forwards or back starting
                in MY 2021. Global Automakers submitted new comments arguing that the
                application of fuel savings adjustment factors to credits carried
                forward or back would not result in a credit windfall. They believed
                this practice would ensure that credits have a consistent value over
                time.\3273\
                ---------------------------------------------------------------------------
                 \3273\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
                12032.
                ---------------------------------------------------------------------------
                 Comments from Global Automakers provided no further justification
                that would persuade NHTSA to consider changing its position on denying
                the application of the adjustment factor to carry-forward and carryback
                credits beginning with MY 2011. NHTSA continues to be concerned about
                the inequitable outcome retroactive adjustments would have on the
                credit market. Therefore, NHTSA is finalizing its decision to deny the
                Alliance/Global request to apply the adjustment factor to credits
                carried forward or carried back within a compliance category
                retroactively beginning as early as MY 2011.
                 Congress expressly required that DOT establish a credit
                ``transferring'' regulation, to allow individual manufacturers to move
                credits from one of their fleets to another (e.g., using a credit
                earned for exceeding the light truck standard for compliance with the
                domestic passenger car standard). Congress also gave DOT discretion to
                establish a credit ``trading'' regulation so that credits may be bought
                and sold between manufacturers.\3274\ Congress specified that trading
                was for earned credits ``to be sold to manufacturers whose automobiles
                fail to achieve the prescribed standards such that the total oil
                savings associated with manufacturers that exceed the prescribed
                standards are preserved.'' \3275\ NHTSA established 49 CFR part 536
                believing it was consistent with the statute for transferred credits to
                be subject to the same ``adjustment factor'' to ensure total oil
                savings are preserved.\3276\ NHTSA believed that no further application
                of the adjustment factor to other credit flexibilities would be
                appropriate at that time. NHTSA sought comments in the NPRM to explore
                the consequences associated with applying the adjustment factor to
                credits carried forward and back starting in MY 2021, but no further
                insight was gained from the comments received. Therefore, NHTSA is
                retaining its existing requirements for the adjustment factor to be
                applied to transferred and traded credits only. NHTSA will continue
                considering potential application of the adjustment factor for all
                types of credit flexibilities in the future, and may consider
                regulatory changes in subsequent rulemakings.
                ---------------------------------------------------------------------------
                 \3274\ 49 U.S.C. 32903(f).
                 \3275\ 49 U.S.C. 32903(f)(1).
                 \3276\ 74 FR 14196, 14434 (Mar. 30, 2009).
                ---------------------------------------------------------------------------
                (6) VMT Estimates for Fuel Savings Adjustment Factor
                 NHTSA uses the vehicle miles traveled (VMT) estimate as part of its
                fuel savings adjustment equation to ensure that when traded or
                transferred credits are used, fuel economy credits are adjusted to
                ensure fuel oil savings is preserved.\3277\ For MYs 2017-2025, NHTSA
                finalized VMT values of 195,264 miles for passenger car credits, and
                225,865 miles for light truck credits.\3278\ These VMT estimates
                harmonized with those used in EPA's CO2 program. For MYs
                2011-2016, NHTSA estimated different VMTs by model year.
                ---------------------------------------------------------------------------
                 \3277\ See 49 CFR 536.4(c).
                 \3278\ 77 FR 63130 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 In the NPRM, NHTSA explained that Alliance/Global requested in
                their 2016 petition that NHTSA apply fixed VMT estimates to the fuel
                savings adjustment factor for MYs 2011-2016 similar to how NHTSA
                handled VMT values for MYs 2017-2025.\3279\ NHTSA rejected a similar
                request from the Auto Alliance in the MY 2017 and later rulemaking,
                citing lack of scope, and expressing concern about the potential loss
                of fuel savings.\3280\
                ---------------------------------------------------------------------------
                 \3279\ Alliance/Global Petition at 5, 11.
                 \3280\ Id.
                ---------------------------------------------------------------------------
                 The Alliance/Global argued that data from MYs 2011-2016 demonstrate
                that no fuel savings would have been lost, as was NHTSA's
                concern.\3281\ Alliance/Global asserted that by not revising the MY
                2012-2016 VMT estimates, credits earned during that timeframe were
                undervalued.\3282\ Therefore, Alliance/
                [[Page 25227]]
                Global argued that NHTSA should retroactively revise its VMT estimates
                to ``reflect better the real-world fuel economy results.'' \3283\
                ---------------------------------------------------------------------------
                 \3281\ Alliance/Global Petition at 11.
                 \3282\ Id.
                 \3283\ Alliance/Global Petition at 11.
                ---------------------------------------------------------------------------
                 Such retroactive adjustments could have unfair adverse effects upon
                manufacturers for decisions they made based on the regulations as they
                existed at the time. As Alliance/Global acknowledged, adjusting VMT
                estimates would disproportionately affect manufacturers that have a
                credit deficit and were part of EPA's Temporary Lead-time Allowance
                Alternative Standards (TLAAS). The TLAAS program sunsets for MYs 2021
                and later. Given that some manufacturers would be disproportionately
                affected were NHTSA to adopt Alliance/Global's proposal, in the NPRM,
                NHTSA tentatively denied Alliance/Global's request to change the
                agency's VMT schedules for MYs 2011-2016 retroactively. Alliance/
                Global's suggestion that a TLAAS manufacturer should be allowed to
                elect either approach does not change the fact that manufacturers in
                the TLAAS program made production decisions based on the regulations as
                understood at the time.\3284\ NHTSA sought comments on the Alliance/
                Global requests in the NPRM.
                ---------------------------------------------------------------------------
                 \3284\ See id. at 11-12, n.12.
                ---------------------------------------------------------------------------
                 However, no further comments were received on this issue in
                response to the NPRM. Therefore, NHTSA is finalizing its decision to
                deny the Alliance/Global request to modify the VMT schedules for MYs
                2011-2016.
                (7) Special Fuel Economy Calculations for Dual and Alternative Fueled
                Vehicles
                 As discussed at length in prior rulemakings, EPCA, as amended by
                EISA, encouraged manufacturers to build alternative-fueled and dual-
                (or flexible-) fueled vehicles by providing special fuel economy
                calculations for ``dedicated'' (that is, 100 percent) alternative
                fueled vehicles and ``dual-fueled'' (that is, capable of running on
                either the alternative fuel or gasoline/diesel) vehicles.
                 Dedicated alternative-fuel automobiles include electric, fuel cell,
                and compressed natural gas vehicles, among others. The statutory
                provisions for dedicated alternative fuel vehicles in 49 U.S.C.
                32905(a) state that the fuel economy of any dedicated automobile
                manufactured after MY 1992 shall be measured ``based on the fuel
                content of the alternative fuel used to operate the automobile. A
                gallon of liquid alternative fuel used to operate a dedicated
                automobile is deemed to contain 0.15 gallon of fuel.'' Under EPCA, for
                dedicated alternative fuel vehicles, there are no limits or phase-out
                for this special fuel economy calculation, unlike for duel-fueled
                vehicles, as discussed below.
                 EPCA's statutory incentive for dual-fueled vehicles at 49 U.S.C.
                32906 and the measurement methodology for dual-fueled vehicles at 49
                U.S.C. 32905(b) and (d) expire after MY 2019; therefore, NHTSA had to
                examine the future of these provisions in the MY 2017 and later CAFE
                rulemaking. NHTSA and EPA concluded that it would be inappropriate to
                measure duel-fueled vehicles' fuel economy like that of conventional
                gasoline vehicles with no recognition of their alternative fuel
                capability, which would be contrary to the intent of EPCA/EISA. The
                agencies determined that for MY 2020 and later vehicles, the general
                statutory provisions authorizing EPA to establish testing and
                calculation procedures provide discretion to set the CAFE calculation
                procedures for those vehicles. The methodology for EPA's approach is
                outlined in the 2012 final rule for MYs 2017 and later at 77 FR 63128
                (Oct. 15, 2012). In the NPRM, NHTSA sought comments on that current
                approach.
                 NHTSA received comments from the Coalition for Renewable Natural
                Gas, NGV America, the American Gas Association, the American Public Gas
                Association, CARB, Ingevity Corporation, Fuel Freedom Foundation, UCS,
                National Farmers Union, Indiana Corn Growers Association, Volkswagen,
                and a joint submission from Ariel Corp. and VNG.co.
                 Fuel Freedom Foundation and the National Farmers Union asserted
                that the agencies should continue offering incentives for emerging
                technology vehicles including natural gas vehicles, internal combustion
                engine (ICE) vehicles that encourage renewable fuel use, electric and
                hydrogen fuel cell vehicles, flex-fuel vehicles (FFVs), and dedicated
                high-octane vehicles designed for compatibility with mid-level ethanol
                blends.\3285\
                ---------------------------------------------------------------------------
                 \3285\ Fuel Freedom Foundation, Detailed Comments, NHTSA-2018-
                0067-12016; National Farmers Union, Detailed Comments, NHTSA-2018-
                0067-11972.
                ---------------------------------------------------------------------------
                 Indiana Corn Growers Association and Fuel Freedom Foundation
                specified that FFVs, as well as vehicles that run on mid-level ethanol
                blends, should receive credit for the petroleum reduction value.\3286\
                For vehicles using higher-ethanol blends, these commenters stated that
                the agencies should establish more accurate petroleum equivalency
                factors for the proportion of ethanol versus gas.\3287\ Clean Fuels
                Development Coalition requested credits for producing ``Engines
                Optimized for High-Octane'' be reinstated.\3288\ Volkswagen made the
                same request and added that a pathway to higher-octane fuel is
                important to it.\3289\
                ---------------------------------------------------------------------------
                 \3286\ Indiana Corn Growers Association, Detailed Comments,
                NHTSA-2018-0067-12003; Fuel Freedom Foundation, Detailed Comments,
                NHTSA-2018-0067-12016.
                 \3287\ Fuel Freedom Foundation, Detailed Comments, NHTSA-2018-
                0067-12016.
                 \3288\ Clean Fuels Development Coalition, Detailed Comments,
                NHTSA-2018-0067-12031.
                 \3289\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
                ---------------------------------------------------------------------------
                 Ariel Corp. and VNG.co, the Coalition for Renewable Natural Gas,
                NGVAmerica, the American Gas Association, and the American Public Gas
                Association commented that the agencies should expand incentives for
                natural gas vehicles in the light-duty sector especially for pick-up
                trucks, work vans, and sport utility vehicles.\3290\ They argued that
                current incentives are not strong enough to induce manufacturers to
                produce natural gas vehicles. They further requested that the market
                penetration rates be removed for light-duty trucks.\3291\
                ---------------------------------------------------------------------------
                 \3290\ Joint submission from Ariel Corp and VNG.co LLC, Detailed
                Comments, NHTSA-2018-0067-7573; Joint submission from the Coalition
                for Renewable Natural Gas, NVG America, the American Gas
                Association, and American Public Gas Association, Detailed Comments,
                NHTSA-2018-0067-11967.
                 \3291\ See, e.g., joint submission from the Coalition for
                Renewable Natural Gas, NGVAmerica, the American Gas Association, and
                the American Public Gas Association, Detailed Comments, NHTSA-2018-
                0067-11967.
                ---------------------------------------------------------------------------
                 The Coalition for Renewable Natural Gas, NGVAmerica, the American
                Gas Association, and the American Public Gas Association argued that an
                AMFA factor of 0.15 is low and because some natural gas vehicles can
                operate at 100 percent natural gas, a higher fuel economy credit is
                justified. They further supported a permanent use of the 0.15 factor
                for dual-fuel vehicles.\3292\ Similarly, Ingevity Corporation, and
                Ariel Corp. and VNG.co argued that natural gas vehicle emissions should
                return to the 0.15 divisor.\3293\
                ---------------------------------------------------------------------------
                 \3292\ Joint submission from the Coalition for Renewable Natural
                Gas, NGVAmerica, the American Gas Association, and the American
                Public Gas Association, Detailed Comments, NHTSA-2018-0067-11967.
                 \3293\ Ingevity Corporation, Detailed Comments, NHTSA-2018-0067-
                8666; Joint submission from Ariel Corp. and VNG.co LLC, Detailed
                Comments, NHTSA-2018-0067-7573.
                ---------------------------------------------------------------------------
                 Ingevity Corporation, Ariel Corp. and VNG.co, the Coalition for
                Renewable
                [[Page 25228]]
                Natural Gas, NGVAmerica, the American Gas Association, and the American
                Public Gas Association requested that the agencies remove the minimum
                driving range of natural gas compared to gasoline and ``drive to
                empty'' design requirements for dual-fueled natural gas vehicles and
                allow higher utility factors based on driving range only, so that dual-
                fuel NGVs are treated similarly to PHEVs. They stated a belief that the
                design constraints for dual-fuel NGVshold NGVs to an unfairly higher
                standard.\3294\ As discussed above in Section IX.B, EPA is removing
                these design constraints for dual-fuel NGVs.
                ---------------------------------------------------------------------------
                 \3294\ Ingevity, Detailed Comments, NHTSA-2018-0067-8666; Joint
                submission from Ariel Corp. and VNG.co LLC, Detailed Comments,
                NHTSA-2018-0067-7573; Joint submission from The Coalition for
                Renewable Natural Gas, NGVAmerica, the American Gas Association, the
                American Public Gas Association, Detailed Comments, NHTSA-2018-0067-
                11967.
                ---------------------------------------------------------------------------
                 CARB argued that flexibilities for natural gas vehicles and high-
                octane blend vehicles are not yet warranted.\3295\ Similarly, UCS
                argued that natural gas is a greenhouse gas and benefits from natural
                gas vehicles are undermined by their costs. UCS further commented that
                natural gas vehicle technology does not need any incentives since it
                has already been deployed and in the market.\3296\
                ---------------------------------------------------------------------------
                 \3295\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
                 \3296\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
                ---------------------------------------------------------------------------
                 In response to comments, NHTSA has determined that EPCA and EISA
                prescribe the incentive that is used for dedicated liquid and gaseous
                alternative fuel vehicles, and the CAFE program will continue to use
                those statutory incentives. For dedicated alternative fuel vehicles,
                the statute provides a significant incentive that only counts 15
                percent of the actual energy used.\3297\ For dual fuel vehicles, NHTSA
                has determined that, for the portion of operation that occurs on an
                alternative fuel, it is consistent to use the same incentive that is
                specified by EPCA and EISA for dedicated fuel vehicles. For example,
                for the hypothetical case of a vehicle that operates 99 percent of the
                time on an alternative fuel, it would be appropriate for that vehicle
                to receive nearly the same incentive as a dedicated alternative fuel
                vehicle that operates 100 percent of the time on alternative fuel.
                Applying the same 15 percent of energy used incentive for both
                dedicated and duel fuel vehicles remains appropriate. NHTSA therefore
                is not adopting any new incentives for any alternative fueled vehicles.
                ---------------------------------------------------------------------------
                 \3297\ 32905(a) ``. . . A gallon of a liquid alternative fuel
                used to operate a dedicated automobile is deemed to contain .15
                gallon of fuel.'' 32905(c) ``. . . One hundred cubic feet of natural
                gas is deemed to contain .823 gallon equivalent of natural gas. The
                Secretary of Transportation shall determine the appropriate gallon
                equivalent of other gaseous fuels. A gallon equivalent of gaseous
                fuel is deemed to have a fuel content of .15 gallon of fuel.''
                ---------------------------------------------------------------------------
                D. Compliance Issues That Affect Both the CO2 and CAFE
                Programs
                 Because the real world CO2 emissions reduction benefits
                of certain technologies cannot be measured or fully measured using 2-
                cycle test procedures, EPA established new compliance flexibilities
                under its CAA authority, starting in MY 2012, that allow manufacturers
                credit for emission compliance for installing these technologies. Those
                flexibilities are designed to recognize improvements in A/C systems
                with greater efficiency and other ``off-cycle'' technologies that
                reduce real world tailpipe CO2 emissions. More specifically,
                real world improvements that cannot be measured or fully measured on 2-
                cycle tests are determined and used to calculate additional
                CO2 credits (in Megagrams (Mg)) for each model type that has
                the technologies. Because these tailpipe CO2 improving
                technologies also impact fuel economy, NHTSA adopted the same
                flexibilities and incentives beginning in MY 2017. EPA and NHTSA also
                established incentives for both the CO2 and CAFE programs
                that give added compliance credits and fuel consumption improvement
                values for the production of strong and mild hybrid full-size pickup
                trucks beginning in MY 2017.\3298\ EPA adjusts manufacturers' CAFE
                performance values using the emissions benefits or incentives provided
                for these technologies. EPA developed a methodology for manufacturers
                to increase their passenger car and light truck fuel economy
                performance in accordance with procedures set forth by EPA in 40 CFR
                part 600. For the NHTSA CAFE program, the CO2 reductions (in
                grams per mile) are converted to fuel consumption improved values
                (FCIVs, gallons per mile) and then the benefits are summed for all the
                model types in the manufacturer's fleets. The total FCIVs are used to
                adjust and increase manufacturers' CAFE (mpg) performance values.
                ---------------------------------------------------------------------------
                 \3298\ See 40 CFR 86.1867-86.1868, 86.1870.
                ---------------------------------------------------------------------------
                 It is important to note that while these flexibilities and
                incentives have similar value for compliance in the CAFE and
                CO2 programs, there are differences in how they are
                accounted for in each of the programs due to differences in the
                structure of the programs. The CAFE program accounts for A/C efficiency
                and off-cycle improvements through EPA measurement procedures that
                determine fuel consumption improvement values (FCIVs). The CAFE A/C
                efficiency and off-cycle provisions do not involve manufacturer
                credits.\3299\ There are no bankable, tradable, or transferrable
                credits earned by a manufacturer for implementing more efficient A/C
                systems or installing an off-cycle technology. In fact, the only
                credits provided for in NHTSA's CAFE program are those earned by
                overcompliance with a standard.\3300\ As discussed above, EPA adjusts
                CAFE performance values based on the FCIVs generated through the use of
                these technologies. Off-cycle technologies and A/C efficiency
                improvements represent adjustments to individual vehicle compliance
                values based on the fuel consumption improvement values of these
                technologies.
                ---------------------------------------------------------------------------
                 \3299\ This is not to be confused with EPA's parallel program,
                which refers to the GHG's consideration of A/C improvements and off-
                cycle technologies as ``credits.''
                 \3300\ 49 U.S.C. 32903.
                ---------------------------------------------------------------------------
                 Illustrative of this confusion, in the 2016 Alliance/Global
                petition, the petitioners asked NHTSA to avoid imposing unnecessary
                restrictions on the use of credits. Alliance/Global referenced language
                from an EPA report that stated compliance is assessed by measuring the
                tailpipe emissions of a manufacturer's vehicles, and then reducing
                vehicle CO2 compliance values depending on A/C efficiency
                improvements and off-cycle technologies.\3301\ This language is
                consistent with NHTSA's statement in the MY 2017 and later final rule,
                which explained how the agencies coordinate and apply off-cycle and A/C
                adjustments. ``There will be separate improvement values for each type
                of credit, calculated separately for cars and for trucks. These
                improvement values are subtracted from the manufacturer's 2-cycle-based
                fleet fuel consumption value to yield a final new fleet fuel
                consumption value, which would be inverted to determine a final fleet
                fuel CAFE value.'' \3302\
                ---------------------------------------------------------------------------
                 \3301\ See Alliance/Global Petition at 15.
                 \3302\ 77 FR 62726 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 In the NPRM, NHTSA proposed to deny Alliance/Global's request
                because what the petitioners refer to as ``technology credits'' are
                actually FCIVs applied to the fuel economy performance of individual
                vehicles.\3303\
                [[Page 25229]]
                Thus, these adjustments are not actually ``credits,'' per the usage of
                ``credit'' in EPCA/EISA and are not subject to the ``carry-forward''
                and ``carryback'' provisions in 49 U.S.C. 32903. To alleviate
                confusion, and to ensure consistency in nomenclature, NHTSA proposed to
                update language in its regulations to reflect that the use of the term
                ``credits'' to refer to A/C efficiency and off-cycle technology
                adjustments should actually be termed fuel consumption improvement
                values (FCIVs). No further comments were received on this issue in
                response to the NPRM. For the final rule, NHTSA is finalizing the
                proposed changes in its regulations to remove the term ``credits'' and
                to replace it with the term ``adjustments'' for the FCIV benefit for A/
                C and off-cycle technologies in the CAFE program.
                ---------------------------------------------------------------------------
                 \3303\ The agencies also refer to A/C and off-cycle technology
                improvement values as ``credits'' sporadically throughout their
                regulations. NHTSA is amending its regulations to reflect these are
                adjustments and not actual credits that can be carried forward or
                back. For a further discussion, see above.
                ---------------------------------------------------------------------------
                 Manufacturers seeking to use these flexibilities and incentives
                start the process each model year by submitting information to EPA and
                seeking any necessary approvals, as appropriate. The use of certain
                technologies only requires submitting information to EPA, whereas
                others require a formal request process for approval. The differences
                are explained in the following sections. The compliance information
                manufacturers must submit to EPA describes the technologies, the
                flexibilities or incentives being used, and the testing approach for
                deriving benefits. Initial information is required as a part of the EPA
                certification process, as specified by 40 CFR 86.1843-01 in advance of
                each model year. For technologies requiring approvals, EPA must confirm
                the manufacturer's testing approach, receive test results to assess the
                benefit of the technology, and then where applicable issue a Federal
                Register notice that invites public comment. EPA review and
                determination usually occurs before the end of the compliance model
                year, if manufacturers provide information to EPA on a timely basis. To
                receive the benefit under the CAFE program for technologies that
                require approvals, manufacturers must concurrently submit to NHTSA the
                same information that is sent to EPA. EPA consults with NHTSA in
                reviewing A/C efficiency and off-cycle adjustments to fuel economy
                performance values that require approval. NHTSA provides EPA its
                assessment of the suitability of a technology considering: (1) Whether
                the technology has a direct impact upon improving fuel economy
                performance; (2) whether the technology is related to crash-avoidance
                technologies, safety critical systems or systems affecting safety-
                critical functions, or technologies designed for the purpose of
                reducing the frequency of vehicle crashes; (3) information from any
                assessments conducted by EPA related to the application, the
                technology, and/or related technologies; and (4) any other relevant
                factors.
                 EPA and NHTSA sought comments on several aspects of the shared
                flexibilities and incentives in the NPRM. Presented in the following
                sections is a summary of the comments received and the agencies final
                decisions for the final rule.
                1. Incentives for Advanced Technologies in Full-Size Pickup Trucks
                 In the 2012 rulemaking for MYs 2017 and beyond, EPA and NHTSA
                created incentives to encourage implementation of hybrid electric full
                size pickup trucks for both the CO2 and CAFE programs.
                CO2 credits and CAFE FCIVs were made available for
                manufacturers that produce full-size pickup trucks with Mild HEV or
                Strong HEV technology, provided the percentage of production with the
                technology is greater than specified percentages.\3304\ In addition,
                CO2 credits and CAFE FCIVs were made available for
                manufacturers that produce full-size pickups with other technologies
                that enables full size pickup trucks to exceed performance of their
                CO2 or CAFE targets based on footprints by specified
                amounts.\3305\ These performance-based incentives created a technology-
                neutral path (as opposed to the other technology-encouraging path) to
                achieve the CO2 credits and CAFE FCIVs, which would
                encourage the development and application of new technological
                approaches.
                ---------------------------------------------------------------------------
                 \3304\ 77 FR 62651 (Oct. 15, 2012).
                 \3305\ Id.
                ---------------------------------------------------------------------------
                 EPA and NHTSA established limits on the vehicles eligible to
                qualify for these incentives; a truck must meet minimum criteria for
                bed size and towing or payload capacity, and meet minimum production
                thresholds (in terms of a percentage of a manufacturer's full-size
                pickup truck fleet) in order to qualify for the incentives. As
                designed, the strong hybrid credit is 20 grams/mile per vehicle,
                available through MY 2025, if installed on at least 10 percent of the
                manufacturer's full-size pickup truck fleet in the model year. The
                program also included an incentive for mild hybrids of 10 grams/mile
                per vehicle during MYs 2017-2021. To be eligible the manufacturer would
                have to show that the mild hybrid technology is utilized in a specified
                portion of its truck fleet beginning with at least 20 percent of a
                company's full-size pickup production in MY 2017 and ramping up to at
                least 80 percent in MY 2021.\3306\
                ---------------------------------------------------------------------------
                 \3306\ 77 FR 62651-2 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 At present, no manufacturer has qualified to use the full-size
                pickup truck incentives. One vehicle manufacturer introduced a mild
                hybrid pickup truck for MY 2019 but did not meet the minimum production
                threshold. Others have announced potential collaborations, or have
                already started production on future hybrid or electric models.\3307\
                ---------------------------------------------------------------------------
                 \3307\ Chrysler released the 2019 Dodge Ram 1500 ``eTorque''
                (see https://www.fueleconomy.gov/feg/Find.do?action=sbs&id=40736&id=40737&id=40394&id=40397) which
                qualifies as a mild hybrid pickup truck by replacing the traditional
                alternator on the engine with a 48-volt Li-on battery-powered, belt-
                driven motor generator that improves performance, efficiency,
                payload, towing capabilities and drivability. The production volume
                of these vehicles did not qualify for the full-size pickup truck
                electric/hybrid incentive for MY 2019. Other vehicle models are
                currently in research or in development for future years but it is
                uncertain whether they will reach the required sales volumes to
                qualify for incentives. For example, the hybrid and battery-electric
                versions of the F-150 pickup, see https://www.trucks.com/2019/09/18/ford-truck-engineer-explains-electric-f-150-pickup-plans (September
                18, 2019), or the new electric pickup truck manufactured by Rivian,
                https://www.trucks.com/2019/04/24/ford-plans-new-electric-truck-rivian-invests-500-million/ (April 24, 2019); or the Tesla all
                electric pickup truck (https://www.cnn.com/2019/11/08/success/tesla-pickup-reveal/index.html) (November 8, 2019).
                ---------------------------------------------------------------------------
                 Prior to the NPRM, the agencies received input from automakers that
                these incentives should be extended and available to all light-duty
                trucks (e.g., cross-over vehicles, minivans, sport utility vehicles,
                and smaller-sized pickups) and not only full-size pickup trucks.\3308\
                Automakers also recommended that the program's eligibility production
                thresholds should be removed because they discourage the application of
                technology since manufacturers cannot be confident of achieving the
                thresholds. Some stakeholders have also suggested an additional
                incentive for strong and mild hybrid passenger cars. In the proposal,
                the agencies sought comment on whether these incentives should be
                expanded along the lines suggested by stakeholders, on the basis that
                perhaps these incentives could lead to additional product offerings of
                strong hybrids, and technologies that offer similar emissions
                reductions, which could enable manufacturers to achieve additional
                long-term CO2 emissions reductions. In addition, the
                agencies sought comment on whether to extend either the incentive for
                hybrid full-size pickup trucks or the performance-based incentive past
                the dates that EPA specified in the 2012 final rule for MY
                [[Page 25230]]
                2017 and later. The agencies also sought comment on eliminating
                incentive programs, as discussed above.
                ---------------------------------------------------------------------------
                 \3308\ 83 FR 43461 (Aug. 24, 2018).
                ---------------------------------------------------------------------------
                 The agencies received a variety of comments on the full-size pickup
                truck incentives. Comments were received from General Motors,
                Volkswagen, Honda, BorgWarner, Fiat Chrysler, Toyota, DENSO
                International, Ford, CARB, Global Automakers, UCS, Electric Drive
                Transportation Association, the Auto Alliance, Ariel Corp. and VNG.co,
                ACEEE, the Coalition for Renewable Natural Gas, NGVAmerica, the
                American Gas Association, and the American Public Gas Association.
                 The Auto Alliance, Toyota, General Motors, BorgWarner, Global
                Automakers, and Volkswagen advocated to expand the full-size pickup
                truck hybrid incentives to all hybrid vehicles.\3309\ They argued that
                prices for all hybrid-drive technologies are projected to remain high
                and consumer demand for these vehicles is still slow to increase.\3310\
                They asserted that expanding the full-size pickup truck hybrid
                incentive to all hybrid vehicles will help encourage investments in
                hybrid technology and continue to help manufacturers address their
                compliance challenges.\3311\ Similarly, these commenters reported that
                the current market, fueled by consumer demand for SUVs and lower than
                expected gas prices, is not conducive to consumer acceptance of or
                demand for electric vehicles.\3312\ For these reasons, they stated
                their belief that it is important to support adjustments and expansion
                of the current incentives to promote hybrid technologies.
                ---------------------------------------------------------------------------
                 \3309\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Toyota, Detailed Comments, NHTSA-2018-0067-12150; General Motors,
                Detailed Comments, NHTSA-2018-0067-11858; BorgWarner, Detailed
                Comments, NHTSA-2018-0067-11895; Global Automakers, Detailed
                Comments, NHTSA-2018-0067-12032; Volkswagen, Detailed Comments,
                NHTSA-2017-0069-0583.
                 \3310\ See, e.g., Auto Alliance, Detailed Comments, NHTSA-2018-
                0067-12073.
                 \3311\ See, e.g., General Motors, Detailed Comments, NHTSA-2018-
                0067-11858.
                 \3312\ See, e.g., Toyota, Detailed Comments, NHTSA-2018-0067-
                12150.
                ---------------------------------------------------------------------------
                 The Auto Alliance, DENSO International, Global Automakers, Fiat
                Chrysler, and Honda also argued for alternative pathways for the
                agencies to consider allowing the full-size pickup truck hybrid
                incentives to be expanded to the light-duty truck segment, but not to
                all passenger vehicles. They argued that hybrid technology has been
                slow to be applied in the light-duty truck segment, but has been
                broadly applied to passenger cars.\3313\
                ---------------------------------------------------------------------------
                 \3313\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                DENSO, Detailed Comments, NHTSA-2018-0067-11880; Global Automakers,
                Detailed Comments, NHTSA-2018-0067-12032; Fiat Chrysler, Detailed
                Comments, NHTSA-2018-0067-11943; Honda, Detailed Comments, NHTSA-
                2018-0067-11818.
                ---------------------------------------------------------------------------
                 Toyota, Global Automakers, and the Auto Alliance suggested the
                incentives for light-duty trucks should amount to 20 grams/mile.\3314\
                Global Automakers added that in addition to expanding full-size pickup
                truck hybrid incentives to light trucks, the agency should consider a
                smaller incentive for hybrid electric passenger vehicles as well.\3315\
                The Auto Alliance and Toyota suggested a 10 grams/mile credit for
                passenger cars.\3316\ Volkswagen further requested the hybrid pickup
                credit to be expanded to all hybrid cars and trucks.\3317\
                ---------------------------------------------------------------------------
                 \3314\ Toyota, Detailed Comments, NHTSA-2018-0067-12150; Global
                Automakers, Detailed Comments, NHTSA-2018-0067-12032; Auto Alliance,
                Detailed Comments, NHTSA-2018-0067-12073.
                 \3315\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
                12032.
                 \3316\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Toyota, Detailed Comments, NHTSA-2018-0067-12150.
                 \3317\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
                ---------------------------------------------------------------------------
                 Toyota, the Auto Alliance, Electric Drive Transportation
                Association, Ford, DENSO International, Global Automakers, Fiat
                Chrysler, and BorgWarner commented that having minimum production
                percentages for hybrid pickup trucks discourages manufacturers from
                investing in hybrid technologies. They requested that the agencies
                consider eliminating the percentage of production requirement and
                provide incentives in proportion to the value of the technology.\3318\
                Ford stated that the minimum production percentages unfairly penalize
                larger manufacturers who must produce more pickup trucks to claim the
                incentives than a smaller volume manufacturer.\3319\
                ---------------------------------------------------------------------------
                 \3318\ Toyota, Detailed Comments, NHTSA-2018-0067-12150; Auto
                Alliance, Detailed Comments, NHTSA-2018-0067-12073; Electric Drive
                Transportation Association, Detailed Comments, NHTSA-2018-0067-1201;
                Ford, Detailed Comments, NHTSA-2018-0067-11928; DENSO, Detailed
                Comments, NHTSA-2018-0067-11880; Global Automakers, Detailed
                Comments, NHTSA-2018-0067-12032; Fiat Chrysler, Detailed Comments,
                NHTSA-2018-0067-11943; BorgWarner, Detailed Comments, NHTSA-2018-
                0067-11895.
                 \3319\ Ford, Detailed Comments, NHTSA-2018-0067-11928.
                ---------------------------------------------------------------------------
                 Ariel Corp. and VNG.co, the Coalition for Renewable Natural Gas,
                NGVAmerica, the American Gas Association, and the American Public Gas
                Association commented the pickup truck incentives should be expanded to
                include natural gas vehicles.\3320\ They suggested a ``Natural Gas
                Pickup'' incentive like the hybrid-electric and performance-based
                pickup credits, but no minimum production requirement.\3321\
                ---------------------------------------------------------------------------
                 \3320\ Joint submission from Ariel Corp. and VNG.co, Detailed
                Comments, NHTSA-2018-0067-7573; Joint submission from The Coalition
                for Renewable Natural Gas, NGVAmerica, the American Gas Association,
                and the American Public Gas Association, Detailed Comments, NHTSA-
                2018-0067-11967.
                 \3321\ See, e.g., Joint submission from Ariel Corp. and VNG.co,
                Detailed Comments, NHTSA-2018-0067-7573.
                ---------------------------------------------------------------------------
                 ACEEE and UCS commented that hybrid technology has been around for
                quite a while and has been applied in every vehicle class. They
                discouraged the agencies from applying more incentives to these
                vehicles.\3322\ Specifically, UCS stated that incentives for electric
                vehicles are mostly driven by state regulation, and EPA and NHTSA
                policies are rewarding manufacturers for meeting standards they were
                already required to meet.\3323\ UCS commented that hybrids are not
                innovators or game-changing vehicles--they are simply one of many
                strategies by which manufacturers can reduce emissions and should not
                receive special treatment.\3324\
                ---------------------------------------------------------------------------
                 \3322\ ACEEE, Detailed Comments, NHTSA-2018-0067-12122-29; UCS,
                Detailed Comments, NHTSA-2018-0067-12039.
                 \3323\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
                 \3324\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
                ---------------------------------------------------------------------------
                 CARB commented that incentives for full-size hybrid pickup trucks
                should remain limited in their scope and that increasing or expanding
                those incentives can erode emissions benefits.\3325\ CARB further
                commented that hybrid electric vehicles (HEVs) are widely available at
                varying levels of power and performance across vehicle sizes, and CARB
                does not believe HEVs deserve special treatment in the CO2
                vehicle regulations.
                ---------------------------------------------------------------------------
                 \3325\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
                ---------------------------------------------------------------------------
                 After carefully considering the comments received, EPA and NHTSA
                are not adopting any new or expanded incentives for hybrid vehicles or
                full-size pickup trucks, and are removing these incentives beginning in
                MY 2022 (the incentive for mild hybrids expires after MY 2021
                regardless, so that does not change). The agencies believe any new or
                expanded incentives would likely not result in any further emissions
                benefits or fuel economy improvements since an increase in sales volume
                would not be expected. The agencies agree with CARB and ACEEE, and UCS
                that hybrids are a well-
                [[Page 25231]]
                established technology that has already been applied to a wide range of
                vehicles and, as such, no further incentives are warranted at this
                time. Further, the agencies believe that incentivizing manufacturers to
                implement specific technologies is inappropriate, as manufacturer fuel
                economy performance should represent actual fuel consumption. The
                agencies believe any new or expanded incentives for hybrids would
                likely not result in any further emissions benefits or fuel economy
                improvements beyond those measured during testing; to the extent that
                manufacturers choose to build full-size pickup trucks that exceed their
                targets, those will reap the benefits of target exceedance in the
                overall fleet averaging. Manufacturers did not provide sufficient
                evidence to support their position in a manner that leads the agencies
                to conclude otherwise, and there does not appear to be any likelihood
                that manufacturers will be able to take advantage of these
                flexibilities beyond MY 2021 that makes it necessary to retain them.
                Therefore, the agencies are removing these flexibilities from the
                program starting with MY 2022.
                2. Flexibilities for Air Conditioning Efficiency
                 A/C systems are virtually standard automotive accessories, and more
                than 95 percent of new cars and light trucks sold in the U.S. are
                equipped with mobile A/C systems. A/C system usage places a load on an
                engine, which results in additional tailpipe CO2 emissions
                and fuel consumption; the high penetration rate of A/C systems
                throughout the light-duty vehicle fleet means that efficient systems
                can significantly impact the total energy consumed and CO2
                emissions. A/C systems also have non-CO2 emissions
                associated with refrigerant leakage.\3326\ Manufacturers can improve
                the efficiency of A/C systems though redesigned and refined A/C system
                components and controls.\3327\ That said, such improvements are not
                measurable or recognized using 2-cycle test procedures, since A/C is
                turned off during 2-cycle testing. Any A/C system efficiency
                improvements that reduce load on the engine and improve fuel economy is
                therefore not measurable on those tests.
                ---------------------------------------------------------------------------
                 \3326\ See Section V for further details. Notably, manufacturers
                cannot claim CAFE-related benefits for reducing A/C leakage or
                switching to an A/C refrigerant with a lower global warming
                potential. While these improvements reduce GHG emissions consistent
                with the purpose of the CAA, they generally do not impact fuel
                economy and, thus, are not relevant to the CAFE program.
                 \3327\ The approach for recognizing potential A/C efficiency
                gains is to utilize, in most cases, existing vehicle technology/
                componentry, but with improved energy efficiency of the technology
                designs and operation. For example, most of the additional A/C-
                related load on an engine is because of the compressor, which pumps
                the refrigerant around the system loop. The less the compressor
                operates, the less load the compressor places on the engine
                resulting in less fuel consumption and CO2 emissions.
                Thus, optimizing compressor operation with cabin demand using more
                sophisticated sensors, controls, and control strategies is one path
                to improving the efficiency of the A/C system. For further
                discussion of A/C efficiency technologies, see Section II.D of the
                NPRM and Chapter 6 of the accompanying PRIA.
                ---------------------------------------------------------------------------
                 The CO2 and CAFE programs include flexibilities to
                account for the real world CO2 emissions and fuel economy
                improvements associated with improved A/C systems and to include the
                improvements for compliance.\3328\ The total of A/C efficiency credits
                is calculated by summing the individual credit values for each
                efficiency improving technology used on a vehicle, as specified in the
                A/C credit menu. The total A/C efficiency credit sum for each vehicle
                is capped at 5.0 grams/mile for cars and 7.2 grams/mile for trucks.
                Additionally, the off-cycle credit program contains credit earning
                opportunities for technologies that reduce the thermal loads on a
                vehicle from environmental conditions (solar loads or parked interior
                air temperature).\3329\ These technologies are listed on a thermal
                control menu that provides a predefined improvement value for each
                technology. If a vehicle has more than one thermal load improvement
                technology, the improvement values are added together, but subject to a
                cap of 3.0 grams/mile for cars and 4.3 grams/mile for trucks.
                ---------------------------------------------------------------------------
                 \3328\ See 40 CFR 86.1868-12.
                 \3329\ See 40 CFR 86.1869-12(b).
                ---------------------------------------------------------------------------
                 EPA requested comment on the A/C caps and on whether A/C efficiency
                technologies and off-cycle thermal control technologies should be
                combined under a single cap, since the technologies directly interact
                with each other. That is, improved thermal control results in reduced
                A/C loads for the more efficient A/C technologies. If the thermal
                credits were removed from the off-cycle menu, they would no longer be
                counted against the 10 grams/mile menu cap discussed above,
                representing a way to provide more room under the menu cap for other
                off-cycle technologies. Specifically, EPA sought comment on replacing
                the current off-cycle thermal efficiency capped value of 10 grams/mile,
                with separate caps of 8 grams/mile for cars and 11.5 grams/mile for
                trucks.
                 Comments concerning the A/C caps were received from the Auto
                Alliance, DENSO, Fiat Chrysler, and Volkswagen. DENSO commented that A/
                C efficiency credits earned through the off-cycle petition process
                should not count toward the A/C credit cap. If A/C credits granted
                through the off-cycle petition process are no longer counted toward the
                A/C credit cap, it stated that manufacturers would be significantly
                incentivized to develop new and innovative technologies.\3330\ Fiat
                Chrysler requested that certain A/C credits for electrical technologies
                (i.e., A/C blower motor controls that limit wasted electrical energy)
                be transferred to the off-cycle credit list.\3331\ Volkswagen further
                supported the removal of the thermal control technology credit caps and
                suggested that implementing caps at the fleet average level, rather
                than per-vehicle, could be less constraining.\3332\ DENSO pointed to an
                NREL study which found that A/C improvements were greater than
                previously thought possible. Therefore, it requested the agencies
                consider increasing the A/C credit cap.\3333\
                ---------------------------------------------------------------------------
                 \3330\ DENSO, Detailed Comments, NHTSA-2018-0067-11880.
                 \3331\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3332\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
                 \3333\ DENSO, Detailed Comments, NHTSA-2018-0067-11880.
                ---------------------------------------------------------------------------
                 Similarly, the Auto Alliance and Fiat Chrysler suggested raising
                the cap on A/C efficiency and thermal control technology by 64 percent
                and combine them under a single cap.\3334\ Additionally, they proposed
                increasing A/C efficiency and thermal control technology credits by up
                to 64 percent.\3335\ They also proposed that the agencies create new
                regulatory provisions to handle additional new A/C and thermal
                technologies.\3336\
                ---------------------------------------------------------------------------
                 \3334\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                 \3335\ See, e.g., Fiat Chrysler, Detailed Comments, NHTSA-2018-
                0067-11943.
                 \3336\ See, e.g., Auto Alliance, Detailed Comments, NHTSA-2018-
                0067-12073.
                ---------------------------------------------------------------------------
                 As with increasing the credit caps, manufacturers and suppliers
                were generally supportive of higher credit caps, or no caps at all, for
                this combined technology group. However, EPA has decided not to adopt
                any changes to the caps, including combining the A/C efficiency and
                thermal controls menu, due to the uncertainty regarding the menu credit
                values. Additional uncertainty exists for these technology groups
                because there are likely synergistic effects between A/C efficiency and
                thermal technologies that would need to be further considered in
                determining appropriate credit levels if
                [[Page 25232]]
                the two groups of technologies are combined under a single cap. Data is
                not currently available to consider these effects. Therefore, the
                agencies are not making any changes to the flexibilities for A/C
                efficiency improvements in the CO2 or CAFE program, but may
                perform research to understand better the relationship between A/C
                efficiency and thermal technologies for consideration in future
                rulemakings.
                3. Flexibilities for Off-Cycle Technologies
                 ``Off-cycle'' technologies are those that reduce vehicle fuel
                consumption and CO2 emissions in the real world, but for
                which the fuel consumption reduction benefits cannot be measured or
                cannot be fully measured under the 2-cycle test procedures (city,
                highway or correspondingly FTP, HFET) used to determine compliance with
                the fleet average standards. The CAFE city and highway test cycles,
                collectively referred to as the 2-cycle laboratory compliance tests (or
                2-cycle tests), were developed in the early 1970s. The city test
                simulates city driving in the Los Angeles area at that time. The
                highway test simulates driving on secondary roads (not expressways).
                The cycles are effective in measuring improvements in most fuel economy
                improving technologies; however, they are unable to measure or
                underrepresent certain fuel economy improving technologies because of
                limitations in the test cycles. For example, off-cycle technologies
                that improve emissions and fuel economy at idle (such as ``stop start''
                systems) and those technologies that improve fuel economy to the
                greatest extent at expressway speeds (such as active grille shutters
                which improve aerodynamics) receive less than their real-world benefits
                in the 2-cycle compliance tests.
                 Starting with MY 2008, EPA began employing a ``five-cycle'' test
                methodology to measure fuel economy for the purpose of improving new
                car window stickers (labels) and giving consumers better information
                about the fuel economy they could expect under real-world driving
                conditions.\3337\ However, for CO2 and CAFE compliance, EPA
                continues to use the established ``two-cycle'' test methodology.\3338\
                As learned through development of the ``five-cycle'' methodology and
                prior rulemakings, there are technologies that provide real-world
                CO2 emissions and fuel consumption improvements, but those
                improvements are not fully reflected on the ``two-cycle'' test. EPA
                established the off-cycle credit program to provide an appropriate
                level of CO2 credit for technologies that achieve
                CO2 reductions, but are normally not chosen as a
                CO2 control strategy because their CO2 benefits
                are not measured on the specified 2-cycle test.
                ---------------------------------------------------------------------------
                 \3337\ https://www.epa.gov/vehicle-and-fuel-emissions-testing/dynamometer-drive-schedules.
                 \3338\ The city and highway test cycles, commonly referred to
                together as the 2-cycle tests are laboratory compliance tests
                required by law for CAFE and are also used for determining
                compliance with the GHG standards.
                ---------------------------------------------------------------------------
                 Currently, EPA has three compliance pathways. The first approach
                allows manufacturers to gain credits without having to prove the
                benefits of the technologies on a case-by-case basis. A predetermined
                list or ``menu'' of credit values for specific off-cycle technologies
                exists and became effective starting in MY 2014.\3339\ This pathway
                allows manufacturers to use credit values established by EPA for a wide
                range of off-cycle technologies, with minimal or no data submittal or
                testing requirements.\3340\ Specifically, EPA established a menu with a
                number of technologies that have real-world CO2 and fuel
                consumption benefits not measured, or not fully measured, by the two-
                cycle test procedures, and those benefits were reasonably quantified by
                the agencies at that time. For each of the pre-approved technologies on
                the menu, EPA established a quantified default value that is available
                without additional testing. Manufacturers must demonstrate that they
                were in fact using the menu technology, but not required to conduct
                testing to quantify the technology's effects, unless they wish to
                receive a credit larger than the default value. The default values for
                these off-cycle credits were largely determined from research,
                analysis, and simulations, rather than from full vehicle testing, which
                would have been both cost and time prohibitive. EPA generally used
                conservative predefined estimates to avoid any potential credit
                windfall.\3341\
                ---------------------------------------------------------------------------
                 \3339\ See 40 CFR 86.1869-12(b).
                 \3340\ The Technical Support Document (TSD) for the 2012 final
                rule for MYs 2017 and beyond provides technology examples and
                guidance with respect to the potential pathways to achieve the
                desired physical impact of a specific off-cycle technology from the
                menu and provides the foundation for the analysis justifying the
                credits provided by the menu. The expectation is that manufacturers
                will use the information in the TSD to design and implement off-
                cycle technologies that meet or exceed those expectations in order
                to achieve the real-world benefits of off-cycle technologies from
                the menu.
                 \3341\ While many of the assumptions made for the analysis were
                conservative, others were ``central.'' For example, in some cases,
                an average vehicle was selected on which the analysis was conducted.
                In that case, a smaller vehicle may presumably deserve fewer credits
                whereas a larger vehicle may deserve more. Where the estimates are
                central, it would be inappropriate for the agencies to grant greater
                credit for larger vehicles, since this value is already balanced by
                smaller vehicles in the fleet. The agencies take these matters into
                consideration when applications are submitted for credits beyond
                those provided on the menu.
                ---------------------------------------------------------------------------
                 For off-cycle technologies not on the pre-defined technology list,
                or obtained through petitioning, EPA created a second pathway which
                allows manufacturers to use 5-cycle testing to demonstrate and justify
                off-cycle CO2 credits.\3342\ EPA established this
                alternative for a manufacturer to demonstrate the benefits of the
                technology using 5-cycle testing. The additional emissions tests allow
                emission benefits to be demonstrated over some elements of real-world
                driving not captured by the CO2 compliance tests, including
                high speeds, rapid accelerations, and cold temperatures. Under this
                pathway, manufacturers submit test data to EPA, and EPA determines
                whether there is sufficient technical basis to approve the off-cycle
                credits. No public comment period is required for manufacturers seeking
                credits using the EPA menu or using 5-cycle testing.
                ---------------------------------------------------------------------------
                 \3342\ See 40 CFR 86.1869-12(c). EPA proposed a correction for
                the 5-cycle pathway in a separate technical amendments rulemaking.
                See 83 FR 49344 (Oct. 1, 2019). EPA is not approving credits based
                on the 5-cycle pathway pending the finalization of the technical
                amendments rule.
                ---------------------------------------------------------------------------
                 The third pathway allows manufacturers to seek EPA approval,
                through a notice and comment process, to use an alternative methodology
                other than the menu or 5-cycle methodology for determining the off-
                cycle technology CO2 credits.\3343\ Manufacturers must
                provide supporting data on a case-by-case basis demonstrating the
                benefits of the off-cycle technology on their vehicle models.
                Manufacturers may also use the third pathway to apply for credits and
                FCIVs for menu technologies where the manufacturer is able to
                demonstrate credits and FCIVs greater than those provided by the menu.
                ---------------------------------------------------------------------------
                 \3343\ See 40 CFR 86.1869-12(d).
                ---------------------------------------------------------------------------
                 Due to the uncertainties associated with combining menu
                technologies and the fact that some uncertainty is introduced because
                off-cycle credits are provided based on a general assessment of off-
                cycle performance, as opposed to testing on the individual vehicle
                models, EPA established caps that limit the amount of credits a
                manufacturer may generate using the EPA menu. Off-cycle technology is
                capped at 10 grams/mile per year on a combined car and truck fleet-wide
                average basis. No caps were established for technologies gaining
                credits through the petitioning or 5-cycle approval methodologies.
                [[Page 25233]]
                a) Consideration of Eliminating A/C and Off-Cycle Adjustments in the
                CO2 and CAFE Programs
                 The agencies sought comments in the NPRM on whether to remove the
                A/C and off-cycle flexibilities from the CAFE program and adjust the
                stringency levels accordingly based upon concern that the flexibilities
                might distort the market. Several commenters provided responses
                concerning the feasibility of removing any of these flexibilities.
                Commenters included the Auto Alliance, the National Automobile Dealers
                Association, Global Automakers, the Alliance for Vehicle Efficiency,
                ACEEE, BorgWarner, Fiat Chrysler, General Motors, International Council
                on Clean Transportation, Toyota, and UCS. Other comments were received
                requesting that the agencies look into expanding the flexibilities by
                including more technologies.
                 There was widespread support from commenters for retaining these
                flexibilities for A/C and off-cycle technologies in the CO2
                and CAFE programs. Commenters preferred that the agencies continue to
                include the flexibilities, believing them to enable real world fuel
                economy improvements and compliance with CO2 and CAFE
                standards with a more cost effective combination of technologies. The
                agencies agree that these programs achieve real world fuel economy
                improvements and that keeping the flexibilities may enable more cost
                effective technology combinations to achieve those real world fuel
                economy improvements. For MY 2017, manufacturers introduced a wide
                variety of low-cost technologies through the A/C and off-cycle
                flexibilities that increased the overall industry's CAFE performance by
                1.1 mpg. The agencies also acknowledge that the continued use of these
                flexibilities under the EPA program since 2012 warrants consideration
                due to automakers' and suppliers' significant investments in developing
                the technologies, which could result in stranded capital should the
                agencies discontinue them and manufacturers choose to remove the
                technologies. For these reasons, the agencies have decided to continue
                allowing manufacturers to use the existing flexibilities for A/C
                efficiency and off-cycle technologies for future model years.
                b) Final Decisions in Response to Manufacturers' and Suppliers'
                Requests
                 Automakers, trade associations, and auto suppliers recommended
                several changes to the current off-cycle credit program.\3344\ Prior to
                the NPRM, automakers and suppliers suggested changes to the off-cycle
                program, including:
                ---------------------------------------------------------------------------
                 \3344\ See generally Alliance/Global Petition.
                ---------------------------------------------------------------------------
                 Streamlining the program in ways that would give auto
                manufacturers more certainty and make it easier for manufacturers to
                earn credits;
                 Expanding the current pre-defined off-cycle credit menu to
                include additional technologies and increasing credit levels where
                appropriate;
                 Eliminating or increasing the credit cap on the pre-
                defined list of off-cycle technologies and revising the thermal
                technology credit cap; and
                 Creating a role for suppliers directly to seek approval of
                their technologies.
                 EPA requested comments on several aspects of the off-cycle credits
                program and, as discussed below, both EPA and NHTSA are adopting some
                modest changes, primarily to help streamline and clarify their
                programs, and to ease the implementation burden for manufacturers and
                the government. The agencies are not adopting a significant expansion
                of the programs in this rule, as also discussed below. EPA and NHTSA
                are taking this relatively conservative approach for their off-cycle
                programs due to the uncertainty that remains in estimating off-cycle
                benefits of technologies and the need to remain cautious to help ensure
                that emissions and fuel economy benefits expected through the off-cycle
                flexibility are realized in the real-world.
                (1) Program Streamlining
                 EPA requested comments on changes to the off-cycle process that
                would streamline the program. Currently, under the third pathway,
                manufacturers submit an application that includes the methodology they
                used to determine the off-cycle credit value and data, which then
                undergoes a public notice and comment process prior to an EPA decision
                regarding the application. Each manufacturer separately submits an
                application to EPA that must undergo a public notice and comment
                process even if the manufacturer uses a methodology previously approved
                by EPA for another manufacturer. For example, under the current
                program, multiple manufacturers have separately submitted applications
                for high-efficiency alternators and advanced A/C compressors using
                similar methodologies and producing similar levels of credits. If
                manufacturers also seek fuel economy improvement values for the CAFE
                program, they are also required to send the submissions to NHTSA, as
                EPA consults with NHTSA in its determinations for the CAFE program.
                NHTSA's involvement is discussed in more detail in Section IX.D.3.b).
                 EPA requested comment on revising the regulations to allow all auto
                manufacturers to make use of a methodology once it has been approved by
                EPA under the public process, without subsequent applications from
                other manufacturers having to undergo the same process. This would
                reduce redundancy in the current program. Manufacturers would need to
                provide EPA with at least the same level of data and detail for the
                technology and methodology as the manufacturer that went through the
                initial public notice and comment process.
                 EPA received supportive comments for streamlining the approval
                process from auto manufacturers and suppliers. The Auto Alliance
                commented that it supports all actions that would shorten the time it
                takes EPA to evaluate and reach decisions on applications through the
                off-cycle alternative methodology pathway, and that manufacturers
                should be allowed to use common data from applications that have
                already been approved.\3345\ Such common data would include ambient
                conditions, general consumer behavior data, and general operating and
                performance data for the same off-cycle technologies. Global Automakers
                also commented that EPA should streamline efforts to avoid
                reduplication of applications in situations where multiple automakers
                have submitted petitions for the same technology and recommended
                blanket approval for applications using the same specific technologies
                and calculation and measurement procedures.\3346\ General Motors
                commented that when a credit for a new technology is approved for one
                manufacturer, the EPA decision document announcing that approval can
                serve as a guidance document that assigns a credit value or calculation
                methodology for the technology for all manufacturers without requiring
                duplicative testing.\3347\ MEMA commented that it would be sufficient
                to uphold the integrity of the off-cycle program to require the next
                vehicle manufacturer's application to provide at least the same level
                of data and details as the original vehicle manufacturer application
                and to validate the level of credit the next vehicle manufacturer is
                [[Page 25234]]
                applying for based on how the technology is applied in its fleet.\3348\
                ---------------------------------------------------------------------------
                 \3345\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3346\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
                12032.
                 \3347\ General Motors, Detailed Comments, NHTSA-2018-0067-11858-
                21.
                 \3348\ MEMA, EPA-HQ-OAR-2018-0283-5692. See https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf.
                ---------------------------------------------------------------------------
                 ACEEE commented that any streamlining of the process by which
                automakers petition for off-cycle credits must maintain the requirement
                that a thorough methodology show real-world benefits and ensure
                adequate opportunity for public review.\3349\ International Council on
                Clean Transportation (ICCT), while not commenting on this specific
                request for comment, commented that the program should remain unchanged
                until potential changes can be further analyzed.\3350\
                ---------------------------------------------------------------------------
                 \3349\ ACEEE, Detailed Comments, NHTSA-2018-0067-12122.
                 \3350\ International Council on Clean Transportation, Detailed
                Comments, NHTSA-2018-0067-11741.
                ---------------------------------------------------------------------------
                 After considering the comments, consistent with its request for
                comment, EPA is streamlining the approval process as follows: Once a
                methodology for a specific off-cycle technology has gone through the
                public notice and comment process and is approved for one manufacturer,
                other manufacturers may follow the same methodology to collect data on
                which to base their off-cycle credits. Once a methodology is approved,
                other manufacturers may submit applications citing the approved
                methodology, but those manufacturers must provide their own necessary
                test data, modeling, and calculations of credit value specific to their
                vehicles, and any other vehicle-specific details pursuant to that
                methodology, to assess an appropriate credit value. This is similar to
                what occurred, for example, with the advanced A/C compressor, where one
                manufacturer applied for credits with data collected through bench
                testing and vehicle testing and subsequent manufacturers applied for
                credits following the same methodology, but by submitting test data
                specific to their vehicle models. However, those subsequent
                applications previously required a public notice and comment process.
                For future applications, as long as the testing is conducted using the
                previously-approved methodology, EPA will evaluate the credit
                application and issue a decision with no additional notice and comment,
                since the first application that established the methodology was
                subject to notice and comment.
                 EPA is not providing blanket approval for a specific credit value,
                nor amending the requirement that manufacturers collect necessary data
                or perform modeling or other analyses on their specific vehicle models
                as the basis for the credit. However, once a methodology has been fully
                vetted and approved through the public process, EPA believes additional
                public review of the identical methodology is unnecessarily
                duplicative. In EPA's experience thus far (for example with high-
                efficiency alternators and advanced A/C compressors for which EPA has
                received applications from several manufacturers based on the same
                methodology), additional public review has yielded no additional
                substantive public comments. EPA believes this change in the program
                will help reduce the time necessary for review of applications. EPA
                will maintain the option to seek additional public comment in cases
                where the agency believes a new application deviates from a previously
                approved methodology or raises new issues on which the agency believes
                it is prudent to seek comment.
                 EPA also requested comment on revising the regulations to allow EPA
                to, in effect, add technologies to the pre-approved credit menu without
                going through a subsequent rulemaking. For example, if one or more
                manufacturers submit applications with sufficient supporting data for
                the same or similar technology, the data from that application(s) could
                potentially be used by EPA as the basis for adding technologies to the
                menu. EPA requested comment on revising the regulations to allow EPA to
                establish through a decision document a credit value, or scalable value
                as appropriate, and technology definitions or other criteria to be used
                for determining whether a technology qualifies for the new menu credit.
                As envisioned in the NPRM, this streamlined process of adding a
                technology to the menu would involve an opportunity for public review
                but not a formal rulemaking to revise the regulations, allowing EPA to
                add technologies to the menu in a timely manner, where EPA believes
                that sufficient data exist to estimate an appropriate credit level for
                that technology across the fleet.
                 EPA received supportive comments regarding this request for
                comments from auto manufacturers and suppliers who believe that the
                change would help streamline the program. EPA also received comments
                from environmental NGOs suggesting that the program should not be
                changed at this time. After consideration of these comments, the
                agencies are not revising the regulations to allow technologies to be
                added to the menu without a rulemaking because EPA believes that menu-
                based off-cycle credits should be based on a robust demonstration of
                the technology, consistent with the regulations. The agencies will
                retain the option to add technologies to the menu through a rulemaking,
                similar to the approach being taken for high-efficiency alternators and
                advanced A/C compressors as discussed below, where sufficient data has
                been collected from multiple manufacturers and vehicle models on which
                to base a menu credit. The menu credits are meant to be conservative.
                The agencies are concerned that basing a menu credit on data from only
                one or a few manufacturers does not guarantee a robust and accurate
                credit level representing vehicles across the fleet. At this time, the
                agencies continue to believe a rulemaking process with full opportunity
                for public comment remains the best approach for adding technologies to
                the menu. A rulemaking ensures that all stakeholders including
                automakers have an opportunity to provide data to support an
                appropriate and conservative credit level for the fleet. This approach
                also provides an incentive for manufacturers to, in the meantime,
                continue to perform testing and provide actual data that could
                eventually be used to inform a rulemaking process to add a technology
                to the menu. The agencies want to preserve that element of the program
                to maintain the integrity of off-cycle credits representing real-world
                reductions.
                (2) A/C and Off-Cycle Application Process
                 The agencies received several comments, in addition to those
                received in the petitions from the Auto Alliance and Global Automakers,
                discussed below, on the application process for approving additional A/
                C and off-cycle credits. Commenters included the Global Automakers, the
                Auto Alliance, Volkswagen, Edison Electric Institute, Ford, Fiat
                Chrysler, NCAT, Toyota, General Motors, and DENSO International.
                 Fiat Chrysler, Ford, Volkswagen, DENSO International, Global
                Automakers, and the Auto Alliance requested that the agencies respond
                more quickly to applications for A/C and off-cycle technologies.\3351\
                They
                [[Page 25235]]
                prefer that petitions be addressed before the close of a model year so
                manufacturers can have a better idea of what credits they will earn.
                ---------------------------------------------------------------------------
                 \3351\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943-
                50; Ford, Detailed Comments, NHTSA-2018-0067-11928-15; Volkswagen,
                Detailed Comments, NHTSA-2017-0069-0583-13; DENSO, Detailed
                Comments, NHTSA-2018-0067-11880-5; Global Automakers, Detailed
                Comments, NHTSA-2018-0067-12032-50; Auto Alliance, Detailed
                Comments, NHTSA-2018-0067-12073-120.
                ---------------------------------------------------------------------------
                 The agencies agree that responding to petitions before the end of a
                model year is beneficial to manufacturers and the government.
                Manufacturers would have a better idea of the approved credits, and the
                government could carry-out its compliance processes more efficiently.
                EPA structured the A/C and off-cycle programs to make it possible to
                complete the processes by the end of the model year so manufacturers
                could submit their final reports within the required deadline, 90 days
                after the calendar year. However, delays currently exist due to the
                timing needed to review and approve technologies for the first time and
                issue Federal Register notices seeking public comments, where
                applicable. The agencies anticipate these problems will resolve
                themselves as the off-cycle program reaches maturity and EPA initiates
                the new streamlining approaches adopted in this final rule, discussed
                in the previous section.
                 The agencies are also aware that delays exist because manufacturers
                frequently submit late applications, new applications, and ask for
                retroactive credits or FCIVs for off-cycle technologies equipped on
                previously-manufactured vehicles after the model year has ended. As
                required under both the CO2 and CAFE programs, manufacturers
                are to submit applications for off-cycle credits and FCIVs before the
                beginning of each compliance model year, to enable the agencies to make
                better informed final decisions before the model year ends.
                 To expedite the process of approvals, the agencies will enforce
                existing EPA and NHTSA regulations requiring manufacturers to notify
                and report information on the technologies before the beginning of the
                model year. Presently, manufacturers must notify EPA in their pre-model
                year reports, and in their applications for certification, of their
                intention to generate any A/C and off-cycle credits before the model
                year, regardless of the methodology for generating credits.\3352\
                Manufacturers choosing to generate credits using the alternative EPA-
                approval methodology are required to submit a detailed analytical plan
                to EPA prior to a model year in which a manufacturer intends to seek
                these credits. The manufacturer may seek EPA input on the proposed
                methodology prior to conducting testing or analytical work, and EPA
                will provide input on the manufacturer's analytical plan. The
                alternative demonstration program must be approved in advance by the
                Administrator. NHTSA has similar provisions for its projections reports
                in which detailed information on the technologies must be included in
                those submissions during the month of December before the model
                year.\3353\ NHTSA's provisions also require manufacturers to submit
                information to NHTSA at the same time as to EPA. Consequently, the
                eligibility of a manufacturer to gain off-cycle CO2 credits
                or CAFE adjustments for a given compliance model year requires
                appropriate submissions to the agencies. The agencies intend to enforce
                these provisions starting with the 2020 compliance model year.
                Manufacturers may resubmit MY 2020 information until May 1, 2020. After
                that time, the agencies will deny any manufacturers' late submissions
                requesting retroactive credits. However, manufacturers who properly
                submit information ahead of time will be allowed to make corrections to
                resolve inadvertent errors during or after the model year. The agencies
                believe that enforcing the existing submission requirements will be the
                most efficient approach to expedite approvals until new regulatory
                deadlines or additional requirements can be adopted.
                ---------------------------------------------------------------------------
                 \3352\ See 40 CFR 86.1869(a) and 40 CFR 1843-01.
                 \3353\ See 49 CFR part 537.7(c)(7) and 49 CFR part 531.6 and
                533.6.
                ---------------------------------------------------------------------------
                 Fiat Chrysler, Volkswagen, Global Automakers, and the Auto Alliance
                further suggested the EPA issue a Federal Register notice for submitted
                off-cycle applications within 30 days and issue a final decision within
                90 days.\3354\
                ---------------------------------------------------------------------------
                 \3354\ Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943;
                Volkswagen, Detailed Comments, NHTSA-2017-0069-0583; Global
                Automakers, Detailed Comments, NHTSA-2018-0067-12032; Auto Alliance,
                Detailed Comments, NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 As mentioned, EPA is addressing the issues raised by commenters by
                streamlining its required regulatory processes to eliminate the need to
                submit multiple Federal Register notices concerning requests from
                different manufacturers for the same technology. Under this streamlined
                process, after a technology is approved for the initial
                manufacturer(s), EPA will approve any subsequent manufacturer requests
                for the same technology upon receipt of data submissions validating the
                benefit specific to their model types.
                 General Motors, Toyota, NCAT, Fiat Chrysler, Ford, Volkswagen,
                DENSO, Edison Electric Institute, Global Automakers, and the Auto
                Alliance further suggested that technologies approved for multiple
                manufacturers, to the extent additional automakers will have the same
                requests, be added to the menu to encourage additional implementation
                of the technology. Doing so would reduce duplicative efforts for the
                agencies, as well as manufacturers.\3355\
                ---------------------------------------------------------------------------
                 \3355\ General Motors, Detailed Comments, NHTSA-2018-0067-11858;
                Toyota, Detailed Comments, NHTSA-2018-0067-12150; NCAT, Detailed
                Comments, NHTSA-2018-0067-11969; Fiat Chrysler, Detailed Comments,
                NHTSA-2018-0067-11943; Ford, Detailed Comments, NHTSA-2018-0067-
                11928; Volkswagen, Detailed Comments, NHTSA-2017-0069-0583; DENSO,
                Detailed Comments, NHTSA-2018-0067-11880; Edison Electric Institute,
                Detailed Comments, NHTSA-2018-0067-11918; Global Automakers,
                Detailed Comments, NHTSA-2018-0067-12032; Auto Alliance, Detailed
                Comments, NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 As mentioned previously, the agencies have decided to allow only
                new technologies to be added to the menu through the regular rulemaking
                processes including the opportunity for notice and public comment.
                 General Motors, DENSO, Global Automakers, and the Auto Alliance
                further suggested that suppliers should be allowed to request a ``grams
                per mile'' value for their off-cycle technologies. They asserted that
                this will provide certainty to manufacturers before they buy that
                technology.\3356\ Toyota and the Auto Alliance suggested that the
                agencies could improve efficiency and reduce burdens by creating a
                ``toolbox,'' methodologies that manufacturers can apply to the analysis
                of off-cycle credit opportunities.\3357\ They stated it would
                additionally help manufacturers if the agency would issue guidance
                letters and decision documents for off-cycle credit approvals.\3358\
                ---------------------------------------------------------------------------
                 \3356\ General Motors, Detailed Comments, NHTSA-2018-0067-11858;
                DENSO, Detailed Comments, NHTSA-2018-0067-11880; Global Automakers,
                Detailed Comments, NHTSA-2018-0067-12032; Auto Alliance, Detailed
                Comments, NHTSA-2018-0067-12073.
                 \3357\ Toyota, Detailed Comments, NHTSA-2018-0067-12150; Auto
                Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3358\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 The agencies believe that developing a ``toolbox'' may not be
                possible due to the development of new and emerging technologies, and
                manufacturers' different approaches for evaluating the benefits of the
                technologies. The agencies may consider additional guidance, if
                feasible, as the programs further matures in the approval process of
                technologies and if the agencies can identify consistent methodologies
                that may help manufacturers analyze off-cycle technologies.
                [[Page 25236]]
                 NCAT and General Motors requested more transparency in the A/C and
                off-cycle approval process. They suggested that the agencies could
                provide reports including off-cycle credits approved by vehicle make
                and model and provide further clarification of data requirements that
                influenced the decision process.\3359\
                ---------------------------------------------------------------------------
                 \3359\ NCAT, Detailed Comments, NHTSA-2018-0067-11969; General
                Motors, Detailed Comments, NHTSA-2018-0067-11858.
                ---------------------------------------------------------------------------
                 EPA and NHTSA have separate approaches for sharing information on
                these flexibilities, to provide public transparency. EPA already
                provides detailed information on manufacturers generation of A/C and
                off-cycle credits for each model year in its end of the year compliance
                report, including the magnitude of credits by manufacturer and by
                credit type, the credits generated by technology type, and the
                penetration of off-cycle technologies in each manufacturer's
                fleet.\3360\ NHTSA plans to share similar information on its PIC and to
                provide projected data on the market penetration rates of the
                technologies as soon as it starts receiving information through its new
                reporting templates for the 2023 compliance model year.
                ---------------------------------------------------------------------------
                 \3360\ ``The 2018 EPA Automotive Trends Report: Greenhouse Gas
                Emissions, Fuel Economy, and Technology since 1975,'' EPA-420-R-19-
                002. March 2019; Figures 5.8 through 5.12, and Tables 5.3 and 5.4.
                ---------------------------------------------------------------------------
                (3) High Efficiency Alternators and Advanced Air Conditioning (A/C)
                Compressors
                 EPA sought comments on modifying the off-cycle menu to add certain
                technologies for which EPA has collected sufficient data to set an
                appropriate credit level. More specifically, EPA received data from
                multiple manufacturers on high-efficiency alternators and advanced air
                conditioning (A/C) compressors that could serve as the basis for new
                menu credits for these technologies.\3361\ EPA requested comments on
                adding these two technologies to the menu including comments on credit
                level and appropriate definitions. EPA also requested comments on other
                off-cycle technologies that EPA could consider adding to the menu
                including supporting data that could serve as the basis for the credit.
                ---------------------------------------------------------------------------
                 \3361\ https://www.epa.gov/vehicle-and-engine-certification/compliance-information-light-duty-greenhouse-gas-ghg-standards.
                ---------------------------------------------------------------------------
                 EPA received only supportive comments on its specific request for
                comments regarding adding high efficiency alternators and advanced A/C
                compressors to the menu. Toyota, General Motors, BorgWarner, Fiat
                Chrysler, the Auto Alliance, Global Automakers, MECA, DENSO, SAFE, and
                Volkswagen submitted responses on the off-cycle menu. General Motors,
                Volkswagen, Fiat Chrysler, Global Automakers, and the Auto Alliance all
                supported adding high-efficiency alternators and advanced A/C
                compressors to the menu.\3362\ They commented that these technologies
                have already been approved for off-cycle credits through the petition
                process multiple times. They contend that it would be less burdensome
                if the technologies would be added to the pre-approved off-cycle credit
                list. That said, they were concerned about being constrained by the
                off-cycle caps.\3363\
                ---------------------------------------------------------------------------
                 \3362\ General Motors, Detailed Comments, NHTSA-2018-0067-11858;
                Volkswagen, Detailed Comments, NHTSA-2017-0069-0583; Fiat Chrysler,
                Detailed Comments, NHTSA-2018-0067-11943; Global Automakers,
                Detailed Comments, NHTSA-2018-0067-12032; Auto Alliance, Detailed
                Comments, NHTSA-2018-0067-12073.
                 \3363\ See, e.g., General Motors, Detailed Comments, NHTSA-2018-
                0067-11858.
                ---------------------------------------------------------------------------
                 The agencies believe that adding high-efficiency alternators and
                advanced A/C compressors to the menu is a reasonable step to help
                streamline the program by allowing manufacturers to select the menu
                credit rather than continuing to seek credits through the public
                approval process. Therefore, EPA is revising the regulations to add
                these two technologies to the menus. The high-efficiency alternator is
                being added to the off-cycle credits menu, and the advanced A/C
                compressor with a variable crankcase valve is being added to the menu
                for A/C efficiency credits. The credit levels are based on data
                previously submitted by multiple manufacturers through the off-cycle
                credits application process, and discussed in the NPRM. The high
                efficiency alternator credit is scalable with efficiency, providing an
                increasing credit value of 0.16 grams/mile CO2 per percent
                improvement as the efficiency of the alternator increases above a
                baseline level of 67 percent efficiency. The advanced A/C compressor
                credit value is 1.1 grams/mile for both cars and light trucks.\3364\
                ---------------------------------------------------------------------------
                 \3364\ For additional details regarding the derivation of these
                credits see EPA's Memorandum to Docket EPA-HQ-OAR-2018-0283
                (``Potential Off-cycle Menu Credit Levels and Definitions for High
                Efficiency Alternators and Advanced Air Conditioning Compressors'').
                ---------------------------------------------------------------------------
                 EPA also received comments from the Auto Alliance, Fiat Chrysler,
                General Motors, Mitsubishi, Gentherm, ITB, and MEMA on a variety of
                individual technologies that they suggest adding to the menu.\3365\
                These commenters provided little data to support their recommended
                credit levels. The Auto Alliance and Alliance for Vehicle Efficiency
                further asserted that flexibility mechanisms are increasingly important
                and there is a need to develop unconventional and non-traditional fuel
                economy technologies to meet standards.\3366\ They requested additional
                pre-defined and pre-approved technologies to be included in this
                regulation.\3367\
                ---------------------------------------------------------------------------
                 \3365\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073-
                48; Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943; General
                Motors, Detailed Comments, NHTSA-2018-0067-11858; Mitsubishi,
                Detailed Comments, NHTSA-2018-0067-12056; MEMA, Detailed Comments,
                MEMA, EPA-HQ-OAR-2018-0283-5692 (See https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20
                Appendices%20Oct%2026%202018.pdf); ITB, Detailed Comments, EPA-HQ-
                OAR-2018-0283-5469; Gentherm, Detailed Comments, EPA-HQ-OAR-2018-
                0283-5058.
                 \3366\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Alliance for Vehicle Efficiency, Detailed Comments, NHTSA-2018-0067-
                11696.
                 \3367\ NHTSA-2018-0067-12073-48.
                ---------------------------------------------------------------------------
                 The agencies have reviewed manufacturers' requests for adding
                additional technologies to the picklist and concluded that there is
                insufficient data in the record at this time on which to base an
                appropriate menu credit value for the technologies. Therefore, none of
                these technologies are being added to the menu at this time. Given the
                limited data and uncertainty, EPA also does not believe it would be
                appropriate to add any of the technologies to the menu without an
                opportunity for public review and comment. Although the agencies are
                not adding these technologies to the menu at this time, manufacturers
                may seek off-cycle credits for these technologies through the other
                program pathways.
                (4) Stop-Start Technology
                 In 2014, EPA approved additional credits for the Mercedes-Benz's
                stop-start system through the off-cycle credit process based on data
                submitted by Mercedes-Benz on fleet idle time and its system's real-
                world effectiveness (i.e., how much of the time the system turns off
                the engine when the vehicle is stopped).\3368\ Prior to proposal,
                multiple auto manufacturers requested that EPA revise the table menu
                value for stop-start technology based solely on one input value EPA
                considered, idle time, in the context of the Mercedes-Benz stop-start
                system. No manufacturers provided additional data on any of the other
                factors evaluated during consideration of a conservative credit value
                for stop-start systems. Stop-start systems vary significantly in
                hardware,
                [[Page 25237]]
                design, and calibration, leading to wide variations in the amount of
                idle time during which the engine is actually turned off in real-world
                driving. EPA has learned that some stop-start systems may be less
                effective in the real-world than the agency estimated in its 2012
                rulemaking analysis, for example, due to systems having a disable
                switch available to the driver, or because stop-start systems can be
                disabled under certain temperature conditions or auxiliary loads, which
                would offset the benefits of the higher idle time estimates. EPA
                requested additional data from manufacturers, suppliers, and other
                stakeholders regarding a comprehensive update to the stop-start off-
                cycle credit table value. EPA did not receive any additional real-world
                system effectiveness data from commenters on which to base an adjusted
                credit level. MEMA commented that EPA should base an increase in the
                credit on the agencies' updated estimated effectiveness of stop-start
                technology in the Draft Technical Assessment Report (TAR), which shows
                a 67 percent increase in effectiveness.\3369 3370\ However, EPA notes
                that this estimate is for system effectiveness over the 2-cycle test
                procedures and, therefore, is not an appropriate basis to adjust the
                off-cycle credits. The agencies are not adjusting the menu credits for
                stop-start systems at this time. Manufacturers may apply for additional
                credits if they are able to collect data demonstrating a system
                effectiveness that would serve as the basis for those credits.
                ---------------------------------------------------------------------------
                 \3368\ ``EPA Decision Document: Mercedes-Benz Off-cycle Credits
                for MY 2012-2016,'' EPA-420-R-14-025 (Sept. 2014).
                 \3369\ Draft Technical Assessment Report: Midterm Evaluation of
                Light-Duty Vehicle Greenhouse Gas Emission Standards and Corporate
                Average Fuel Economy Standards for Model Years 2022-2025, EPA-420-D-
                16-900 (July 2016).
                 \3370\ MEMA, EPA-HQ-OAR-2018-0283-5692. See https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf.
                ---------------------------------------------------------------------------
                (5) Menu Credit Cap
                 The off-cycle menu currently includes a fleetwide cap on credits of
                10 grams/mile to address the uncertainty surrounding the data and
                analysis used as the basis of the menu credits.\3371\ Prior to
                proposal, some stakeholders expressed concern that the current cap may
                constrain manufacturers' future ability to fully utilize the menu
                especially if the menu is expanded to include additional technologies,
                as described above. For example, Global Automakers suggested raising
                the cap from 10 grams/mile to 15 grams/mile.\3372\ EPA requested
                comments on increasing the current cap, for example, from the current
                10 grams/mile to 15 grams/mile to accommodate increased use of the
                menu. EPA also requested comment on a concept that would replace the
                current menu cap with an individual manufacturer cap that would scale
                with the manufacturer's average fleetwide target levels. The cap would
                be based on a percentage of the manufacturer's fleetwide 2-cycle
                emissions performance, for example at five to ten percent of
                CO2 of a manufacturer's emissions fleet-wide target. With a
                cap of five percent for a manufacturer with a 2-cycle fleetwide average
                CO2 level of 200 grams/mile, for example, the cap would be
                10 grams/mile.
                ---------------------------------------------------------------------------
                 \3371\ 40 CFR 86.1869-12(b)(2).
                 \3372\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
                12032.
                ---------------------------------------------------------------------------
                 There was widespread support from automakers and suppliers for
                removing the cap entirely or raising the cap from 10 grams/mile to 15-
                20 grams/mile. Toyota, General Motors, BorgWarner, Fiat Chrysler, the
                Auto Alliance, Global Automakers, MECA, DENSO, SAFE, and Volkswagen
                submitted responses on the off-cycle cap to EPA.\3373\ They argued that
                the 2-cycle test does not always account for all the benefits a
                technology provides.\3374\ General Motors, Fiat Chrysler, the Auto
                Alliance, Global Automakers, and Volkswagen agreed that EPA should
                remove the 10 grams/mile cap and, if they must keep the cap, increasing
                it to 15 grams/mile.\3375\
                ---------------------------------------------------------------------------
                 \3373\ Toyota, Detailed Comments, NHTSA-2018-0067-12150; General
                Motors, Detailed Comments, NHTSA-2018-0067-11858; BorgWarner,
                Detailed Comments, NHTSA-2018-0067-11895; Fiat Chrysler, Detailed
                Comments, NHTSA-2018-0067-11943; Auto Alliance, Detailed Comments,
                NHTSA-2018-0067-12073; Global Automakers, Detailed Comments, NHTSA-
                2018-0067-12032; MECA, Detailed Comments, NHTSA-2018-0067-11994;
                DENSO, Detailed Comments, NHTSA-2018-0067-11880; SAFE, Detailed
                Comments, NHTSA-2018-0067-11981; Volkswagen, Detailed Comments,
                NHTSA-2017-0069-0583.
                 \3374\ See, e.g., DENSO, Detailed Comments, NHTSA-2018-0067-
                11880.
                 \3375\ General Motors, Detailed Comments, NHTSA-2018-0067-11858;
                Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943; Auto
                Alliance, Detailed Comments, NHTSA-2018-0067-12073; Global
                Automakers, Detailed Comments, NHTSA-2018-0067-12032; Volkswagen,
                Detailed Comments, NHTSA-2017-0069-0583.
                ---------------------------------------------------------------------------
                 Global Automakers commented that, as more technology receives off-
                cycle credit values, the cap will restrict innovation and therefore EPA
                should lift the cap now in anticipation of increased use of
                technologies.\3376\ General Motors similarly commented that the cap was
                an arbitrary limit without any technical justification and that, if the
                agency was to add emission reduction technologies to the menu these
                devices could not be effectively incentivized if the 10 grams/mile cap
                remains in place, since there would be no room under the cap.\3377\
                General Motors suggested that as the program continues, manufacturers
                will continue to find new technologies and will be limited by the cap.
                They stated that the cap will stifle additional investments for
                technologies. MEMA commented that if EPA expands the off-cycle
                technologies menu and continually adds off-cycle technologies to the
                menu, it is critical that EPA increase or eliminate the cap on the
                credits gained from the off-cycle menu.\3378\
                ---------------------------------------------------------------------------
                 \3376\ Global Automakers, Detailed Comments, NHTSA-2018-0067-
                12032.
                 \3377\ General Motors, Detailed Comments, NHTSA-2018-0067-11858.
                 \3378\ MEMA, EPA-HQ-OAR-2018-0283-5692. See https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf.
                ---------------------------------------------------------------------------
                 The Auto Alliance argued that putting caps on emerging new
                technologies will hinder further vehicle investments and improvements.
                The planning cycle is implemented years out and without a guarantee
                they will see benefits, the Auto Alliance stated that manufacturers
                lack incentivization to work toward large technological advances.\3379\
                The Auto Alliance and Alliance for Vehicle Efficiency further asserted
                that flexibility mechanisms are increasingly important and there is a
                need to develop unconventional and non-traditional fuel economy
                technologies.\3380\
                ---------------------------------------------------------------------------
                 \3379\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3380\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Alliance for Vehicle Efficiency, Detailed Comments, NHTSA-2018-0067-
                11696.
                ---------------------------------------------------------------------------
                 ACEEE commented that the off-cycle credit menu cap should not be
                increased or modified without the agency first defining any other
                changes it might consider making to the off-cycle credit program and
                this should be done through a separate NPRM and public review
                process.\3381\ ICCT commented that if the agencies allow more use of
                off-cycle credits without clear validation of their real-world
                benefits, the regulations cannot serve their intended objectives to
                reduce CO2 and fuel use.\3382\
                ---------------------------------------------------------------------------
                 \3381\ ACEEE, Detailed Comments, NHTSA-2018-0067-12122.
                 \3382\ ICCT, Detailed Comments, NHTSA-2018-0067-11741-43.
                ---------------------------------------------------------------------------
                 EPA also received a few comments warning about the risks of
                removing the caps and over incentivizing the CAFE and CO2 programs.
                ACEEE pointed out that while expanding and updating the flexibilities
                that incentivize innovation
                [[Page 25238]]
                and research is a great method to increase fuel efficiency, it is
                important to put a time limit on those incentives and carefully design
                them so manufacturers do not take advantage. ACEEE argued that, if
                these flexibilities are not implemented thoughtfully, they can end up
                reducing the program benefits. UCS commented that, given the potential
                interaction from multiple incentives, it is important to consider the
                combined impacts of flexibilities on the overall stringency of the
                regulation. UCS stated that given the potential for widespread harm,
                credits within the program should be severely limited, and the
                agencies' assessment of the impacts of such incentives should be
                extremely conservative in order to promote increased environmental
                benefits of the fuel economy and carbon dioxide emissions
                standards.\3383\
                ---------------------------------------------------------------------------
                 \3383\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
                ---------------------------------------------------------------------------
                 The agencies are not increasing the 10 grams/mile menu credit cap
                at this time. EPA established the 10 grams/mile credit cap to address
                the uncertainty surrounding the data and analysis used as the basis of
                the menu credits, and agrees with ACEEE, ICCT, and UCS that sufficient
                uncertainty remains such that increasing the current cap is not
                justified. As noted in the 2012 final rule, EPA included the fleet-wide
                cap because the default credit values were based on limited data, and
                also because the agencies recognized that some uncertainty is
                introduced when credits are provided based on a general assessment of
                off-cycle performance as opposed to testing on the individual vehicle
                models.\3384\ That uncertainty has not significantly diminished since
                the 2012 final rule. Also, over the course of implementing the program,
                EPA has encountered issues with the regulatory definitions currently in
                place for some technologies. The regulations specify that manufacturers
                may claim credits for technologies that meet the regulatory
                definitions. However, there have been instances where manufacturers
                have claimed credits for a technological approach that they have argued
                meets the regulatory definition, but EPA found that the technology was
                not implemented consistent with the technological approach envisioned
                when the off-cycle program was established. This has raised questions
                of whether the credits for the technological approach in question truly
                represent real-world reductions, and whether the credits should
                ultimately be allowed. These types of issues have resulted in
                uncertainty, which can lead to delays in credit calculations,
                competitive inequities, as well as increased burden on the agency to
                review and resolve issues. The caps continue to serve as an important
                measure against the loss of emissions reductions and fuel savings given
                the uncertainty in the credit values as the program is implemented.
                Since the agencies are not expanding the menu beyond the two
                technologies discussed above, the agencies believe there remains enough
                room under the cap such that the menu may continue to serve its purpose
                as a source of off-cycle credits. Although a few manufacturers
                approached the cap limit in MY 2018, the fleet average menu credit was
                4.7 grams/mile, less than half the cap value.\3385\ If the agencies
                undertake a rulemaking in the future to modify the menu or regulatory
                definitions, the agencies may re-evaluate the cap levels at that time.
                The agencies note that the cap only applies to credits based on the
                menu. Under the current program, manufacturers may apply for credits
                beyond the cap through other available pathways based on a
                demonstration of off-cycle technology emission reduction data for their
                fleets.
                ---------------------------------------------------------------------------
                 \3384\ 77 FR 62834 (Oct. 15, 2012).
                 \3385\ The 2018 EPA Automotive Trends Report, Greenhouse Gas
                Emissions, Fuel Economy, and Technology since 1975, EPA-420-R-19-002
                (Mar. 2019).
                ---------------------------------------------------------------------------
                 As noted above, the agencies have decided to continue the option to
                add technologies to the menu only through the rulemaking process and,
                for this final rule, have decide to add two new menu items; one for
                high-efficiency alternators and another for advanced A/C compressors.
                The agencies stated that they will only add technologies when
                sufficient data has been collected from multiple manufacturers and
                vehicle models on which to base a menu credit. Accordingly, the
                agencies believe this approach ensures that conservative, robust and
                accurate credit levels are being added representing vehicles ``on
                average'' across the fleet.
                 Finally, NHTSA has been studying how the combination of
                flexibilities and incentives may adversely affect the stringency of the
                CAFE regulations. NHTSA is aware of an instance in which combining
                incentives for alternative fueled vehicles and adjustments for A/C and
                off-cycle technologies allowed one manufacturer to increase in CAFE
                fleet performance to a combined average of 516.8 mpg for MY 2017, a
                curious result. NHTSA iscontinuing to evaluate the issue of combining
                incentives and flexibilities and may address this issue further in the
                future.
                (6) Eligibility
                 Though, in the NPRM, EPA did not explicitly request comment on the
                eligibility criteria for determining what technologies are eligible for
                off-cycle credits, EPA received comments on this topic. UCS commented
                that regulations should be clarified so that the program does not
                result in unwarranted credits for baseline technologies, noting that in
                the 2012 final rule EPA stated that technologies integral or inherent
                to the basic vehicle design were not eligible for credits and
                specifically excluded technologies identified by the agency as
                technologies a manufacturer may use to meet the two-cycle
                CO2 standards.\3386\ ACEEE commented that off-cycle credits
                should be limited to new and innovative technologies and, that to be
                eligible for credit, a technology must reduce emissions from the
                vehicle receiving the credit (as opposed to other vehicles on the road,
                for example, through system effects of technologies designed for crash
                avoidance or improving traffic flow).\3387\ The Auto Alliance also
                commented in the area of eligibility, suggesting regulatory changes
                that would allow off-cycle credits for any technology where the
                manufacturer could demonstrate an off-cycle emissions benefit.\3388\
                The Auto Alliance commented that the program is intended to provide
                credit for technologies that provide more fuel economy and
                CO2 emissions reduction benefit in the real-world than is
                realized in FTP and HFET on-cycle testing and that a baseline
                technology should be eligible for such credits.
                ---------------------------------------------------------------------------
                 \3386\ UCS, Detailed Comments, NHTSA-2018-0067-12039.
                 \3387\ ACEEE, Detailed Comments, NHTSA-2018-0067-12122.
                 \3388\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 Given the various public comments on eligibility of technologies
                for off-cycle credits, the agencies are clarifying the regulations
                regarding technology eligibility, consistent with the intent and EPA's
                interpretation of the 2012 rule, as expressed in the preamble to the
                proposed and final rules. The agencies believe that clarifying the
                regulations will reduce confusion among manufacturers as to what
                technologies are eligible and reduce the overall program burden
                associated with EPA staff giving continued guidance to manufacturers
                regarding eligibility, as detailed in the 2012 rule preamble.
                Eligibility was thoroughly addressed in the 2012 final rule preamble,
                but the regulations were not as clear, which has led to confusion on
                the part of some manufacturers and delays in reviewing
                [[Page 25239]]
                credit applications.\3389\ The agencies are not establishing a new
                policy regarding eligibility, only amending the language reflecting the
                existing policy in the regulations for sake of clarity.
                ---------------------------------------------------------------------------
                 \3389\ 77 FR 62726-36, 62835-37.
                ---------------------------------------------------------------------------
                 As noted in the 2012 final rule preamble, the goal of the off-cycle
                credits program is to provide ``an incentive for the development and
                use of additional technologies to achieve real-world reductions in
                CO2 emissions.'' \3390\ EPA further stated that the intent
                of the program is to ``provide an incentive for CO2 and fuel
                consumption reducing off-cycle technologies that would otherwise not be
                developed because they do not offer a significant 2-cycle benefit.''
                \3391\ The regulation at 40 CFR 86.1869-12(a) provides that
                manufacturers may generate credits for CO2 reducing
                technologies ``where the CO2 reduction benefit for the
                technology is not adequately captured on the Federal Test Procedure
                and/or Highway Fuel Economy Test.'' The regulation continues: ``[t]hese
                technologies must have a measurable, demonstrable, and verifiable real-
                world CO2 reduction that occurs outside the conditions of
                the Federal Test Procedure and the Highway Fuel Economy Test.''
                ---------------------------------------------------------------------------
                 \3390\ 77 FR 62833.
                 \3391\ 77 FR 62836.
                ---------------------------------------------------------------------------
                 Off-cycle credits are available for technologies that are not
                utilized when performing FTP and HFET tests because their operation is
                linked to a condition not found during the 2-cycle testing. For
                example, heating and cooling systems are not operated during the 2-
                cycle test, and therefore, efficiency improvements to these systems are
                not captured at all on the 2-cycle tests. As the 2012 rule's language
                indicates, off-cycle credits are not necessarily limited to
                technologies listed on the menu or off-cycle technologies with no
                measurable benefit on the FTP and/or HFET. Off-cycle credits may be
                available for some technologies whose performance is measurable to some
                extent on the FTP and/or HFET but which perform measurably better off-
                cycle. Active aerodynamic and stop-start technologies (menu item) are
                examples. However, there are limits on what the agencies would consider
                to be an off-cycle technology eligible for credits, as discussed below.
                 Just as the regulations and preamble to the 2012 final rule listed
                technologies that the agencies considered to be off-cycle technologies,
                the preamble also discussed technologies that the agency would not
                consider off-cycle technologies--i.e., technologies the agencies
                consider to be ``adequately captured'' by the FTP and therefore not
                eligible for off-cycle credits. The preamble specifically noted that
                engine, transmission, mass reduction, passive aerodynamic design, and
                base tire technologies are not considered to be off-cycle technologies
                eligible for credits.\3392\ These are technologies that are considered
                to be ``integral or inherent to basic vehicle design.'' \3393\ In
                response to comments in the final rule, the agencies further clarified
                that advanced combustion concepts, such as camless engines, variable
                compression ratio engines, micro air/hydraulic launch assist devices,
                would not be considered to be eligible for credits.\3394\ This
                limitation to eligibility further extends to other engine designs,
                transmission designs, and electrification systems not specifically
                contemplated in the rulemaking, such as Atkinson combustion engines,
                and 9 and 10 speed transmissions, as well as to other hybrid systems
                such as 48 Volt technologies. Further, the 2012 final rule preamble
                stated that technologies included in the agencies' assessment for
                purposes of developing the standard would not be allowed to generate
                off-cycle credits and cites the technologies described in Chapter 3 of
                the 2012 final rule TSD.\3395\ Finally, off-cycle credits are not
                available for technologies required to be used by Federal Law or for
                crash avoidance systems, safety critical systems, or technologies that
                may reduce the frequency of vehicle crashes.\3396\
                ---------------------------------------------------------------------------
                 \3392\ 77 FR 62732, 62836.
                 \3393\ 77 FR 62732, 62836/1; 81 FR 73499.
                 \3394\ 77 FR 62732.
                 \3395\ 77 FR 62836.
                 \3396\ 40 CFR 86.1869-12(a); 77 FR 62836.
                ---------------------------------------------------------------------------
                 The preamble to the 2012 final rule provides the rationale for what
                the agency considers an off-cycle technology and, therefore, eligible
                for credits. Technologies that are integral or inherent to the vehicle
                are, by necessity, well represented on the 2-cycle test.\3397\ Examples
                provided in the preamble are engine, transmission, mass reduction,
                passive aerodynamic design, and base tire technologies. The control
                logic for these powertrain components, like the components themselves
                (i.e. engine and transmission), are constantly active, fully
                functioning, and operating over the entirety of the FTP and HFET.
                Similarly, an automatic transmission, regardless of whether it has 6-
                speeds or 8-speeds, would still be constantly active, fully functioning
                and operating over the entirety of the FTP and HFET.\3398\ This would
                also be true for base engine technologies, advanced combustion
                concepts, engine components (pistons, valves, camshafts, crankshafts,
                oil pumps, etc.), and driveline components (individual components of
                the transmission, axle, and differential).\3399\
                ---------------------------------------------------------------------------
                 \3397\ 77 FR 62732, 62836.
                 \3398\ 76 FR 75024 (Dec. 1, 2011).
                 \3399\ 77 FR 62732/2.
                ---------------------------------------------------------------------------
                 Further, even if these technologies have greater benefits on
                supplemental test cycles, EPA has explained that it would be difficult
                to devise accurate A/B testing (i.e., with and without the technology)
                for these technologies.\3400\ The 2012 preamble states that ``EPA is
                limiting the off-cycle program to technologies that can be identified
                as add-on technologies conducive to A/B testing,'' partly because it
                would be very difficult accurately to parse out the off-cycle benefits
                for some integral technologies.\3401\ Because the technology is
                integral to the vehicle, there would not be an appropriate baseline
                (i.e., without the technology) vehicle to use for comparison. Vehicles
                are not built without tires, engines, passive aerodynamics or
                transmissions.
                ---------------------------------------------------------------------------
                 \3400\ 76 FR 75024.
                 \3401\ 77 FR 62836.
                ---------------------------------------------------------------------------
                 Also, because these technologies are inherent to the vehicle
                design, their performance is already reflected in the stringency of the
                standard and giving credits for these inherent technologies would be a
                type of double-counting windfall.\3402\ ``[S]ince these methods are
                integral to basic vehicle design, there are fundamental issues as to
                whether they would ever warrant off-cycle credits. Being integral,
                there is no need to provide an incentive for their use, and (more
                importantly), these technologies would be incorporated regardless.
                Granting credits would be a windfall.'' \3403\ As such, EPA has laid
                out a clear basis that technological improvements to integral and
                inherent components are considered to be adequately captured on the FTP
                and HFET test.
                ---------------------------------------------------------------------------
                 \3402\ 77 FR 62732.
                 \3403\ See also 76 FR 75024.
                ---------------------------------------------------------------------------
                 EPA is clarifying the regulations in a manner that is consistent
                with the intent and our interpretation of the 2012 rule, as expressed
                in the preambles to the proposed and final rules. The regulations are
                revised to specify that technologies used primarily to meet the 2-cycle
                standards are not eligible for off-cycle credits and that only
                technologies primarily installed for reducing off-cycle emissions would
                be eligible. The revised regulations specify that the technologies must
                not be integral or inherent to the basic vehicle design, such as, for
                example, engine,
                [[Page 25240]]
                transmission, mass reduction, passive aerodynamic design, and tire
                technologies. Exceptions to these general provisions include
                technologies already specified on the menu, including engine idle stop-
                start, active aerodynamic improvements, and high-efficiency
                alternators. These technologies may provide some benefit on the 2-cycle
                test, but EPA determined in the 2012 rule that they are eligible for
                off-cycle credits because they are technologies that could be added to
                vehicles to provide discernable off-cycle reductions.
                 Regulatory text at 40 CFR 86.1869-12(a) states: ``Manufacturers may
                generate credits for CO2 reducing technologies where the
                CO2 reduction benefit of the technology is not adequately
                captured on the Federal Test Procedure and/or the Highway Fuel Economy
                Test,'' to which EPA is adding, ``such that the technology would not be
                otherwise installed for purposes of reducing emissions (directly or
                indirectly) over those test cycles (i.e., on-cycle) for compliance with
                the [CO2] standards.'' EPA is also adding text to this
                paragraph of the regulations specifying: ``The technologies must not be
                integral or inherent to the basic vehicle design, such as engine,
                transmission, mass reduction, passive aerodynamic design, and tire
                technologies. Technologies installed for non-off-cycle emissions
                related reasons are also not eligible as they would be considered part
                of the baseline vehicle design. The technology must not be inherent to
                the design of occupant comfort and entertainment features except for
                technologies related to reducing passenger A/C demand and improving A/C
                system efficiency. Notwithstanding the provisions of this paragraph
                (a), off-cycle menu technologies included in paragraph (b) of this
                section remain eligible for credits.''
                 The agencies believe the above regulatory changes will help reduce
                confusion over what technologies are eligible for off-cycle credits,
                refocusing the program on technologies that manufacturers would install
                on vehicles for purposes of reducing off-cycle emissions rather than
                obtaining additional credits for technologies installed primarily for
                2-cycle emissions reduction or for other reasons not related to
                emissions. This approach is consistent with the intent of the program
                as stated in the 2012 final rule to provide an incentive to develop and
                employ off-cycle technologies not adequately captured on the 2-cycle
                test procedure.
                 Of the technologies recommended by manufacturers to be added to the
                menu, cooled EGR is an example of a technology that would not be
                eligible because it is an integral 2-cycle technology that EPA noted in
                its technology assessment in the MY 2012 rule. Cooled EGR is often an
                integral component of turbo charged gasoline direct injection engines
                which is a primary CO2 reduction strategy used by
                manufacturers to reduce 2-cycle emissions. The technologies are
                calibrated to act as a system such that is not possible to separate
                them in a way that would allow for a clear indication of the off-cycle
                benefit of cooled EGR as a stand-alone technology.
                 EPA also received comments from the Auto Alliance regarding several
                technologies they believe should qualify as active warm-up off-cycle
                technologies. The Auto Alliance commented that systems that use waste
                heat from the exhaust gas stream should receive additional credits
                beyond the menu credits currently established for active engine and
                transmission warm-up.\3404\ However, when EPA established the menu
                credits for active transmission and engine warm-up in the 2012 rule,
                EPA envisioned waste heat from the exhaust as the primary source of
                heat to quickly bring the system to operating temperature as the basis
                for the warm-up technology credits.\3405\ Therefore, EPA does not
                believe additional credits, as suggested by the Auto Alliance, are
                warranted. EPA further notes that the definitions for active engine and
                transmission warm-up specify that ``waste heat'' be used in active
                warm-up technologies in order to qualify for the credits.\3406\ If a
                system first directs heat to warm the engine oil or warm the interior
                cabin, and only then to the engine or transmission, thereby delaying
                active warm-up, EPA would not view that heat as waste heat since it is
                serving other purposes during initial vehicle warm-up. EPA would also
                not consider this approach to be warming up the engine or transmission
                ``quickly'' due to the potentially significant delay in warm-up
                activation. In developing the active warm-up credits, EPA focused on
                systems using heat from the exhaust as a primary source of waste heat
                because that heat would be available quickly and also be exhausted by
                the vehicle and otherwise unused.
                ---------------------------------------------------------------------------
                 \3404\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3405\ See Joint Technical Support Document: Final Rulemaking
                for 2017-2025 Light-Duty Vehicle Greenhouse Gas Emission Standards
                and Corporate Average Fuel Economy Standards, EPA-420-R-12-901,
                August 2012, p. 5-96--5-100.
                 \3406\ 40 CFR 86.1869-12(b)(4)(v) and (vi).
                ---------------------------------------------------------------------------
                 EPA allowed for the possible use of other sources of heat such as
                coolant as the basis for credits as long as those methods would
                ``provide similar performance'' as extracting the heat directly from
                the exhaust system.\3407\ However, EPA may require manufacturers to
                demonstrate that the system is based on ``waste heat'' or heat that is
                not being preferentially used by the engine or other systems to warm-up
                other areas like engine oil or the interior cabin. Systems using waste
                heat from the coolant do not qualify for credits if their operation
                depends on, and is delayed by, engine oil temperature or interior cabin
                temperature. As the engine and transmission components are warming up,
                the engine coolant and transmission oil do not have any `waste' heat
                available for warming up anything else on the vehicle. During engine
                and transmission warm-up, the only waste heat source in a vehicle with
                an internal combustion engine is the engine exhaust as the transmission
                and coolant have not reached warmed-up operating temperature and
                therefore do not have any heat to share. Conserving heat in a
                transmission is not a rapid transmission warm-up using waste heat.
                Unless the component with lubricating oil and coolant is operating at
                its fully warmed-up design temperature, by EPA's definition, that
                component does not have any waste heat available for transfer from the
                lubricating oil or coolant to any other device until it has reached its
                fully warmed-up operating temperature (i.e. the temperature when the
                cooling system is enabled). A qualifying system may involve a second
                cooling loop that operates independent of the primary coolant system
                and is not dependent on or otherwise delayed by, for example, cabin
                temperature. Evaluating whether such systems qualify for menu credits
                often requires additional information regarding system design to
                understand better how the system uses waste heat. Given the complexity
                of these systems and the need to sometimes consider the details of how
                a system operates, EPA is not making any changes to the menu regarding
                warm-up technologies.
                ---------------------------------------------------------------------------
                 \3407\ See Joint Technical Support Document: Final Rulemaking
                for 2017-2025 Light-Duty Vehicle Greenhouse Gas Emission Standards
                and Corporate Average Fuel Economy Standards, p. 5-99, EPA-420-R-12-
                901, August 2012.
                ---------------------------------------------------------------------------
                 The Auto Alliance further commented that active transmission bypass
                valves should qualify for active transmission warm-up credits.\3408\
                The Auto Alliance
                [[Page 25241]]
                commented that traditional transmission oil coolers are always active
                and sized for extreme or worst-case hot ambient conditions. The coolers
                will, in colder ambient conditions, keep the transmission temperatures
                well outside of their most efficient operating range. The bypass valve
                circumvents the cooler when the transmission is relatively cold
                preserving the transmission heat, so the transmission warms more
                quickly. EPA disagrees that this type of approach should be eligible
                for active transmission warm-up because it does not use waste heat to
                add heat to the transmission. Instead, it prevents useful heat already
                present in the transmission from being unnecessarily removed. Also, EPA
                does not view this type of bypass valve as an off-cycle technology but
                rather as part of a good engineering design of a transmission cooler
                system. Many vehicles already are designed with transmission cooler
                bypass valves. EPA does not believe existing coolers qualify as warm-up
                technologies simply because they are disabled under cold conditions.
                This approach does not represent the addition of a new off-cycle warm-
                up technology but the disabling of an existing cooling technology.
                ---------------------------------------------------------------------------
                 \3408\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 Although the agencies did not consider changes to the program to
                allow credits for safety-related technologies and autonomous vehicle
                technologies in the proposal, comments were received both in favor of
                and not in favor of allowing such credits.\3409\ The agencies note that
                the rationale for not allowing off-cycle credits for safety-related or
                crash avoidance technologies has not changed since the 2012 rule and,
                therefore, in the proposed rule the agencies did not consider making
                any changes to allow off-cycle credits for safety-related
                technologies.\3410\ The agencies continue to believe that there is a
                very significant distinction between technologies providing direct and
                reliably quantifiable improvements to fuel economy and CO2
                emission reductions, and technologies which provide those improvements
                by indirect means, where the improvement is not reliably quantifiable,
                and may be speculative (or in many instances, non-existent), or may
                provide benefit to other vehicles on the road more than for themselves.
                The agencies also continue to believe that the advancement of crash-
                related and crash avoidance systems specifically is best left to
                NHTSA's exercise of its vehicle safety authority.
                ---------------------------------------------------------------------------
                 \3409\ See, e.g., SAFE, Detailed Comments, NHTSA-2018-0067-
                11981; AAA, Detailed Comments, NHTSA-2018-0067-11979.
                 \3410\ 77 FR 62733.
                ---------------------------------------------------------------------------
                 Auto manufacturers and suppliers also commented that EPA should
                adopt ``eco-innovation'' credits approved in the European Union (EU)
                vehicle CO2 reduction program as part of the off-cycle
                credits program.\3411\ No data was provided as to why the credits would
                be appropriate for the U.S. vehicle fleet. EPA did not consider or
                request comment on the EU credits program and does not believe the
                credit levels would necessarily be appropriate for the U.S. fleet given
                the very different vehicle use and driving patterns between Europe and
                the U.S. Thus, there is no assurance that the credits would be based on
                real-world emissions reductions.
                ---------------------------------------------------------------------------
                 \3411\ See, e.g., Mitsubishi, Detailed Comments, NHTSA-2018-
                0067-12056.
                ---------------------------------------------------------------------------
                 EPA received comments from the Auto Alliance and Global Automakers
                that EPA should automatically award credits if the agency does not take
                final action within 90 days of receiving a request for credits.\3412\
                Regarding these comments, EPA does not believe such a provision is in
                keeping with maintaining the integrity of the off-cycle credits
                program. As discussed above, EPA often requires time to sort through
                complex issues to determine if the technologies meet the regulatory
                requirements for receiving credits and whether the credits have been
                quantified appropriately. In some instances, EPA has received public
                comments and manufacturer rebuttals to those comments that takes
                additional time to consider before making a final decision. EPA's goal
                continues to be to evaluate applications for credits in as timely a
                manner as is possible given the issues that must be addressed and
                within the resources available. While EPA's need carefully to consider
                applications may slow down the approval process or result in credits
                not being approved, it remains paramount to ensure credits are not
                provided to technologies that do not provide actual off-cycle benefits,
                and thereby do not meet the regulations. In the past, longer time
                frames for EPA review have not caused manufacturers to lose credits
                where credits are determined by EPA to be warranted under the
                regulations. EPA believes that the changes EPA is making to the program
                will help streamline the program and reduce confusion, thus helping to
                reduce the time necessary to evaluate applications and provide final
                decisions to manufacturers.
                ---------------------------------------------------------------------------
                 \3412\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Global Automakers, Detailed Comments, NHTSA-2018-0067-12032.
                ---------------------------------------------------------------------------
                (7) Supplier Role in the Off-Cycle Credits Program
                 Prior to proposal, EPA heard from many suppliers and their trade
                associations about an interest in allowing suppliers to have a formal,
                regulatorily defined role in the off-cycle credits program.\3413\ EPA
                requested comment on providing a pathway for suppliers, along with at
                least one auto manufacturer partner, to submit off-cycle applications
                for EPA approval. As described in the proposal, under such an approach,
                an application submitted by a supplier and vehicle manufacturer would
                establish a credit and/or methodology for demonstrating credits that
                all auto manufacturers could then use in their subsequent applications.
                EPA requested comment on requiring that the supplier be partnered in a
                substantive way with one or more auto manufacturers to ensure that
                there is a practical interest in the technology prior to EPA investing
                resources in the approval process. The supplier application would be
                subject to public review and comment prior to an EPA decision. However,
                once approved, subsequent auto manufacturer applications requesting
                credits based on the supplier methodology would not be subject to
                public review. Under this concept, the credits would be available
                provisionally for a limited period of time, allowing manufacturers to
                implement the technology and collect data on their vehicles in order to
                support a continuation of credits for the technology in the longer
                term. Also, as envisioned by EPA in its request for comment, the
                provisional credits could be included under the menu credit cap since
                they would be based on a general analysis of the technology rather than
                manufacturer-specific data.
                ---------------------------------------------------------------------------
                 \3413\ 83 FR 43461.
                ---------------------------------------------------------------------------
                 Auto manufacturers' and suppliers' comments were generally
                supportive of an expanded role for suppliers in the off-cycle credit
                program. The Auto Alliance supported allowing a supplier to lead the
                application process but did not support the provisional credit concept
                since the follow-up testing conducted by manufacturers may not support
                the level of credits initially claimed by the supplier, resulting in a
                lower than anticipated credit.\3414\ Instead, the Auto Alliance
                suggested a separate cap for supplier-based credits and noted that
                manufacturers could submit their own data if they wanted to pursue
                credits levels that exceeded the cap. General Motors similarly
                disagreed with the provisional credits that might
                [[Page 25242]]
                be rescinded if subsequent testing does not fully validate the value of
                the technology.\3415\ MEMA supported the request for comments regarding
                a supplier-led process but did not support requiring that suppliers
                have an auto manufacturer partner.\3416\ MEMA commented that there
                would be no incentive for a supplier to go through the product/
                technology development process, collect the necessary data, and
                undertake the full application process for a product/technology that
                would not generate manufacturer interest.
                ---------------------------------------------------------------------------
                 \3414\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3415\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                 \3416\ MEMA, EPA-HQ-OAR-2018-0283-5692. See https://www.mema.org/sites/default/files/resource/MEMA%20CAFE%20and%20GHG%20Vehicle%20Comments%20FINAL%20with%20Appendices%20Oct%2026%202018.pdf.
                ---------------------------------------------------------------------------
                 At this time, EPA believes additional discussions with interested
                parties and an opportunity for public comment, both of which are beyond
                the scope of this rulemaking, are needed. EPA continues to believe such
                an approach could encourage the further development of off-cycle
                technologies, but must be done in a reasonable way that ensures the
                credits are based on real-world emissions reductions.
                 Under the approach suggested by the Auto Alliance, manufacturers
                could claim supplier-based credits indefinitely and EPA might never
                receive any manufacturer data substantiating the credits unless that
                data supported a credit that exceeded the level established through the
                supplier process. EPA is concerned such a one-way ratchet approach
                could result in the loss of emissions benefits and undermine the
                integrity of the off-cycle credit program. EPA also remains concerned
                about the potential for a significantly increased volume of credit
                applications, including the potential for applications for proposed
                technologies that manufacturers might in reality have no interest in
                adopting. EPA understands MEMA's perspective on the issue of requiring
                a manufacturer partner, but a supplier-only process would potentially
                open the door to many requests such that the agency would need to
                expend considerable additional resources. EPA notes that nothing in the
                current regulations prevents collaboration between manufacturers and
                suppliers. Suppliers can initiate this process; manufacturer
                participation will be necessary to complete an application. EPA will
                provide additional clarity about this process through a subsequent
                technical amendments rulemaking.
                (8) Other Considerations
                 Avista Oil commented that EPA should provide an opportunity for
                credits based on the use of recycled engine oil. Avista Oil commented
                that there are CO2 emissions reductions associated with the
                use of recycled used engine oil and that vehicle manufacturers should
                be awarded credits for the use of recycled oil. Avista Oil's comment is
                not within the scope of the rulemaking. The off-cycle credits program
                focuses on providing credits for technologies that, when applied to the
                vehicle, the result is lower quantifiable real-world emissions from the
                vehicle. According to Avista Oil's comment, their recycled oil
                technology benefits are associated with the recycling process rather
                than lowering vehicle emissions on the road. Therefore, EPA would not
                view the technology as eligible for off-cycle credits, and EPA did not
                propose any other credit specific to the use of recycled engine oil.
                 Several commenters recommended that EPA raise the credit caps and
                credit values for thermal controls based on recent work by the National
                Renewable Energy Lab (NREL). Commenters suggested that credit values
                should be raised by 64 percent. In response, as discussed in the
                preamble, EPA is retaining the current menu credit caps and menu credit
                values due to uncertainties involved with the emissions projections and
                estimated credit values. Manufacturers may generate additional credits
                through the off-cycle credits program using the other two pathways by
                providing individual vehicle data. EPA recognizes additional modeling
                analysis has been performed by NREL that indicates the potential
                benefit of all thermal technologies including glazing. EPA designed the
                thermal control program and related caps based on previous NREL work
                and applied the thermal caps at the current levels to account for the
                wide range of uncertainties--including the uncertainty of the benefit
                from the combination of thermal technologies and the uncertainty
                highlighted by the different credit levels across the NREL studies. EPA
                believes the separate current thermal menu program cap and AC
                efficiency program cap continue to be reasonable for application across
                the fleet given these uncertainties.
                 Enhanced Protective Glass Automotive Association (EPGAA) and Vitro
                commented that the regulations established by the 2012 rule included an
                oversight in defining the baseline Tts (the metric used to evaluate
                thermal reflectivity of glass). EPGAA commented that there was an
                omission in the case of trucks, where the regulations do allow the use
                of privacy glass in locations other than the windshield and the front
                doors. The commenter discussed that the reference baseline glass for
                trucks, SUVs, and CUVs should have already included privacy glass for
                some of the rearward windows. In response, EPA recognized when the
                thermal credit program was finalized in 2012 that some of the vehicles
                within the reference fleet upon which the credits were based were
                already composed of vehicles with this type of thermal reflective
                glass. However, the agency found it difficult to estimate what portion
                of the fleet contained privacy glass and what the Tts rating was for
                privacy glass across the fleet. Because of this lack of specificity in
                the fleet composition and glass ratings, the agencies determined that
                the most appropriate approach was to allow credit for any glass meeting
                the finalized Tts requirements, and the total thermal cap was designed
                to account for this and other uncertainties.
                 Ford and others commented that thermal control technology credit
                caps should be implemented on a fleet average basis rather than on a
                ``per VIN'' basis. These commenters argued that the per VIN basis
                creates a reporting burden that is misaligned with the current
                reporting structure and creates program complexity and unnecessary
                workload. In response, EPA continues to believe that applying the
                thermal control credit cap on a per vehicle (per VIN) basis is
                appropriate due to the synergistic effects among these technologies.
                The CO2 reduction potential of applying thermal control technologies is
                limited within any given vehicle. The program has been implemented in
                this manner since MY2014, and manufacturers have in fact reported the
                necessary information to generate thermal control credits.
                 Gentherm, GM, MEMA, and The ITB Group commented that cooled seats
                should be added to the menu based on the approved GM off-cycle credits
                application and NREL study. EPA and NHTSA are not adding cooled seat
                technology to the menu because the agencies have received data from
                only a single manufacturer. By contrast, for the technologies EPA and
                NHTSA are adding to the menu in this final rule, the agencies have
                assessed data from multiple manufacturers. EPA notes however that the
                streamlining provisions being finalized in this action should
                facilitate other manufacturers in being able to apply for off-cycle
                credits by using GM's methodology.
                 Finally, on October 1, 2018, EPA proposed a technical correction
                separate from the SAFE Vehicles rulemaking for
                [[Page 25243]]
                the off-cycle credits pathway based on 5-cycle testing (83 FR 49344).
                This proposal would correct an error in the regulations established as
                part of the 2012 final rule. Some commenters expressed their support
                for the correction as part of their SAFE Vehicles rule comments. EPA
                notes that this correction continues to be part of a separate
                rulemaking and is not being addressed in the SAFE Vehicles final rule.
                c) Final Decisions on the 2016 Alliance/Global Petition
                (1) Retroactive A/C and Off-Cycle CAFE Adjustments
                 In 2016, the Alliance and Global submitted a petition for
                rulemaking, which included requests that: (1) NHTSA allow retroactive
                credits for A/C and off-cycle incentives for MYs 2012 to 2016; and (2)
                NHTSA and EPA revisit the average A/C efficiency benefit calculated by
                EPA applicable to MYs 2012 through 2016. The Alliance/Global argued
                that A/C efficiency improvements were not properly acknowledged in the
                CAFE program, and that manufacturers had exceeded the A/C efficiency
                improvements estimated by the agencies. The petitioners requested that
                EPA also amend its regulations such that manufacturers would be
                entitled to additional A/C efficiency improvement benefits
                retroactively. The petitioners also argued that NHTSA incorrectly
                stated the agency had taken off-cycle adjustments into consideration
                when setting standards for MYs 2017 through 2025, but not for MYs 2010-
                2016. The Alliance/Global further contended that because neither NHTSA
                nor EPA considered off-cycle adjustments in formulating the stringency
                of the MY 2012-2016 standards, NHTSA should retroactively grant
                manufacturers off-cycle adjustments for those model years as EPA did.
                Doing so, they said, would maintain consistency between the agencies'
                programs.
                 Of the two agencies, EPA was the first to establish an off-cycle
                technology program. For MYs 2012 through 2016, EPA allowed
                manufacturers to request off-cycle credits for ``technologies that
                achieve [CO2] reductions that are not reflected on current
                test procedures . . .'' \3417\ In the subsequent MY 2017 and later
                rulemaking, NHTSA joined EPA and included an off-cycle program for CAFE
                compliance. The Alliance/Global petition cited a statement in the MYs
                2012-2016 final rule as affirmation that NHTSA took off-cycle
                adjustments into account in formulating the MYs 2012-2016 stringencies,
                and therefore should allow manufacturers to earn off-cycle benefits in
                model years that have already passed.
                ---------------------------------------------------------------------------
                 \3417\ 75 FR 25341, 25344 (May 7, 2010). EPA had also provided
                an option for manufacturers to claim ``early'' off-cycle credits in
                the 2009-2011 time frame.
                ---------------------------------------------------------------------------
                 In the NPRM, NHTSA tentatively decided to retain the structure of
                the existing A/C efficiency program and not extend it to MYs 2010
                through 2016. For the rulemaking for MYs 2012 through 2016, NHTSA
                determined it was unable to consider improvements manufacturers made to
                passenger car A/C efficiency in calculating CAFE
                compliance.3418 3419 However, EPA did consider passenger car
                improvements to A/C efficiency for that timeframe. To allow
                manufacturers to build one fleet that complied with both EPA and NHTSA
                standards, the CAFE and CO2 standards were offset to account
                for the differences borne out of A/C efficiency improvements.
                Specifically, the agencies converted EPA's grams/mile standards to
                NHTSA mpg (CAFE) standards. EPA then estimated the average amount of
                improvement manufacturers were expected to earn via improved A/C
                efficiency. From there, NHTSA took EPA's converted mpg standard and
                subtracted the average improvement attributable to improvement in A/C
                efficiency. NHTSA set its standard at this level to allow manufacturers
                to comply with both standards with similar levels of technology.\3420\
                ---------------------------------------------------------------------------
                 \3418\ At that time, NHTSA stated ``[m]odernizing the passenger
                car test procedures, or even providing similar credits, would not be
                possible under EPCA as currently written.'' 75 FR 25557 (May 7,
                2010).
                 \3419\ 74 FR 49700 (Sept. 28, 2009).
                 \3420\ Id.
                ---------------------------------------------------------------------------
                 Likewise, EPA tentatively decided in the NPRM not to modify its
                regulations to change the way to account for A/C efficiency
                improvements. EPA believed this was appropriate as manufacturers
                decided what fuel economy-improving technologies to apply to vehicles
                based on the standards as finalized in 2010.\3421\ This included
                deciding whether to apply traditional tailpipe technologies, A/C
                efficiency improvements, or both. Granting A/C efficiency adjustments
                to manufacturers retroactively could result in arbitrarily varying
                levels of adjustments granted to manufacturers, similar to the
                Alliance/Global request regarding retroactive off-cycle adjustments.
                Thus, the existing A/C efficiency improvement structure for MYs 2010
                through 2016 would remain unchanged.
                ---------------------------------------------------------------------------
                 \3421\ In the MY 2017 and later rulemaking, NHTSA reaffirmed its
                position it would not extend A/C efficiency improvement benefits to
                earlier model years. 77 FR 62720 (Oct. 15, 2012).
                ---------------------------------------------------------------------------
                 NHTSA also tentatively decided manufacturers should not be granted
                retroactive off-cycle adjustments for MYs 2010 through 2016, and
                presented a number of clarifications to justify the denial. In
                particular, Alliance/Global pointed to a general statement where NHTSA,
                while discussing consideration of ``the effect of other motor vehicle
                standards of the Government on fuel economy,'' stated that that
                rulemaking resulted in consistent standards across the program.\3422\
                The Alliance/Global petition took this statement as a blanket assertion
                that NHTSA's consideration of all ``relevant technologies'' included
                off-cycle technologies. To the contrary, as quoted above, NHTSA
                explicitly stated it had not considered these off-cycle
                technologies.\3423\
                ---------------------------------------------------------------------------
                 \3422\ Id.
                 \3423\ Likewise, EPA stated it had not considered off-cycle
                technologies in finalizing the MYs 2012-2016 rule. ``Because these
                technologies are not nearly so well developed and understood, EPA is
                not prepared to consider them in assessing the stringency of the
                CO2 standards.'' Id. at 25438.
                ---------------------------------------------------------------------------
                 The fact that NHTSA had not taken off-cycle adjustments into
                consideration in setting its MYs 2012-2016 standards makes granting the
                Alliance/Global request inappropriate. Doing so could result in a
                question as to whether the MY 2012-2016 standards were maximum feasible
                under 49 U.S.C. 32902(b)(2)(B). If NHTSA had considered industry's
                ability to earn off-cycle adjustments--an incentive that allows
                manufacturers to utilize technologies other than those that were being
                modeled as part of NHTSA's analysis--the agency might have concluded
                more stringent standards were maximum feasible. Additionally, granting
                off-cycle adjustments to manufacturers retroactively raises questions
                of equity. NHTSA issued its MYs 2012-2016 standards without an off-
                cycle program, and manufacturers had no reason to anticipate that NHTSA
                would allow the use off-cycle technologies to meet fuel economy
                standards. Therefore, manufacturers made fuel economy compliance
                decisions with the expectation that they would have to meet fuel
                economy standards using on-cycle technologies. Generating off-cycle
                adjustments retroactively would arbitrarily reward some (and
                potentially disadvantage other) manufacturers for compliance decisions
                they made without the knowledge such technologies would be eligible for
                NHTSA's off-cycle program. Thus, NHTSA tentatively decided to deny
                Alliance/Global's request for retroactive off-cycle adjustments.
                [[Page 25244]]
                 It is worth noting that in the MYs 2017 and later rulemaking, NHTSA
                and EPA did include off-cycle technologies in establishing the
                stringency of the standards. As Alliance/Global noted, NHTSA and EPA
                limited their consideration to stop-start and active aerodynamic
                features because of limited technical information on these
                technologies.\3424\ At that time, the agencies stated they ``have
                virtually no data on the cost, development time necessary,
                manufacturability, etc. [sic] of these technologies. The agencies thus
                cannot project that some of these technologies are feasible within the
                2017-2025 timeframe.'' \3425\
                ---------------------------------------------------------------------------
                 \3424\ Alliance/Global Petition at 7.
                 \3425\ Draft Joint Technical Support Document: Rulemaking for
                2017-2025 Light-Duty Vehicle Greenhouse Gas Emission Standards and
                Corporate Average Fuel Economy Standards (November 2011), p. 5-57.
                ---------------------------------------------------------------------------
                 As described above, NHTSA first allowed manufacturers to generate
                off-cycle technology fuel consumption improvement values equivalent to
                CO2 off-cycle credits in MY 2017.\3426\ In finalizing the
                rule covering MYs 2017 and later, NHTSA declined to retroactively
                extend its off-cycle program to apply to model years 2012 through
                2016,\3427\ explaining ``NHTSA did not take [off-cycle credits] into
                account when adopting the CAFE standards for those model years. As
                such, extending the credit program to the CAFE program for those model
                years would not be appropriate.'' \3428\
                ---------------------------------------------------------------------------
                 \3426\ 77 FR 62840 (Oct. 15, 2012).
                 \3427\ See id.; EPA decided to extend provisions from its MY
                2017 and later off-cycle program to the 2012-2016 model years.
                 \3428\ Id.
                ---------------------------------------------------------------------------
                 In the NPRM, NHTSA and EPA sought any further comments on the
                tentative denials of the retroactive requests in the Alliance/Global.
                The Auto Alliance and Fiat Chrysler provided additional comments on the
                tentative denial of the petition requests from the Alliance/Global. The
                commenters cited that the widening gap between the regulatory standards
                and actual industry-wide new vehicle average fuel economy that has
                become evident since 2016, despite the growing use of improvement
                ``credits'' from various flexibility mechanisms, such as off-cycle
                technology credits, mobile air conditioner efficiency credits, mobile
                air conditioner refrigerant leak reduction credits and credits from
                electrified vehicles.\3429\ The commenters believe that applying
                retroactive credits for the new flexibilities for MYs 2012 to 2016 can
                address the current compliance deficiencies.
                ---------------------------------------------------------------------------
                 \3429\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073;
                Fiat Chrysler, Detailed Comments, NHTSA-2018-0067-11943.
                ---------------------------------------------------------------------------
                 Upon consideration of the issue, NHTSA is finalizing its decision
                to deny any retroactive off-cycle adjustments in the CAFE program for
                MYs 2012-2016. As mentioned in the NPRM, NHTSA is concerned about the
                negative impact of allowing retroactive credits, which could undermine
                the stringency of the MYs 2012-2016 standards. EPA is finalizing its
                decision not to modify its regulations to change the benefits for A/C
                efficiency improvements. As mentioned by EPA, the current approach
                creates uniformity and objectivity in determining A/C efficiency
                benefits. Consequently, because EPA is maintaining the current A/C
                determination methodology and NHTSA already considered those A/C
                adjustments in its MYs 2012-2016 CAFE standards, NHTSA is also
                finalizing its decisions in this rule to deny any retroactive A/C
                adjustments in the CAFE program for MYs 2012-2016.
                (2) Petition Requests on A/C Efficiency and Off-Cycle Program
                Administration
                 As discussed above, NHTSA and EPA jointly administer the off-cycle
                program. The 2016 Alliance/Global petition requested that EPA and NHTSA
                make various adjustments to the off-cycle program; specifically, the
                petitioners requested that the agencies should:
                 re-affirm that technologies meeting the stated definitions
                are entitled to the off-cycle credit at the values stated in the
                regulation;
                 re-acknowledge that technologies shown to generate more
                emissions reductions than the pre-approved amount are entitled to
                additional credit;
                 confirm that technologies not in the null vehicle set but
                which are demonstrated to provide emissions reductions benefits
                constitute off-cycle credits; and
                 modify the off-cycle program to account for unanticipated
                delays in the approval process by providing that applications based on
                the 5-cycle methodology are to be deemed approved if not acted upon by
                the agencies within a specified timeframe (for instance 90 days),
                subject to any subsequent review of accuracy and good faith.\3430\
                ---------------------------------------------------------------------------
                 \3430\ Alliance/Global Petition at 20.
                ---------------------------------------------------------------------------
                 With respect to Alliance/Global's request regarding off-cycle
                technologies that demonstrate emissions reductions greater than what is
                allowable from the menu, this final rule retains that capability. As
                was the case for MYs 2017-2021, a manufacturer may still apply for
                FCIVs and CO2 credits beyond the values listed on the menu,
                provided the manufacturer demonstrates the CO2 and fuel
                economy improvement.\3431\ This includes the two-alternative processes
                for demonstrating CO2 reductions and fuel economy
                improvement for gaining benefits using either the 5-cycle or
                alternative approval methodologies.\3432\
                ---------------------------------------------------------------------------
                 \3431\ 77 FR 62837 (Oct. 15, 2012).
                 \3432\ 40 CFR 86.1869-12.
                ---------------------------------------------------------------------------
                 The agencies have considered Alliance/Global's requests to
                streamline aspects of the A/C efficiency and off-cycle programs in
                response to the issues outlined above. Among other things, Alliance/
                Global requested that the agencies consider providing for a default
                acceptance of petitions for off-cycle credits after a specified period
                of time, provided that all required information has been provided, to
                accelerate the processing of off-cycle credit requests. While the
                agencies agree with the merits of A/C efficiency and off-cycle
                programmatic improvements, there are significant concerns with the
                concept of approving petition requests by default because such requests
                may not address program issues like uncertainty in quantifying program
                benefits, or general program administration.
                 Based on its consideration of the issues raised by the Alliance/
                Global, EPA has adopted in this final rule new processes for
                streamlining the compliance mechanisms for approving off-cycle and
                applications as discussed in the preceding section.
                (3) Other EPA Responses to Alliance Requests
                 One issue raised in the Alliance/Global Automakers June 2016
                petition (item 6 titled ``Refrain from Imposing Unnecessary
                Restrictions on the Use of Credits'') for EPA's consideration concerns
                how credits are managed within the CO2 program. The Alliance
                and Global Automakers suggested that EPA allow more flexibility in
                using credits generated under the various credit programs such as air
                conditioning or off-cycle credits by allowing them to be carried
                forward or back independently. Under this approach, a manufacturer
                would be allowed, for example, to carry their air conditioning credits
                back to cover a previous deficit while running a deficit in a current
                model year. The Alliance referred to this petition request in their
                comments, noting they believe the request ``remains pertinent in the
                context of this rulemaking.''
                 In response, EPA did not raise this issue or any related
                programmatic changes in the proposal and therefore
                [[Page 25245]]
                these comments are not within the scope of the rulemaking. EPA notes
                the GHG and CAFE programs are harmonized on the aggregation of credits.
                 The automakers' petition also requested that EPA correct the
                multiplier equation in the regulations so that manufacturers may
                generate the intended number of credits (item 8, ``Correct the
                Multiplier for BEVs, PHEVs, FCVs, and CNGs''). This request concerns an
                error in the regulations established in the 2012 Final Rule that
                results in manufacturers generating fewer than intended for MY 2017-
                2021 vehicles in some cases. In October 2018, in response to this
                petition request, EPA issued a proposed rule separate from the SAFE
                Vehicles NPRM to correct the error in the previously established
                regulations. EPA will continue to address this issue and related
                comments in that separate rulemaking. CAFE does not include multiplier
                credits and therefore this is not a harmonization issue.
                4. Specialty Vehicles With Low Mileage (SVLM)
                 In response to the NPRM, Volkswagen submitted comments seeking to
                adopt a new flexibility for specialty vehicles with low mileage
                (SVLM).\3433\ The flexibility would apply to specialty vehicles
                produced at low volumes and produced for infrequent use. They argued
                these specialty vehicles do not approach the vehicle miles traveled of
                typical vehicles. They requested that NHTSA and EPA allow the SVLM
                flexibility for vehicles that demonstrate limited predicted driving
                use. The flexibility would allot each manufacturer a limited annual
                production of 5,000 SVLM vehicles. It was also proposed that, within
                this limited product volume, each SVLM would retain its footprint
                derived performance target (per model type), but would utilize a
                modified VMT for determining any credits or debits associated with the
                performance of these vehicles within the manufacturer's fleet.
                ---------------------------------------------------------------------------
                 \3433\ Volkswagen, Detailed Comments, NHTSA-2017-0069-0583.
                ---------------------------------------------------------------------------
                 The agencies have considered the request from Volkswagen for
                credits or debits and fuel economy adjustments for SVLM vehicles and
                are denying the request. NHTSA notes that Congress prescribed
                alternative (reduced) CAFE standards for low-volume manufacturers,
                codified in 49 CFR part 525. Low-volume manufacturers' vehicles are
                often high-end sports cars and are not typically driven by their owners
                for long distances. Congress limited this exemption under the CAFE
                program to manufacturers of fewer than 10,000 passenger
                automobiles.\3434\ EPA has a similar program for smallvolume
                manufacturers which are defined as manufacturers with average sales for
                the three most recent consecutive model years of less than 5,000
                vehicles.\3435\ The flexibility proposed by Volkswagen would presumably
                be in addition to these existing provisions, but Volkswagen does not
                identify a source of authority for it. The agencies also have a number
                of questions about how specifically a SVLM concept might be
                implemented, such as whether every manufacturer would simply identify
                the 5,000 vehicles with the lowest projected VMT or lowest fuel economy
                and therefore qualify for credits for 5,000 vehicles every model year,
                or whether there should be additional criteria for vehicles to be
                included. The NPRM did not seek comment on a SVLM concept and the
                agencies did not receive other comments on the requested program.
                Therefore, the agencies are not adopting the SVLM concept suggested by
                Volkswagen.
                ---------------------------------------------------------------------------
                 \3434\ 49 U.S.C. 32902(d)(1).
                 \3435\ 40 CFR 86.1818-12(g).
                ---------------------------------------------------------------------------
                E. CO2 and CAFE Compliance Issues Not Addressed in the NPRM
                1. CO2 and CAFE Adjustments for 5-Cycle Testing
                 EPA and NHTSA received several comments requesting that the
                agencies revise current CAFE test procedures to use EPA's 5-cycle test
                procedures in place of the 2-cycle test procedures that have been
                largely unchanged since the inception of the CAFE program, or offset
                measured 2-cycle test fuel economy and CO2 emissions for
                CO2 and CAFE compliance. Walter Kreucher commented ``some
                technologies (Hybrid Electric) have penalties on the road that are not
                reflected on the tests used to determine CAFE compliance. . . . If the
                Agencies want to provide adjustment factors for A/C and other `Off-
                Cycle' conditions it must do so in both the positive and negative
                direction'' (sic).\3436\ AVE commented that the agencies should use 5-
                cycle procedures rather than 2-cycle procedures, arguing that the 5-
                cycle model better demonstrates real-world driving conditions and would
                lead to a more simplified credit allocation system.\3437\ BorgWarner
                echoed those comments, stating that the 5-cycle test is more accurate
                than the 2-cycle test and would reduce the need for credit
                adjustments.\3438\ Jeremy Michalek commented that the fuel economy
                values the public sees reflected on vehicles for purchase (e.g., on the
                Monroney label or in new car advertising) is calculated from the 5-
                cycle test; updating the 2-cycle test to capture more of the vehicle's
                fuel efficiency factors would allow for better consistency and a more
                accurate fuel efficiency measure.\3439\ The Auto Alliance proposed that
                the EPA revise its methodology for calculating off-cycle improvements
                when using the 5-cycle methodology by subtracting the 2-cycle benefit
                from the 5-cycle benefit to ensure credits are calculated
                properly.\3440\
                ---------------------------------------------------------------------------
                 \3436\ Walter Kreucher, Detailed Comments, NHTSA-2018-0067-0444.
                 \3437\ AVE, Detailed Comments, NHTSA-2018-0067-11696.
                 \3438\ BorgWarner, Detailed Comments, NHTSA-2018-0067-11895.
                 \3439\ Jeremy Michalek, et al., Detailed Comments, NHTSA-2018-
                0067-11903.
                 \3440\ Auto Alliance, Detailed Comments, NHTSA-2018-0067-12073.
                ---------------------------------------------------------------------------
                 The NPRM did not seek comment on revising compliance test
                procedures to use 5-cycle test procedures in place of 2-cycle test
                procedures, either entirely or broadly. Such a change would require
                extensive assessment and analysis to consider how changes could be
                implemented and what standards might be maximum feasible for CAFE and
                appropriate and reasonable for CO2 for new test procedures.
                There has been no analysis conducted to estimate the impacts of such a
                change on the levels of the standards. Therefore, making these
                requested changes is outside the scope of this rulemaking.
                2. National Zero Emissions Vehicle Concept
                 Although the agencies did not discuss or request comment on a
                National Zero Emissions Vehicle (NZEV) program concept, several
                organizations commented on that topic. Some discussed ideas from a task
                force that was formed by the governors of nine States who signed a
                memorandum of understanding (MOU) committing to undertake joint
                cooperative actions to build a robust market for ZEVs under their
                individual state programs. Collectively, these States have committed to
                having at least 3.3 million ZEVs operating on their roadways by 2025.
                ZEVs include battery-electric vehicles (BEVs), plug-in hybrid electric
                vehicles (PHEVs), and hydrogen fuel-cell electric vehicles (FCEVs).
                Comments on an NZEV concept were received from General Motors, CARB,
                Edison Electric Institute, Honda, NCAT, Workhorse Group, and Volvo.
                [[Page 25246]]
                 General Motors offered comments supporting an NZEV program, stating
                that it continues to expect California to be the leader of the EV
                market but hopes a national effort will be put forth, making the U.S. a
                global leader in EV technology development and deployment. \3441\
                General Motors stated it believes an NZEV program would further U.S.
                national security interests, make the U.S. more competitive with China,
                which already has an NZEV program, and reduce U.S. dependence on
                foreign petroleum. General Motors requested that EPA incentivize EV
                deployment, including providing credits for autonomous EVs and EVs that
                are used in rideshare programs.\3442\ General Motors outlined their
                proposed NZEV program which would include increasing ZEV requirements
                annually, establishing credit banks for manufacturers based on national
                ZEV sales, and ZEV multipliers for vehicles over 5,250 lbs., autonomous
                vehicles using EV, and EVs in rideshare programs. General Motors also
                proposed that requirements would be revisited if EV battery cell were
                not available at the costs Argonne National Lab forecasts by 2025.
                General Motors also suggested implementing a Zero Emissions Task Force
                that would promote complementary policies. General Motors acknowledged
                that the NZEV program would have to be subject to acceleration or delay
                depending on how quickly technologies are incentivized like battery
                cost.
                ---------------------------------------------------------------------------
                 \3441\ General Motors, Detailed Comments, NHTSA-2018-0067-11858.
                 \3442\ General Motors, Detailed Comments, NHTSA-2018-0067-11858.
                ---------------------------------------------------------------------------
                 CARB recommended a national ZEV multiplier, stating that a national
                incentive would help ensure ZEVs and PHEVs were being produced for sale
                beyond the ten States that have ZEV programs.\3443\ The Edison Electric
                Institute supported increasing stringency of fuel economy and
                CO2 standards and incorporating policies from ZEV States to
                create a ``One National Program.'' \3444\ Workhorse Group commented
                that a national ZEV mandate, where agencies progressively increase the
                mandated percentage of electric vehicles in every fleet, merits serious
                consideration by the agencies. They contended that an NZEV would have
                to work with the current State ZEV mandates and not preempt the
                progress already made.\3445\ Volvo, and Honda were proponents of
                incorporating ZEV standards into a national program. Volvo requested
                nationwide credits for ZEVs since there are 40 States without ZEV
                mandates.\3446\ Honda mentioned that incorporating California's ZEV
                credits into the national program would reduce compliance costs for
                manufacturers while incentivizing technological development.\3447\ NCAT
                recommended in their comment that EPA provide enhanced credits for EVs,
                PHEVs, and FCVs that are more stringent than California (and other
                States) ZEV mandates, making the national program credits
                ``additional'' to state ZEV compliance credits.\3448\
                ---------------------------------------------------------------------------
                 \3443\ CARB, Detailed Comments, NHTSA-2018-0067-11873.
                 \3444\ Edison Electric Institute, Detailed Comments, NHTSA-2018-
                0067-11918.
                 \3445\ Workhorse Group, Detailed Comments, NHTSA-2018-0067-
                12215.
                 \3446\ Volvo, Detailed Comments, NHTSA-2018-0067-12036.
                 \3447\ Honda, Detailed Comments, NHTSA-2018-0067-11818.
                 \3448\ NCAT, Detailed Comments, NHTSA-2018-0067-11969.
                ---------------------------------------------------------------------------
                 Northeast States for Coordinated Air Use Management (NESCAUM)
                commented that an aggressive reduction in emissions will not occur
                without national ZEV standards which will drive development of advanced
                clean vehicle technologies.\3449\
                ---------------------------------------------------------------------------
                 \3449\ NESCAUM, Detailed Comments, NHTSA-2018-0067-11691.
                ---------------------------------------------------------------------------
                 The NPRM did not propose or request comment on an NZEV concept or
                program, as such, and establishing such a program would be outside the
                scope of this rulemaking. Such a concept would require thorough
                assessment and full rulemaking notice and comment. There are also
                policy questions about what the appropriate level of potential
                incentives should be and whether certain technologies should receive
                greater incentives than other technologies, and if so, on what basis
                and by what amounts. Also, for the CAFE program, incentives for
                technologies are almost entirely prescribed by statute, and there are
                questions about how the CAFE program could implement an NZEV program in
                alignment with EPCA and EISA. Therefore, the agencies have decided not
                to implement an NZEV program as part of this rulemaking.
                3. CO2 In-Use Requirements
                 Current in-use regulations outlined in 86.1845-04 provide
                flexibility in determining the applicable number of test vehicles per
                test group. Each large volume manufacturer is provided the flexibility
                to employ small volume sampling allowances for a limited number of
                total annual production units. In response to the NPRM, Volkswagen is
                proposing to modify 86.1845-04 to provide a separate, additional small
                volume sampling allowance allocation of annual production volume for a
                manufacturer's plug-in hybrid vehicles. This additional allowance would
                only be applicable through the 2025 model year and would only be
                applicable to CO2 testing requirements under the in use
                regulations.
                 The basis for this flexibility is rooted in the continuing
                evolution and development of traction drive battery cell chemistries
                and battery management systems. This ongoing development is aimed at
                continuously improving such features as energy density, power, cost,
                and durability. As such, the engineering processes for understanding
                and quantifying long-term performance are still developing and subject
                to reevaluation as new chemistries are examined. Manufacturers such as
                Volkswagen have allocated significant capital in battery testing to
                ensure that performance is maintained for consumers and are also
                providing longer term battery warranty provisions.
                 Volkswagen believes that the targeted flexibility will provide
                additional time to continue evaluating chemistries and reduce
                administrative testing burdens for a very limited production allocation
                per manufacturer. This provision will further support plug-in hybrid
                technology development and deployment. Volkswagen proposed modifying
                86.1845-04 table SO4-07 footnote 2, to read as follows:
                 \2\ Total annual production of groups eligible for testing under
                small volume sampling plan is capped at a maximum of 14,999 vehicle 49
                or 50 state annual sales, or a maximum of 4,500 vehicle California only
                sales per model year, per large volume manufacturer. Through model year
                2025, a separate total annual production of plug-in hybrid electric
                vehicle groups shall be eligible for testing under small volume
                sampling plan as described above. This allocation shall only be
                applicable to exhaust CO2 emission standards under this
                subpart.\3450\
                ---------------------------------------------------------------------------
                 \3450\ See EPA-HQ-OAR-2018-0283-5689-A1, p.32.
                ---------------------------------------------------------------------------
                 Regarding comments from VW on CO2 in-use requirements,
                EPA did not consider the change recommended by VW in the proposal and
                is not finalizing such a change. EPA believes the current program
                provides enough flexibility. EPA's general approach for this final rule
                is also to avoid providing incentives or other unique flexibilities to
                specific technologies.
                [[Page 25247]]
                F. Medium and Heavy-Duty Fuel Efficiency Technical Amendments
                 NHTSA proposed in the NPRM to make minor technical revisions to
                correct typographical mistakes and improper references adopted in the
                agency's 2016 Phase 2 medium- and heavy-duty fuel efficiency
                rule.\3451\ The proposed changes were as follows:
                ---------------------------------------------------------------------------
                 \3451\ 81 FR 73478 (Oct. 25, 2016).
                ---------------------------------------------------------------------------
                 NHTSA heavy-duty vehicles and engine fuel consumption
                credit equations. In each credit equation in 49 CFR 535.7, the minus-
                sign in each multiplication factor was omitted in the final version of
                the rule sent to the Federal Register. For example, the credit equation
                in Part 535.7(b)(1) should be specified as, Total MY Fleet FCC
                (gallons) = (Std-Act) x (Volume) x (UL) x (10-2) instead of (102), as
                currently exists. NHTSA proposed to correct these omissions.
                 The CO2 to gasoline conversion factor: In 49
                CFR 535.6(a)(4)(ii) and (d)(5)(ii), NHTSA provides the methodology and
                equations for converting the CO2 FELs/FCLs for heavy-duty
                pickups and vans (gram per mile) and for engines (grams per hp-hr) to
                their gallon-of-gasoline equivalence. In each equation, NHTSA proposed
                to correct the conversion factor to 8,887 grams per gallon of gasoline
                fuel instead of a factor of 8,877 as currently specified.
                 Curb weight definition: In 49 CFR 523.2, the reference in
                the definition for curb weight is incorrect. NHTSA proposed to correct
                the definition to incorporate a reference to 40 CFR 86.1803 instead of
                49 CFR 571.3.
                 No public comments were received in response to NHTSA's proposed
                technical corrections. Therefore, NHTSA is finalizing these amendments
                and incorporating them into its heavy-duty regulations.
                X. Regulatory Notices and Analyses
                A. Executive Order 12866, Executive Order 13563
                 Executive Order 12866, ``Regulatory Planning and Review'' (58 FR
                51735, Oct. 4, 1993), as amended by Executive Order 13563, ``Improving
                Regulation and Regulatory Review'' (76 FR 3821, Jan. 21, 2011),
                provides for making determinations whether a regulatory action is
                ``significant'' and therefore subject to the Office of Management and
                Budget (OMB) review and to the requirements of the Executive Order. One
                comment requested that the agencies provide ``a far more robust cost/
                benefit analysis as required by Executive Order (E.O.) 12866 and Office
                of Management and Budget Circular A-4.'' \3452\ The NPRM and this final
                rule satisfy the requirements of Executive Order 12866, ``Regulatory
                Planning and Review'' (58 FR 51735, Oct. 4, 1993), as amended by
                Executive Order 13563, ``Improving Regulation and Regulatory Review''
                (76 FR 3821, Jan. 21, 2011). Under these Executive Orders, this action
                is an ``economically significant regulatory action'' because it is
                likely to have an annual effect on the economy of $100 million or more.
                Accordingly, EPA and NHTSA submitted this action to the OMB for review
                and any changes made in response to OMB recommendations have been
                documented in the docket for this action. The benefits and costs of
                this proposal are described above and in the Final Regulatory Impact
                Analysis (FRIA), which is located in the docket and on the agencies'
                websites.
                ---------------------------------------------------------------------------
                 \3452\ See Anonymous Comment, Docket No. EPA-HQ-OAR-2018-0283-
                3896, at 4-5 (footnote and citation omitted). As an example, the
                comment critiqued the NPRM's discussion of the ``diminishing
                returns'' of fuel economy benefits, alleging that the discussion
                ``is not backed by reference to data or studies regarding how this
                conclusion was made.'' Id. at 5. Contrary to the comment's
                allegation, the conclusion is supported by the analysis from U.S.
                Energy Information Administration's (EIA's) Annual Energy Outlook
                (AEO) that was cited in the discussion. Id. As noted in the NPRM,
                the EIA--the statistical and analytical agency within the U.S.
                Department of Energy (DOE)--is the nation's premier source of energy
                information, and every fuel economy rulemaking since 2002 (and every
                joint CAFE and CO2 rulemaking since 2009) has applied
                fuel price projections from EIA's AEO. Id. at 42992 n.24.
                ---------------------------------------------------------------------------
                B. DOT Regulatory Policies and Procedures
                 The rule is also significant within the meaning of the Department
                of Transportation's Regulatory Policies and Procedures. The benefits
                and costs of this proposal are described above and in the FRIA, which
                is located in the docket and on NHTSA's website.
                C. Executive Order 13771 (Reducing Regulation and Controlling
                Regulatory Costs)
                 This rule is an E.O. 13771 deregulatory action. Per OMB Memorandum
                M-17-21, because this rule is deregulatory, it is not required to be
                offset by two deregulatory actions, as one comment suggested.\3453\
                ---------------------------------------------------------------------------
                 \3453\ Anonymous Comment, Docket No. EPA-HQ-OAR-2018-0283-3896,
                at 8.
                ---------------------------------------------------------------------------
                D. Executive Order 13211 (Energy Effects)
                 Executive Order 13211 applies to any rule that: (1) is determined
                to be economically significant as defined under E.O. 12866, and is
                likely to have a significant adverse effect on the supply,
                distribution, or use of energy; or (2) that is designated by the
                Administrator of the Office of Information and Regulatory Affairs as a
                significant energy action. If the regulatory action meets either
                criterion, the agencies must evaluate the adverse energy effects of the
                rule and explain why the regulation is preferable to other potentially
                effective and reasonably feasible alternatives considered.
                 The rule establishes passenger car and light truck fuel economy
                standards and tailpipe carbon dioxide and related emissions standards.
                An evaluation of energy effects of the action and reasonably feasible
                alternatives considered is provided in NHTSA's EIS and in the FRIA. To
                the extent that EPA's CO2 standards are substantially
                related to fuel economy and, accordingly, petroleum consumption, the
                EIS and FRIA analyses also provide an estimate of impacts of EPA's
                rule.
                E. Environmental Considerations
                1. National Environmental Policy Act (NEPA)
                 Concurrently with this final rule, NHTSA is releasing a Final
                Environmental Impact Statement (FEIS), pursuant to the National
                Environmental Policy Act, 42 U.S.C. 4321-4347, and implementing
                regulations issued by the Council on Environmental Quality (CEQ), 40
                CFR part 1500, and NHTSA, 49 CFR part 520. NHTSA prepared the FEIS to
                analyze and disclose the potential environmental impacts of the
                proposed CAFE standards and a range of alternatives. The FEIS analyzes
                direct, indirect, and cumulative impacts and analyzes impacts in
                proportion to their significance. It describes potential environmental
                impacts to a variety of resources, including fuel and energy use, air
                quality, climate, land use and development, hazardous materials and
                regulated wastes, historical and cultural resources, noise, and
                environmental justice. The FEIS also describes how climate change
                resulting from global carbon emissions (including CO2
                emissions attributable to the U.S. light duty transportation sector
                under the alternatives considered) could affect certain key natural and
                human resources. Resource areas are assessed qualitatively and
                quantitatively, as appropriate, in the FEIS.
                 Some commenters provided feedback on the ``flaws'' they identified
                in the CAFE model, concluding that because it played a significant role
                in modeling for the DEIS, the DEIS itself was flawed and
                [[Page 25248]]
                should be withdrawn and reissued.\3454\ The agencies address the
                comments regarding the CAFE model above in this preamble and in the
                FRIA. Ultimately, the findings on potential environmental impacts
                presented in the FEIS are of the same level of intensity and
                significance as those presented in the DEIS. While in some cases, the
                directionality of potential air quality emissions changed, the overall
                impact was generally small. NHTSA concludes that the CAFE model
                results, as used in the FEIS, do not result in the FEIS providing
                significant new information for the decisionmaker or the public
                compared to the DEIS.\3455\ NHTSA therefore concludes that a
                supplemental DEIS is not required.
                ---------------------------------------------------------------------------
                 \3454\ States of California, Connecticut, Delaware, Hawaii,
                Iowa, Illinois, Maine, Maryland, Minnesota, North Carolina, New
                Jersey, New Mexico, New York, Oregon, Rhode Island, Vermont, and
                Washington; the Commonwealths of Massachusetts, Pennsylvania, and
                Virginia; the District of Columbia; and the Cities of Los Angeles,
                New York, Oakland, San Francisco, and San Jose (``California et.
                al.--Detailed NEPA Comments''), Docket No. NHTSA-2017-0069-0625, at
                6-11; Environmental Defense Fund, Docket No. NHTSA-2018-0067-11996,
                at 3-4; and Center for Biological Diversity, et al., Docket No.
                NHTSA-2018-0067-12123, at 19.
                 \3455\ 40 CFR 1502.9(c)(1)(ii).
                ---------------------------------------------------------------------------
                 NHTSA also performed a national-scale photochemical air quality
                modeling and health benefit assessment for the FEIS; it is included as
                Appendix E. The purpose of this assessment was to use air quality
                modeling and health-related benefits analysis tools to examine the
                potential air quality-related consequences of the alternatives
                considered in its Draft Environmental Impact Statement (DEIS). In a
                comment on the DEIS, the South Coast Air Quality Management District
                stated that performing the photochemical modeling for the FEIS ``comes
                too late for the public to be able to comment on that analysis,'' and
                that the EIS must be recirculated to allow such public comment.\3456\
                However, NHTSA publicly stated its intent to conduct the analysis as
                part of the FEIS in its scoping notice published on July 26,
                2017.\3457\ The agency noted that this approach was consistent with
                past practice and resulted from the substantial time required to
                complete such an analysis. NHTSA also announced that, due to the
                substantial lead time required, the analysis would be based on the
                modeling of the alternatives presented in the DEIS, not of the
                alternatives as presented in the FEIS. NHTSA received no public
                comments in response to the scoping notice addressing this analytical
                approach, and the agency proceeded accordingly. Furthermore, while
                photochemical modeling provides spatial and temporal detail for
                estimating changes in ambient levels of air pollutants and their
                associated impacts on human health and welfare, the analysis affirms
                the estimates that appear in the EIS and does not provide significant
                new information for the decisionmaker or the public. For these reasons,
                NHTSA concludes that inclusion of the photochemical modeling and health
                benefit assessment in the FEIS is appropriate, and recirculation of the
                EIS is not required.
                ---------------------------------------------------------------------------
                 \3456\ South Coast Air Quality Management District, Docket No.
                NHTSA-2018-0067-5666, at 10. See also North Carolina Department of
                Environmental Quality, Docket No. NHTSA-2018-0067-12025, at 35-37.
                 \3457\ NHTSA, ``Notice of Intent to Prepare an Environmental
                Impact Statement for Model Year 2022-2025 Corporate Average Fuel
                Economy Standards,'' 82 FR 34740, 34743 fn. 15 (Jul. 26, 2017).
                ---------------------------------------------------------------------------
                 NHTSA has considered the information contained in the FEIS in
                making the final decision described in this final rule.\3458\ This
                preamble and final rule constitute NHTSA's Record of Decision (ROD)
                under 40 CFR 1505.2 for its promulgation of CAFE standards for MYs
                2021-2026. NHTSA has authority to issue its FEIS and ROD simultaneously
                pursuant to 49 U.S.C. 304a(b) and U.S. Department of Transportation,
                Office of Transportation Policy, Guidance on the Use of Combined Final
                Environmental Impact Statements/Records of Decision and Errata Sheets
                in National Environmental Policy Act Reviews (April 25, 2019).\3459\
                NHTSA has determined that neither the statutory criteria nor
                practicability considerations preclude simultaneous issuance.
                ---------------------------------------------------------------------------
                 \3458\ The FEIS is available for review in the public docket for
                this action and in Docket No. NHTSA-2017-0069.
                 \3459\ The guidance is available at https://www.transportation.gov/sites/dot.gov/files/docs/mission/transportation-policy/permittingcenter/337371/feis-rod-guidance-final-04302019.pdf.
                ---------------------------------------------------------------------------
                 As required by the CEQ regulations,\3460\ this final rule (as the
                ROD) sets forth the following: (1) The agency's decision (Sections V
                and VIII above); (2) alternatives considered by NHTSA in reaching its
                decision, including the environmentally preferable alternative
                (Sections V, VII, and VIII above); (3) the factors balanced by NHTSA in
                making its decision, including essential considerations of national
                policy (Section VIII.B above); (4) how these factors and considerations
                entered into its decision (Section VIII.B above); and (5) the agency's
                preferences among alternatives based on relevant factors, including
                economic and technical considerations and agency statutory missions
                (Section VIII.B.4 above). This section also briefly addresses
                mitigation\3461\ and whether all practicable means to avoid or minimize
                environmental harm from the alternative selected have been adopted.
                ---------------------------------------------------------------------------
                 \3460\ 40 CFR 1505.2.
                 \3461\ See 40 CFR 1508.20(b) (``Mitigation includes . . . (b)
                Minimizing impacts by limiting the degree or magnitude of the action
                and its implementation. . .'')
                ---------------------------------------------------------------------------
                 In the DEIS and in the FEIS, the agency identified a Preferred
                Alternative. In the DEIS, the Preferred Alternative was identified as
                Alternative 1 (0.0 Percent Annual Increase in Fuel Economy, MYs 2021-
                2026), which were the standards the agency proposed in the NPRM. In the
                FEIS, the Preferred Alternative was identified as Alternative 3 (1.5
                Percent Annual Increase in Fuel Economy, MYs 2021-2026). As the FEIS
                notes, under the Preferred Alternative, on an mpg basis, the estimated
                annual increases in the average required fuel economy levels between
                MYs 2021 and 2026 is 1.5 percent for both passenger cars and light
                trucks.\3462\ After carefully reviewing and analyzing all of the
                information in the public record, the FEIS, and comments submitted on
                the DEIS and the NPRM, NHTSA has decided to finalize the Preferred
                Alternative described in the FEIS for the reasons described in this
                ROD.
                ---------------------------------------------------------------------------
                 \3462\ Because the standards are attribute-based, average
                required fuel economy levels, and therefore rates of increase in
                those average mpg values, depend on the future composition of the
                fleet, which is uncertain and subject to change. When NHTSA
                describes a percent increase in stringency, we mean in terms of
                shifts in the footprint functions that form the basis for the actual
                CAFE standards (as in, on a gallon per mile basis, the CAFE
                standards change by a given percentage from one model year to the
                next).
                ---------------------------------------------------------------------------
                 NHTSA has considered environmental considerations as part of its
                balancing of the statutory factors to set maximum feasible fuel economy
                standards. As a result, the agency has limited the degree or magnitude
                of the action as appropriate in light of its statutory
                responsibilities. NHTSA's authority to promulgate fuel economy
                standards does not allow it to regulate criteria polluants from
                vehicles or refineries, nor can NHTSA regulate other factors affecting
                those emissions, such as driving habits. Consequently, NHTSA must set
                CAFE standards but is unable to take further steps to mitigate the
                impacts of these standards. Chapter 9 of the FEIS provides a further
                discussion of mitigation measures in the context of NEPA.
                 One commenter states that NHTSA, at a minimum, ``must include a
                thorough discussion of all reasonable mitigation measures and detail
                the appropriate agencies that could implement such
                [[Page 25249]]
                measures.'' \3463\ As examples, the commenter listed: ``creating tax
                breaks for transit and biking, expanding transportation demand
                management programs for federal employees, implementing a social
                marketing campaign regarding VMT reduction, increasing dedicated
                funding for transit and active modes, requiring VMT as a performance
                measure for federal funding, and providing NEPA guidance on evaluating
                VMT impacts of federal projects.'' Each of the examples listed is
                beyond NHTSA's statutory authority. Furthermore, documenting the myriad
                measures that could reduce VMT or address criteria pollutant or carbon
                dioxide emissions would provide no added benefit to the decisionmaker
                or the public. Each of these actions requires their own extensive cost-
                benefit anlaysis, are beyond the purview of this action, and are beyond
                the legal responsibility of NHTSA. NHTSA concludes that the commenter's
                request is beyond the bounds of NEPA's ``rule of reason.'' \3464\
                ---------------------------------------------------------------------------
                 \3463\ California et. al.--Detailed NEPA Comments, Docket No.
                NHTSA-2017-0069-0625, at 31.
                 \3464\ Dep't of Transp. v. Pub. Citizen, 541 U.S. 752, 772
                (2004).
                ---------------------------------------------------------------------------
                 Another commenter disputes NHTSA's conclusion that it lacks
                statutory authority to mitigate the impacts of its CAFE standards.
                Specifically, the commenter cites to its very authority to set fuel
                economy standards: ``It is axiomatic that fuel efficiency standards set
                at levels of the No Action Alternative or at more stringent levels
                would eliminate the additional pollution created by the proposed
                freeze.'' \3465\ This, however, mischaracterizes mitigation as nothing
                more than a choice among alternatives. NHTSA is already considering a
                range of reasonable alternatives and has concluded that alternatives
                more stringent than the No Action Alternative are beyond reasonable.
                Furthermore, NHTSA disputes that more stringent fuel economy standards
                will axiomatically lead to lower levels of criteria pollutant
                emissions. In fact, because of the rebound effect, higher levels of
                stringency may result in higher VMT, which may result in criteria
                pollutant emission increases.
                ---------------------------------------------------------------------------
                 \3465\ Center for Biological Diversity, et al., Docket No.
                NHTSA-2018-0067-12123, at 55-56.
                ---------------------------------------------------------------------------
                 The North Carolina Department of Environmental Quality commented
                that the proposed changes to the CAFE standards could undermine the
                integrity of many of the assumptions in various NEPA documents across
                the United States, in part because EPA required the use of the
                MOVES2014 model (or a subsequent revision) for transportation
                conformity determinations.\3466\ That version of MOVES incorporates
                CAFE and CO2 standards based on the agencies' actions in
                2012 and does not reflect the actions being finalized in this rule. The
                implication of the commenter's assertion, however, is that neither
                NHTSA nor EPA could take any regulatory action regarding CAFE or
                CO2 standards, regardless of whether such action was to
                increase or decrease such standards. Clearly neither agency can be
                paralyzed from undertaking its statutory obligations because of the
                independent NEPA obligations related to other ongoing Federal actions.
                For those actions, responsible officials may need to assess whether
                this final rule triggers the need for a supplemental NEPA document.
                However, it is not unique for Federal agencies to take actions or for
                new information to become available that affects the underlying inputs
                in models, such as EPA's MOVES model, on which NEPA and conformity
                analyses rely. Over time, those models will be updated to reflect these
                actions and information. EPA is responsible for approving the
                availability of models for the use in State implementation plans and
                transportation conformity analyses. EPA will evaluate and address, as
                appropriate, the impact of this action on future SIP approval actions.
                Currently approved emission factor models remain approved for SIPs and
                transportation conformity analyses, and EPA will work with DOT on the
                appropriate implementation of Federal requirements based on current and
                available information.
                ---------------------------------------------------------------------------
                 \3466\ North Carolina Department of Environmental Quality,
                Docket No. NHTSA-2018-0067-12025, at 37. See also Southern
                Environmental Law Center, EPA-HQ-OAR-2018-0283-0887, at 2-4.
                ---------------------------------------------------------------------------
                2. Clean Air Act (CAA) as Applied to NHTSA's Action
                 The CAA (42 U.S.C.[thinsp]7401 et seq.) is the primary Federal
                legislation that addresses air quality. Under the authority of the CAA
                and subsequent amendments, EPA has established National Ambient Air
                Quality Standards (NAAQS) for six criteria pollutants, which are
                specifically identified pollutants that have recognized adverse effects
                on ambient air quality and that can accumulate in the atmosphere as a
                result of human activity. EPA is required to review each NAAQS every
                five years and to revise those standards as may be appropriate
                considering new scientific information.
                 The air quality of a geographic region is usually assessed by
                comparing the levels of criteria air pollutants found in the ambient
                air to the levels established by the NAAQS (taking into account, as
                well, the other elements of a NAAQS: averaging time, form, and
                indicator). Concentrations of criteria pollutants within the air mass
                of a region are measured in parts of a pollutant per million parts
                (ppm) of air or in micrograms of a pollutant per cubic meter ([mu]g/
                m\3\) of air present in repeated air samples taken at designated
                monitoring locations using specified types of monitors. These ambient
                concentrations of each criteria pollutant are compared to the levels,
                averaging time, and form specified by the NAAQS in order to assess
                whether the region's air quality is in attainment with the NAAQS.
                 When the measured concentrations of a criteria pollutant within a
                geographic region are below those permitted by the NAAQS, EPA
                designates the region as an attainment area for that pollutant, while
                regions where concentrations of criteria pollutants exceed Federal
                standards are called nonattainment areas. Former nonattainment areas
                that are now in compliance with the NAAQS are designated as maintenance
                areas. Each State with a nonattainment area is required to develop and
                implement a State Implementation Plan (SIP) documenting how the region
                will reach attainment levels within time periods specified in the CAA.
                For maintenance areas, the SIP must document how the State intends to
                maintain compliance with the NAAQS. When EPA revises a NAAQS, each
                State must revise its SIP to address how it plans to attain the new
                standard.
                 No Federal agency may ``engage in, support in any way or provide
                financial assistance for, license or permit, or approve'' any activity
                that does not ``conform'' to a SIP or Federal Implementation Plan after
                EPA has approved or promulgated it.\3467\ Further, no Federal agency
                may ``approve, accept, or fund'' any transportation plan, program, or
                project developed pursuant to title 23 or chapter 53 of title 49,
                U.S.C., unless the plan, program, or project has been found to
                ``conform'' to any applicable implementation plan in effect.\3468\ The
                purpose of these conformity requirements is to ensure that Federally
                sponsored or conducted activities do not interfere with meeting the
                emissions targets in SIPs, do not cause or contribute to new violations
                of the NAAQS, and do not impede the ability of a State to attain or
                maintain the NAAQS or delay any interim milestones. EPA has issued two
                sets of
                [[Page 25250]]
                regulations to implement the conformity requirements:
                ---------------------------------------------------------------------------
                 \3467\ 42 U.S.C. 7506(c)(1).
                 \3468\ 42 U.S.C. 7506(c)(2).
                ---------------------------------------------------------------------------
                 (1) The Transportation Conformity Rule\3469\ applies to
                transportation plans, programs, and projects that are developed,
                funded, or approved under title 23 or chapter 53 of title 49, U.S.C.
                ---------------------------------------------------------------------------
                 \3469\ 40 CFR part 51, subpart T, and part 93, subpart A.
                ---------------------------------------------------------------------------
                 (2) The General Conformity Rule\3470\ applies to all other federal
                actions not covered under transportation conformity. The General
                Conformity Rule establishes emissions thresholds, or de minimis levels,
                for use in evaluating the conformity of an action that results in
                emissions increases.\3471\ If the net increases of direct and indirect
                emissions are lower than these thresholds, then the project is presumed
                to conform and no further conformity evaluation is required. If the net
                increases of direct and indirect emissions exceed any of these
                thresholds, and the action is not otherwise exempt, then a conformity
                determination is required. The conformity determination can entail air
                quality modeling studies, consultation with EPA and state air quality
                agencies, and commitments to revise the SIP or to implement measures to
                mitigate air quality impacts.
                ---------------------------------------------------------------------------
                 \3470\ 40 CFR part 51, subpart W, and part 93, subpart B.
                 \3471\ 40 CFR 93.153(b).
                ---------------------------------------------------------------------------
                 The CAFE standards and associated program activities are not
                developed, funded, or approved under title 23 or chapter 53 of title
                49, United States Code. Accordingly, this action and associated program
                activities are not subject to the Transportation Conformity Rule. Under
                the General Conformity Rule, a conformity determination is required
                where a Federal action would result in total direct and indirect
                emissions of a criteria pollutant or precursor originating in
                nonattainment or maintenance areas equaling or exceeding the rates
                specified in 40 CFR 93.153(b)(1) and (2). As explained below, NHTSA's
                action results in neither direct nor indirect emissions as defined in
                40 CFR 93.152.
                 The General Conformity Rule defines direct emissions as ``those
                emissions of a criteria pollutant or its precursors that are caused or
                initiated by the Federal action and originate in a nonattainment or
                maintenance area and occur at the same time and place as the action and
                are reasonably foreseeable.'' \3472\ Because NHTSA's action would set
                fuel economy standards for light duty vehicles, it would cause no
                direct emissions consistent with the meaning of the General Conformity
                Rule.\3473\
                ---------------------------------------------------------------------------
                 \3472\ 40 CFR 93.152.
                 \3473\ Dep't of Transp. v. Pub. Citizen, 541 U.S. at 772
                (``[T]he emissions from the Mexican trucks are not `direct' because
                they will not occur at the same time or at the same place as the
                promulgation of the regulations.''). NHTSA's action is to establish
                fuel economy standards for MY 2021-2026 passenger car and light
                trucks; any emissions increases would occur in a different place and
                well after promulgation of the final rule.
                ---------------------------------------------------------------------------
                 Indirect emissions under the General Conformity Rule are ``those
                emissions of a criteria pollutant or its precursors (1) That are caused
                or initiated by the federal action and originate in the same
                nonattainment or maintenance area but occur at a different time or
                place as the action; (2) that are reasonably foreseeable; (3) that the
                agency can practically control; and (4) for which the agency has
                continuing program responsibility.'' \3474\ Each element of the
                definition must be met to qualify as indirect emissions. NHTSA has
                determined that, for purposes of general conformity, emissions that may
                result from its final fuel economy standards would not be caused by
                NHTSA's action, but rather would occur because of subsequent activities
                the agency cannot practically control. ``[E]ven if a Federal licensing,
                rulemaking, or other approving action is a required initial step for a
                subsequent activity that causes emissions, such initial steps do not
                mean that a Federal agency can practically control any resulting
                emissions.'' \3475\
                ---------------------------------------------------------------------------
                 \3474\ 40 CFR 93.152.
                 \3475\ 40 CFR 93.152.
                ---------------------------------------------------------------------------
                 As the CAFE program uses performance-based standards, NHTSA cannot
                control the technologies vehicle manufacturers use to improve the fuel
                economy of passenger cars and light trucks. Furthermore, NHTSA cannot
                control consumer purchasing (which affects average achieved fleetwide
                fuel economy) and driving behavior (i.e., operation of motor vehicles,
                as measured by VMT). It is the combination of fuel economy
                technologies, consumer purchasing, and driving behavior that results in
                criteria pollutant or precursor emissions. For purposes of analyzing
                the environmental impacts of the alternatives considered here and under
                NEPA, NHTSA has made assumptions regarding all of these factors. The
                agency's FEIS predicts that increases in air toxic and criteria
                pollutants would occur in some nonattainment areas under certain
                alternatives. However, the standards and alternatives do not mandate
                specific manufacturer decisions, consumer purchasing, or driver
                behavior, and NHTSA cannot practically control any of them.\3476\
                ---------------------------------------------------------------------------
                 \3476\ See, e.g., Dep't of Transp. v. Pub. Citizen, 541 U.S.
                752, 772-73 (2004); S. Coast Air Quality Mgmt. Dist. v. Fed. Energy
                Regulatory Comm'n, 621 F.3d 1085, 1101 (9th Cir. 2010).
                ---------------------------------------------------------------------------
                 In addition, NHTSA does not have the statutory authority to control
                the actual VMT by drivers. As the extent of emissions is directly
                dependent on the operation of motor vehicles, changes in any emissions
                that result from NHTSA's CAFE standards are not changes the agency can
                practically control or for which the agency has continuing program
                responsibility. Therefore, the final CAFE standards and alternative
                standards considered by NHTSA would not cause indirect emissions under
                the General Conformity Rule, and a general conformity determination is
                not required.
                 As this analysis was presented in the NPRM, some commenters
                disagreed with NHTSA's conclusion. One commenter cited two reasons for
                concluding that the General Conformity Rule applies to NHTSA's
                action.\3477\ First, the commenter argues that NHTSA used
                ``inappropriate modeling'' in its analysis. However, this is irrelevant
                to the agency's analysis, which is based on the Federal regulations and
                the applicable case law. Second, the commenter asserts that NHTSA
                ``cannot have it both ways'' by alleging that it cannot control the
                technologies that automobile manufacturers would use or consumer
                purchasing behavior, yet justifies its rulemakings based on consumer
                purchasing and emissions implications.3478 3479 The
                rulemaking analysis presents a feasible pathway for manufacturers to
                comply with the rules, based on a series of assumptions about consumer
                behavior; it is not sufficiently foreseeable to trigger application of
                the General Conformity Rule. Furthermore, NHTSA cannot directly control
                these behaviors, and the chain of causation is too attenuated to be
                responsible for the resulting emissions. Another commenter stated that
                NHTSA has continuing
                [[Page 25251]]
                program responsibility for motor vehicle criteria pollutant emissions
                because it ``retain[s] authority to revise [its] standards in a way
                that affects future emission levels.'' \3480\ However, NHTSA disagrees
                with this assertion. First, the agency does not have statutory
                authority to regulate criteria pollutant emissions from motor vehicles.
                Second, the fact that NHTSA could establish CAFE standards for
                separate, future motor vehicles does not establish continuing program
                responsibility over emissions that could result from the vehicles
                regulated by this action.
                ---------------------------------------------------------------------------
                 \3477\ California et. al.--Detailed NEPA Comments, Docket No.
                NHTSA-2017-0069-0625, at 21-22.
                 \3478\ The commenter also quotes CBD v. NHTSA, 538 F.3d at 1217,
                for the proposition that NHTSA's regulations are the proximate cause
                of the emissions because they allow particular fuel economy levels
                that ``translate directly into particular tailpipe emissions.''
                However, that quote was referencing carbon dioxide emissions, which
                are predictable based on fuel used. NHTSA can directly regulate fuel
                economy for passenger cars and light trucks. On the other hand,
                criteria pollutant emissions are more significantly impacted by VMT,
                technology choices, and other factors that are not directly within
                the control of NHTSA.
                 \3479\ See also Joint Submission from the States of California
                et al. and the Cities of Oakland et al., Docket No. NHTSA-2018-0067-
                11735, at 35.
                 \3480\ Id.
                ---------------------------------------------------------------------------
                 NHTSA and EPA further discuss their obligations under the General
                Conformity Rule, and further address comments received, in Section
                VI.D.3 above.
                3. National Historic Preservation Act (NHPA)
                 The NHPA (54 U.S.C. 300101 et seq.) sets forth government policy
                and procedures regarding ``historic properties''--that is, districts,
                sites, buildings, structures, and objects included on or eligible for
                the National Register of Historic Places. Section 106 of the NHPA
                requires Federal agencies to ``take into account'' the effects of their
                actions on historic properties.\3481\ In the NPRM, the agencies
                concluded that the NHPA is not applicable to this rulemaking because
                the promulgation of CAFE and CO2 emissions standards for
                light duty vehicles is not the type of activity that has the potential
                to cause effects on historic properties.
                ---------------------------------------------------------------------------
                 \3481\ Section 106 is now codified at 54 U.S.C. 306108.
                Implementing regulations for the Section 106 process are located at
                36 CFR part 800.
                ---------------------------------------------------------------------------
                 Two commenters wrote that ``[c]limate change and air pollution
                imperil historic properties throughout the country via direct
                degradation, sea level rise, fire, flood, and other forms of harm.''
                Therefore, the commenters concluded that NHTSA and EPA must consult
                with the relevant Federal and State authorities and fully disclose any
                impacts to historic properties.\3482\ However, as this final rule
                establishes CAFE and CO2 standards that increase each year
                for MYs 2021-2026, this action will result in reductions in climate
                change-related impacts and most air pollutants compared to the absence
                of regulation. Furthermore, any impacts to particular historic
                properties that could be related to emissions changes associated with
                this rulemaking are not reasonably certain to occur, would be de
                minimis in their level of impact if they did occur, and are too
                attenuated to be attributed directly to this action. (See also Section
                X.E.6 below.) There is no evidence that the changes in air pollution or
                CO2 emissions associated with this rulemaking, in and of
                themselves, would alter the characteristics of a historic property
                qualifying it for inclusion in or eligibility for the National
                Register.\3483\ Nevertheless, NHTSA includes a brief, qualitative
                discussion of the impacts of the alternatives on historical and
                cultural resources in Section 7.3 of the FEIS. For the foregoing
                reasons, the agencies continue to conclude that any potential impacts
                have been accounted for in the associated analyses of this rulemaking
                and that no consultation is required under the NHPA.
                ---------------------------------------------------------------------------
                 \3482\ CARB, Docket No. NHTSA-2018-0067-11873, at 411;
                California et. al.--Detailed NEPA Comments, Docket No. NHTSA-2017-
                0069-0625, at 30.
                 \3483\ 36 CFR 800.16(i).
                ---------------------------------------------------------------------------
                4. Fish and Wildlife Conservation Act (FWCA)
                 The FWCA (16 U.S.C. 2901 et seq.) provides financial and technical
                assistance to States for the development, revision, and implementation
                of conservation plans and programs for nongame fish and wildlife. In
                addition, the Act encourages all Federal departments and agencies to
                utilize their statutory and administrative authorities to conserve and
                to promote conservation of nongame fish and wildlife and their
                habitats. The agencies conclude that the FWCA is not applicable to this
                final rule because this rulemaking does not involve the conservation of
                nongame fish and wildlife and their habitats. NHTSA has, however,
                conducted a qualitative review in its FEIS of the related direct,
                indirect, and cumulative impacts, positive or negative, of the
                alternatives on potentially affected resources, including nongame fish
                and wildlife and their habitats.
                5. Coastal Zone Management Act (CZMA)
                 The Coastal Zone Management Act (16 U.S.C. 1451 et seq.) provides
                for the preservation, protection, development, and (where possible)
                restoration and enhancement of the Nation's coastal zone resources.
                Under the statute, States are provided with funds and technical
                assistance in developing coastal zone management programs. Each
                participating State must submit its program to the Secretary of
                Commerce for approval. Once the program has been approved, any activity
                of a Federal agency, either within or outside of the coastal zone, that
                affects any land or water use or natural resource of the coastal zone
                must be carried out in a manner that is consistent, to the maximum
                extent practicable, with the enforceable policies of the State's
                program.\3484\
                ---------------------------------------------------------------------------
                 \3484\ 16 U.S.C. 1456(c)(1)(A).
                ---------------------------------------------------------------------------
                 In the NPRM, the agencies concluded that the CZMA is not applicable
                to this rulemaking because this rulemaking does not involve an activity
                within, or outside of, the Nation's coastal zones that affects any land
                or water use or natural resource of the coastal zone. CARB commented
                that California's coast is vulnerable to sea level rise from climate
                change and that the proposal would exacerbate that threat. Therefore,
                the commenter claimed that the proposal violated California's policies
                and obligations in its management program to preserve, protect, and
                enhance its coastline.\3485\ However, in its FEIS, NHTSA estimates that
                the sea-level rise in 2100 associated with Alternative 1 (0 percent
                annual average increase for both passenger cars and light trucks for
                MYs 2021-2026), the least stringent alternative considered, would be
                0.7 mm. Such a level is too small to have any meaningful impact on land
                or water use or a natural resource of the coastal zone. Furthermore, as
                this final rule establishes CAFE and CO2 standards that
                increase each year for MYs 2021-2026, this action will result in
                reductions in sea level rise resulting from climate change compared to
                the absence of regulation. Therefore, the agencies continue to conclude
                that the CZMA is not applicable to this rulemaking. NHTSA has, however,
                conducted a qualitative review in its FEIS of the related direct,
                indirect, and cumulative impacts, positive or negative, of the
                alternatives on potentially affected resources, including coastal
                zones.
                ---------------------------------------------------------------------------
                 \3485\ CARB, Docket No. NHTSA-2018-0067-11873, at 411.
                ---------------------------------------------------------------------------
                6. Endangered Species Act (ESA)
                 Under Section 7(a)(2) of the Endangered Species Act (ESA), Federal
                agencies must ensure that actions they authorize, fund, or carry out
                are ``not likely to jeopardize the continued existence'' of any
                Federally listed threatened or endangered species (collectively,
                ``listed species'') or result in the destruction or adverse
                modification of the designated critical habitat of these species.\3486\
                In general, if a Federal agency determines that an agency action may
                affect a listed species or designated critical habitat, it must
                initiate consultation with the
                [[Page 25252]]
                appropriate Service--the U.S. Fish and Wildlife Service (FWS) of the
                Department of the Interior (DOI) and/or the National Oceanic and
                Atmospheric Administration's National Marine Fisheries Service (NMFS)
                of the Department of Commerce (together, ``the Services''), depending
                on the species involved--in order to ensure that the action is not
                likely to jeopardize the species or destroy or adversely modify
                designated critical habitat.\3487\ Under this standard, the Federal
                agency taking action evaluates the possible effects of its action and
                determines whether to initiate consultation.\3488\
                ---------------------------------------------------------------------------
                 \3486\ 16 U.S.C. 1536(a)(2).
                 \3487\ See 50 CFR 402.14.
                 \3488\ See 50 CFR 402.14(a) (``Each Federal agency shall review
                its actions at the earliest possible time to determine whether any
                action may affect listed species or critical habitat.'').
                ---------------------------------------------------------------------------
                 In the NPRM, the agencies noted that they had considered the
                effects of the proposed standards and alternatives in light of
                applicable ESA regulations, case law, and guidance to determine what,
                if any, impact there might be to listed species or designated critical
                habitat. The agencies also considered the discussion in the DEIS, where
                NHTSA incorporated by reference its response to a public comment on
                page 9-101 of the MY 2017-2025 CAFE Standards Final EIS.\3489\ Based on
                that assessment, the agencies determined that the actions of setting
                CAFE and CO2 emissions standards did not require
                consultation under Section 7(a)(2) of the ESA. Accordingly, the
                agencies wrote that they had concluded their review of this action
                under Section 7 of the ESA.
                ---------------------------------------------------------------------------
                 \3489\ For the final rule for MY 2017 and beyond CAFE standards,
                NHTSA concluded that a Section 7(a)(2) consultation was not required
                because any potential for a specific impact on particular listed
                species and their habitats associated with emission changes achieved
                by that rulemaking were too uncertain and remote to trigger the
                threshold for such a consultation. In the Draft EIS, NHTSA wrote
                that this conclusion, based on the discussion and analysis cited,
                applied equally to the current rulemaking.
                ---------------------------------------------------------------------------
                 Several commenters disagreed with the agencies' assessment. In
                general, commenters stated that the agencies' proposed action would
                increase emissions of CO2 and criteria air pollutants (e.g.,
                nitrogen oxide [NOX] and sulfur dioxide
                [SO2]\3490\), that these emissions would have direct or
                indirect (i.e., through climate change) impacts on listed species and
                critical habitats, that the threshold for a finding of ``may affect''
                is extremely low, and that the agencies therefore have a duty to
                consult with the Services under the ESA.\3491\
                ---------------------------------------------------------------------------
                 \3490\ In fact, in Section 4.2.1.1 of NHTSA's FEIS, the agency
                reports that any of the action alternatives would result in
                decreased emissions of sulfur dioxide in 2025, 2035, and 2050
                compared to the No Action Alternative.
                 \3491\ See Center for Biological Diversity, Earthjustice,
                Natural Resources Defense Council, and Sierra Club, Docket Nos.
                NHTSA-2017-0069-0605 and NHTSA-2018-0067-12127; Center for
                Biological Diversity, Sierra Club, and Public Citizen, Inc., Docket
                No. NHTSA-2018-0067-12378; Center for Biological Diversity,
                Earthjustice, Environmental Law and Policy Center, Natural Resources
                Defense Council, Public Citizen, Inc., Safe Climate Campaign, Sierra
                Club, Southern Environmental Law Center, and Union of Concerned
                Scientists, Docket No. NHTSA-2018-0067-12123, at 69; States of
                California, Connecticut, Delaware, Hawaii, Iowa, Illinois, Maine,
                Maryland, Minnesota, New Jersey, New Mexico, New York, North
                Carolina, Oregon, Rhode Island, Vermont, and Washington, the
                Commonwealths of Massachusetts, Pennsylvania, and Virginia, the
                District of Columbia, and the Cities of Los Angeles, New York,
                Oakland, San Francisco, and San Jose, Docket Nos. NHTSA-2018-0067-
                11735, at 47-48; and California Air Resources Board, Docket Nos.
                NHTSA-2018-0067-11873, at 411.
                ---------------------------------------------------------------------------
                 In light of these comments, the agencies re-evaluated their
                obligations under the ESA and applicable regulations, case law, and
                guidance. Ultimately, for the following reasons, the agencies arrive at
                the same conclusion. Although there is a general association between
                the actions undertaken in this final rule and environmental impacts, as
                described in this preamble and the FEIS, the agencies' actions result
                in no effects on listed species or designated critical habitat and
                therefore do not require consultation under Section 7(a)(2) of the ESA.
                Furthermore, the agencies lack sufficient discretion or control to
                bring these actions under the consultation requirement of the ESA. The
                agencies' review under the ESA is concluded.
                a) The Agencies' Actions Have No Effects on Listed Species or Critical
                Habitat and Do Not Trigger ESA Consultation
                 Commenters have stated that CO2 and criteria air
                pollutant emissions are relevant to Section 7(a)(2) consultation
                because of the potential impacts of climate change or the pollutants
                themselves on listed species or critical habitat. The agencies have
                considered the potential impacts of this action to listed species or
                designated critical habitat of these species and conclude that any such
                impacts cannot be attributed to the agencies' actions (e.g., they are
                too uncertain and attenuated). Because the agencies conclude there are
                ``no effects,'' Section 7(a)(2) consultation is not required. The
                agencies base this conclusion both on the language of the Section
                7(a)(2) implementing regulations and on the long history of actions and
                guidance provided by DOI.
                 The Section 7(a)(2) implementing regulations require consultation
                if a Federal agency determines its action ``may affect'' listed species
                or critical habitat.\3492\ The recently revised regulations define
                ``effects of the action'' as ``all consequences to listed species or
                critical habitat that are caused by the proposed action, including the
                consequences of other activities that are caused by the proposed
                action. A consequence is caused by the proposed action if it would not
                occur but for the proposed action and it is reasonably certain to
                occur.'' \3493\ The revised definition made explicit a ``but for'' test
                and the concept of ``reasonably certain to occur'' for all
                effects.\3494\ However, in the preamble to the final rule, the Services
                emphasized that the ``but for'' test and ``reasonably certain to
                occur'' are not new or heightened standards.\3495\ In this context,
                ```but for' causation means that the consequence in question would not
                occur if the proposed action did not go forward . . . . In other words,
                if the agency fails to take the proposed action and the activity would
                still occur, there is no `but for' causation. In that event, the
                activity would not be considered an effect of the action under
                consultation.'' \3496\
                ---------------------------------------------------------------------------
                 \3492\ 50 CFR 402.14(a). The Services recently issued a final
                rule revising the regulations governing the ESA Section 7
                consultation process. 84 FR 44976 (Aug. 27, 2019). The effective
                date of the new regulations was subsequently delayed to October 28,
                2019. 84 FR 50333 (Sep. 25, 2019). As discussed in the text that
                follows, the agencies believe that their conclusion would be the
                same under both the current and prior regulations.
                 \3493\ 50 CFR 402.02 (emphasis added), as amended by 84 FR
                44976, 45016 (Aug. 27, 2019).
                 \3494\ The Services' prior regulations defined ``effects of the
                action'' in relevant part as ``the direct and indirect effects of an
                action on the species or critical habitat, together with the effects
                of other activities that are interrelated or interdependent with
                that action, that will be added to the environmental baseline.'' 50
                CFR 402.02 (as in effect prior to Oct. 28, 2019). Indirect effects
                were defined as ``those that are caused by the proposed action and
                are later in time, but still are reasonably certain to occur.'' Id.
                 \3495\ 84 FR at 44977 (``As discussed in the proposed rule, the
                Services have applied the `but for' test to determine causation for
                decades. That is, we have looked at the consequences of an action
                and used the causation standard of `but for' plus an element of
                foreseeability (i.e., reasonably certain to occur) to determine
                whether the consequence was caused by the action under
                consultation.'').
                 \3496\ Id. We note that as the Services do not consider this to
                be a change in their longstanding application of the ESA, this
                interpretation applies equally under the prior regulations (which
                were effective through October 28, 2019, and the current
                regulations.
                ---------------------------------------------------------------------------
                 The revised ESA regulations also provide a framework for
                determining whether consequences are caused by a proposed action and
                are therefore ``effects'' that may trigger consultation. The
                regulations provide in part:
                [[Page 25253]]
                 To be considered an effect of a proposed action, a consequence
                must be caused by the proposed action (i.e., the consequence would
                not occur but for the proposed action and is reasonably certain to
                occur). A conclusion of reasonably certain to occur must be based on
                clear and substantial information, using the best scientific and
                commercial data available. Considerations for determining that a
                consequence to the species or critical habitat is not caused by the
                proposed action include, but are not limited to:
                 (1) The consequence is so remote in time from the action under
                consultation that it is not reasonably certain to occur; or
                 (2) The consequence is so geographically remote from the
                immediate area involved in the action that it is not reasonably
                certain to occur; or
                 (3) The consequence is only reached through a lengthy causal
                chain that involves so many steps as to make the consequence not
                reasonably certain to occur.\3497\
                ---------------------------------------------------------------------------
                 \3497\ 50 CFR 402.17(b).
                The regulations go on to make clear that the action agency must factor
                these considerations into its assessments of potential effects.\3498\
                ---------------------------------------------------------------------------
                 \3498\ 50 CFR 402.17(c) (``Required consideration. The
                provisions in paragraphs (a) and (b) of this section must be
                considered by the action agency and the Services.'').
                 DOI, the agency charged with co-administering the ESA, previously
                evaluated whether CO2 emissions associated with a specific
                proposed Federal action triggered ESA Section 7(a)(2) consultation. The
                agencies have reviewed the long history of actions and guidance
                provided by DOI. To that point, the agencies incorporate by reference
                Appendix G of the MY 2012-2016 CAFE standards EIS.\3499\ That analysis
                relied on the significant legal and technical analysis undertaken by
                FWS and DOI. Specifically, NHTSA looked at the history of the Polar
                Bear Special Rule and several guidance memoranda provided by FWS and
                the U.S. Geological Survey. Ultimately, DOI concluded that a causal
                link could not be made between CO2 emissions associated with
                a proposed Federal action and specific effects on listed species;
                therefore, no Section 7(a)(2) consultation would be required.
                ---------------------------------------------------------------------------
                 \3499\ Available on NHTSA's Corporate Average Fuel Economy
                website at https://one.nhtsa.gov/Laws-&-Regulations/CAFE-%E2%80%93-Fuel-Economy/Final-EIS-for-CAFE-Passenger-Cars-and-Light-Trucks,-Model-Years-2012%E2%80%932016.
                ---------------------------------------------------------------------------
                 Subsequent to the publication of that Appendix, a court vacated the
                Polar Bear Special Rule on NEPA grounds, though it upheld the ESA
                analysis as having a rational basis.\3500\ FWS then issued a revised
                Final Special Rule for the Polar Bear.\3501\ In that final rule, FWS
                provided that for ESA Section 7, the determination of whether
                consultation is triggered is narrow and focused on the discrete effect
                of the proposed agency action. FWS wrote, ``[T]he consultation
                requirement is triggered only if there is a causal connection between
                the proposed action and a discernible effect to the species or critical
                habitat that is reasonably certain to occur. One must be able to
                `connect the dots' between an effect of a proposed action and an impact
                to the species and there must be a reasonable certainty that the effect
                will occur.'' \3502\ The statement in the revised Final Special Rule is
                consistent with the prior guidance published by FWS and remains valid
                today.\3503\ Likewise, the current regulations identify remoteness in
                time, geography, and the causal chain as factors to be considered in
                assessing whether a consequence is ``reasonably certain to occur.'' If
                the consequence is not reasonably certain to occur, it is not an
                ``effect of a proposed action'' and does not trigger the consultation
                requirement.
                ---------------------------------------------------------------------------
                 \3500\ In re: Polar Bear Endangered Species Act Listing and
                Section 4(D) Rule Litigation, 818 F.Supp.2d 214 (D.D.C. Oct. 17,
                2011).
                 \3501\ 78 FR 11766 (Feb. 20, 2013).
                 \3502\ 78 FR at 11784-11785.
                 \3503\ See DOI Solicitor's Opinion No. M-37017, ``Guidance on
                the Applicability of the Endangered Species Act Consultation
                Requirements to Proposed Actions Involving the Emissions of
                Greenhouse Gases'' (Oct. 3, 2008).
                ---------------------------------------------------------------------------
                 The agencies' actions establishing CAFE and CO2
                standards for passenger cars and light trucks do not directly affect
                listed species or critical habitat. The regulations promulgated by the
                agencies are used to calculate average standards for manufacturers
                based on the vehicles they produce for sale in the United States. Any
                potential effects of this action on listed species or designated
                critical habitat would be a result of changes to CO2 or air
                pollutant emissions that are caused by the individual choices of
                manufacturers in producing these vehicles and of consumers in
                purchasing and operating those vehicles. The agencies are not
                requiring, authorizing, funding, or carrying out the operation of motor
                vehicles (i.e., the proximate cause of downstream emissions), the
                production or refining of fuel (i.e., a proximate cause of upstream
                emissions),\3504\ the use of any land that is critical habitat for any
                purpose, or the taking of any listed species or other activity that may
                affect any listed species. Ultimately, the relevant decisions that
                result in emissions are taken by third parties, and any on-the-ground
                activities to implement and carry out those decisions are undertaken by
                such third parties. These decisions are influenced by a complex series
                of market factors that, though influenced by the agencies' actions,
                independently could result in the same series of decisions by consumers
                that commenters attribute to the agencies' actions (such as increased
                VMT and therefore increased emissions). This complex and lengthy chain
                of causality, which is highly dependent on market factors and therefore
                uncertain, leads the agencies to conclude that the resulting impacts of
                their actions to listed species or critical habitat do not satisfy the
                ``but for'' test or are ``reasonably certain to occur.''
                ---------------------------------------------------------------------------
                 \3504\ The agencies note that upstream emissions sources, such
                as oil extraction sites and fuel refineries, remain subject to the
                ESA. As future non-federal activities become reasonably certain,
                Section 7 and/or other sections of the ESA may provide protection
                for listed species and designated critical habitats. For example,
                new oil exploration or extraction activity may result in permitting
                or construction activities that would trigger consultation or other
                activities for the protection of listed species or designated
                critical habitat, as impacts may be more direct and more certain to
                occur.
                ---------------------------------------------------------------------------
                 With regard to climate change, EPA and NHTSA are not able to make a
                causal link for purposes of Section 7(a)(2) that would ``connect the
                dots'' between their actions, vehicle emissions from motor vehicles
                affected by their actions, climate change, and particular impacts to
                listed species or critical habitats. The agencies' actions are to set
                standards that are effectively footprint curves, which are used as part
                of a complex calculation based on the vehicles produced by
                manufacturers for sale in the United States to determine a corporate
                average standard for each manufacturer. This approach, dictated by the
                Federal statute, gives manufacturers significant discretion to design,
                produce, and sell motor vehicles to meet consumer demand. Because
                manufacturers could choose to produce more vehicles with larger
                footprints (and therefore less stringent standards), fleet-average
                CO2 emissions could increase to some extent year-over-year
                independently of where the agencies set standards. Or the opposite may
                be true, and a shift in consumer preferences could lead to increased
                production of vehicles with smaller footprints (and therefore more
                stringent standards), resulting in overall declines in CO2
                emissions in the future compared to what the agencies are forecasting.
                Importantly, consumers not only choose which vehicles to purchase
                across a range of available fuel economies, they also choose how much
                to operate those vehicles (and therefore the quantity of fuel used and
                CO2 emitted)
                [[Page 25254]]
                independently of any action undertaken by the
                agencies.3505 3506
                ---------------------------------------------------------------------------
                 \3505\ While VMT is affected by the cost of driving associated
                with fuel economy (i.e., the rebound effect), it is also affected by
                several market factors, such as economic conditions, that are far
                beyond the agencies' control and arguably have a greater influence
                than this rulemaking.
                 \3506\ The fact that overall CO2 emissions are
                influenced so heavily by consumer preferences and behavior further
                supports the agencies' conclusion that impacts are not ``reasonably
                certain to occur.''
                ---------------------------------------------------------------------------
                 Even with so many third parties in the causal chain making
                independent choices influenced by independent factors, the mechanics of
                climate change further break the chain of causality between the
                agencies' actions and specific effects on listed species or designated
                critical habitat. Climate change is a global phenomenon, impacted by
                greenhouse gas emissions that could occur anywhere throughout the
                world. As these gases accumulate in the atmosphere, radiative forcing
                increases, resulting in various potential impacts to the global climate
                system (e.g., warming temperatures, droughts, and changes in ocean pH)
                over long time scales. These changes could directly or indirectly
                impact listed species and/or designated critical habitat over time.
                Although this is a simplified explanation of a complex phenomenon
                subject to a significant degree of scientific study, it illustrates
                that the potential climate change-related consequences of this
                rulemaking on listed species and designated critical habitat are not
                ``reasonably certain to occur'' under any of the three tests in the ESA
                regulations and listed above. Not only are the consequences to listed
                species or designated critical habitat geographically and temporally
                remote from the emissions that result from regulated vehicles, the
                chain of causality is simply too lengthy and complex. Because impacts
                to listed species and designated critical habitat result from climate
                shifts that, in and of themselves, result from the accumulation over
                time of greenhouse gas emissions from anywhere in the world, there is
                simply no way to ``connect the dots'' between the emissions from a
                regulated vehicle and those impacts. While the potential impacts of
                climate change have been well-documented, there is no degree of
                certainty that this action (as distinct from any other source of
                CO2 emissions) would be the cause of any particular impact
                to listed species or critical habitats. Because greenhouse gas
                emissions continue to occur from other sectors within the U.S. and from
                other sources globally, there is simply no scientific way to apportion
                any impact to a listed species or designated critical habitat to the
                agencies' actions.\3507\
                ---------------------------------------------------------------------------
                 \3507\ See 50 CFR 402.17(b) (``A conclusion of reasonably
                certain to occur must be based on clear and substantial information,
                using the best scientific and commercial data available.'')
                ---------------------------------------------------------------------------
                 One comment to the NPRM documented the potential impacts of climate
                change on Federally protected species and included a five-page table of
                species listed during 2006 to 2015 for which the commenters claim
                climate change was a listing factor.\3508\ This conflates the
                requirements of ESA Section 4 (governing ESA listing) and ESA Section 7
                (addressing the obligations of Federal agencies). Section 4 requires
                FWS or NMFS to assess all threats to species regardless of the origin
                of those threats. 16 U.S.C. 1533(a)(1). In contrast, the focus of
                Section 7(a)(2) is narrower and requires agencies to assess only
                effects on species that are attributable to the specific agency action.
                16 U.S.C. 1536(a)(2). That climate change was considered as a factor in
                a determination to list a species does not speak to the separate
                inquiry of whether the specific agency action is impacting a listed
                species. Here, the agencies believe this comment inappropriately
                attributes the entire issue of climate change, including all
                CO2 emissions no matter which sector generated them, to
                NHTSA and EPA's actions. In fact, NHTSA and EPA's actions would have
                only very small impacts on climate attributes, such as average
                temperatures, precipitation, and sea-level rise. The likelihood that
                these very small impacts, which are described above and in NHTSA's
                FEIS, would jeopardize listed species or adversely modify designated
                critical habitat is simply too remote to be cognizable under the ESA
                consultation requirements.\3509\ The fact that the agencies would
                exacerbate the impacts of climate change to a very small degree is not
                enough to determine that impacts on listed species or designated
                critical habitat are reasonably certain to occur.3510 3511
                ---------------------------------------------------------------------------
                 \3508\ Center for Biological Diversity, Sierra Club, and Public
                Citizen, Inc., Docket No. NHTSA-2018-0067-12378, at 25-30.
                 \3509\ Ground Zero Center for Non-Violent Action v. U.S. Dept.
                of Navy, 383 F.3d 1082 (2004).
                 \3510\ Such a broad interpretation of the ESA would ensnare
                every Federal action that resulted in even an additional ounce of
                additional carbon dioxide emissions into the Section 7(a)(2)
                consultation process. See, e.g., 78 FR 11766, 11785 (Feb. 20, 2013)
                (``Without the requirement of a causal connection between the action
                under consultation and effects to species, literally every agency
                action that contributes CO2 emissions to the atmosphere
                would arguably result in consultation with respect to every listed
                species that may be affected by climate change.'').
                 \3511\ The agencies also disagree that, for purposes of
                compliance with the ESA, this action would exacerbate climate change
                impacts on listed species or critical habitat. This final rule
                establishes CAFE and CO2 standards that increase in
                stringency on a year-by-year basis. While these standards are less
                stringent than the standards considered and set forth in the 2012
                rulemaking, the ESA does not serve as a one-way ratchet when
                agencies use their inherent authority to reconsider decisions that
                have not yet taken effect.
                ---------------------------------------------------------------------------
                 As noted above, for consultation to be required, there must exist a
                sufficient nexus between the agency activity and the impact on listed
                species that the ESA intends to avoid. The Services have defined that
                nexus as ``but for'' causation. However, there is no ``but for''
                causation associated with this final rule as the impacts of climate
                change will occur regardless of this action. In fact, even if the
                agencies were to set CAFE and CO2 standards at levels that
                would eliminate all CO2 emissions from motor vehicles made
                available for sale in the United States, the impacts of climate change
                are still projected to occur due to emissions from other sectors in the
                United States and other sources globally. Changes to tailpipe
                greenhouse gas emissions or associated upstream emissions related to
                this rulemaking and the alternatives considered would be very small
                compared to global CO2 emissions, which would continue. The
                agencies also note that because third parties (as described above)
                undertake most of the decisions that result in emissions, increased
                greenhouse gas emissions could occur regardless of the agencies'
                actions in this final rule. This further demonstrates the lack of ``but
                for'' causality in this case.
                 Criteria air pollutant emissions from passenger cars and light
                trucks differ from greenhouse gas emissions in many ways. Most
                significantly, because passenger cars and light trucks are subject to
                gram-per-mile emissions standards for criteria pollutants, more fuel-
                efficient (and, correspondingly, less CO2-intensive)
                vehicles are not necessarily, from the standpoint of air quality,
                ``cleaner'' vehicles. Therefore, to the extent that CAFE and
                CO2 standards lead to changes in overall quantities of
                vehicular emissions that impact air quality, these are dominated by
                induced changes in highway travel. Changes in overall fuel consumption
                do lead to changes in emissions from ``upstream'' processes involved in
                supplying fuel to vehicles. Depending on how total vehicular emissions
                and total upstream emissions change in response to less stringent
                standards, overall emissions could increase or decrease.
                 While small in magnitude, net impacts could also vary considerably
                [[Page 25255]]
                among different geographic areas depending on the locations of upstream
                emission sources and where changes in highway travel occur. This is
                important because of another significant difference between criteria
                air pollutant emissions and greenhouse gas emissions: Criteria air
                pollutant emissions are localized \3512\ whereas CO2
                emissions contribute to global atmospheric concentrations and climate
                change no matter where they occur. As reported in Section 4.1.1 of the
                FEIS, concentrations of many air pollutants emitted from motor vehicles
                are elevated in ambient air within approximately 1,000 to 2,000 feet of
                major roadways. With meteorological conditions that tend to inhibit the
                dispersion of emissions, concentrations of traffic-generated air
                pollutants can be elevated for as much as about 8,500 feet downwind of
                roads.3513 3514 But this means that impacts of criteria
                pollutant emissions are dependent on where they occur, to a degree much
                more significant than greenhouse gas emissions. Although the agencies
                anticipate increased fuel use as a result of this final rule (compared
                to the standards described in the 2012 final rule),\3515\ NHTSA and EPA
                have no way to know with reasonable certainty where additional fuel
                extraction and refining will occur. The agencies also cannot calculate
                with reasonable certainty where changes in highway travel will occur,
                as those impacts may not be uniform across the country. In fact,
                changes in land use patterns could exacerbate or reduce criteria
                pollutant emissions in any particular area, and such local changes are
                more uncertain. Therefore, even with the best scientific and commercial
                data available, the agencies cannot draw conclusions on impacts on
                particular listed species or designated critical habitat.
                ---------------------------------------------------------------------------
                 \3512\ Criteria pollutant emissions contribute to local,
                regional, cross-state, and cross-national air pollution. Ultimately,
                however, the physical distance impacted by the pollutants is much
                smaller than for CO2 emissions, which affect the global
                atmosphere.
                 \3513\ Hu, S., S. Fruin, K. Kozawa, S. Mara, S.E. Paulson, and
                A.M. Winer. A Wide Area of Air Pollutant Impact Downwind of a
                Freeway during Pre-sunrise Hours. Atmospheric Environment.
                43(16):2541-49 (2009). doi:10.1016/j.atmosenv.2009.02.033.
                 \3514\ Hu, S., S.E. Paulson, S. Fruin, K. Kozawa, S. Mara, and
                A.M. Winer. Observation of Elevated Air Pollutant Concentrations in
                a Residential Neighborhood of Los Angeles California Using a Mobile
                Platform. Atmospheric Environment. 51:311-319 (2012). doi:10.1016/
                j.atmosenv.2011.12.055. Available at: http://europepmc.org/backend/ptpmcrender.fcgi?accid=PMC3755476&blobtype=pdf.
                 \3515\ Although, again, the agencies note that average fleet-
                wide fuel economy is projected to improve under any of the
                alternatives considered in this action.
                ---------------------------------------------------------------------------
                 In short, the impacts of CAFE and CO2 standards on
                criteria pollutant emissions is indirect, and the impacts on air
                quality at any particular location (such as where a listed species or
                designated critical habitat is located) are more ambiguous than for
                global atmospheric concentrations of CO2 over the long term.
                Therefore, the agencies reach the same conclusion for criteria
                pollutant emissions as for CO2 emissions and climate change.
                For example, the causal chain between the agencies' actions and any
                impacts to listed species or designated critical habitat is attenuated
                by the fact that independent third parties must choose not only how
                much to operate their motor vehicles, but where to operate those motor
                vehicles as well. And the agencies cannot meaningfully conclude that
                any impact to a listed species and designated critical habitat would be
                caused by criteria pollutant emissions from the vehicles regulated by
                this rule rather than by another source. Finally, the impacts on
                criteria pollutant emissions as a result of this rule, especially in
                light of other emissions sources besides the regulated vehicles, are
                small\3516\ and the likelihood of jeopardy or the adverse modification
                of designated critical habitat is too remote. Current modeling tools
                available are not designed to trace fluctuations in ambient
                concentration levels of criteria and toxic air pollutants to potential
                impacts on particular endangered species. The agencies therefore cannot
                conclude that impacts are ``reasonably certain to occur.'' \3517\
                ---------------------------------------------------------------------------
                 \3516\ For more information, see Chapter 4 of the FEIS.
                 \3517\ See 50 CFR 402.17 (``A conclusion of reasonably certain
                to occur must be based on clear and substantial information, using
                the best scientific and commercial data available'').
                ---------------------------------------------------------------------------
                 Finally, the agencies also note the potential uncertainty related
                to changes in total air pollutant and CO2 emissions as a
                result of the flexibilities in the CAFE and CO2 programs.
                Both programs allow manufacturers to trade and apply credits that have
                been earned from over-compliance in lieu of meeting the applicable
                standards for a particular model year, and manufacturers may have
                planned to rely on credits to comply with the standards for the model
                years regulated by this action. This could offset any changes in
                emissions that would result from the agencies' final decision.
                Furthermore, NHTSA's CAFE program allows manufacturers to pay civil
                penalties to cover any shortfall in compliance, further offsetting
                potential improvements in fuel economy (and, therefore, changes in air
                pollutant and CO2 emissions) that might have occurred under
                the augural standards. The existence of these flexibilities further
                supports the agencies' conclusion that they can establish neither ``but
                for'' causation nor a reasonable certainty that impacts will occur on
                listed species or designated critical habitat.
                 The agencies have considered this analysis and conclude that any
                consequence to specific listed species or designated critical habitats
                from climate change or other air pollutant emissions is too remote and
                uncertain to be attributable to the agencies' actions here. These
                consequences are not ``effects'' for purposes of consultation under
                Section 7(a)(2). NHTSA and EPA therefore conclude that this final rule
                has no effect on listed species or their critical habitats.
                (b) The Agencies Lack Sufficient Discretion or Control To Bring These
                Actions Under the Consultation Requirement of the ESA
                 The primary purpose of EPCA, as amended by EISA, and codified at 49
                U.S.C. chapter 329, is energy conservation, and NHTSA is statutorily
                obligated to set attribute-based CAFE standards for each model year at
                the levels it determines are ``maximum feasible.'' \3518\ But ``maximum
                feasible'' is a balancing of several factors, and Congress clearly did
                not envision that the CAFE program would ``solve'' energy conservation
                in a single rulemaking action.\3519\ Fuel economy standards have the
                related benefit of reducing CO2 emissions, and may also
                result in reduced emissions of many criteria air pollutants. Similarly,
                EPA has found that the elevated concentrations of greenhouse gases in
                the atmosphere may reasonably be anticipated to endanger public health
                and welfare. As a result of these findings, CAA section 202(a) requires
                the agency to issue standards applicable to emissions of such gases
                from motor vehicles. Although not a statutory requirement, EPA has
                given weight to the policy goal of establishing CO2
                [[Page 25256]]
                standards that are coordinated with NHTSA's CAFE standards.\3520\
                ---------------------------------------------------------------------------
                 \3518\ See 49 U.S.C. 32902(a) (``At least 18 months before the
                beginning of each model year, the Secretary of Transportation shall
                prescribe by regulation average fuel economy standards for
                automobiles manufactured by a manufacturer in that model year. Each
                standard shall be the maximum feasible average fuel economy level
                that the Secretary decides the manufacturers can achieve in that
                model year.'').
                 \3519\ See, e.g., 49 U.S.C. 32902(b)(2) (setting separate
                requirements for CAFE standards for MYs 2011 through 2020 and MYs
                2021 through 2030).
                 \3520\ See Mass. v. EPA, 549 U.S. 497, 532 (2007) (``. . .there
                is no reason to think the two agencies cannot both administer their
                obligations and yet avoid inconsistency.'')
                ---------------------------------------------------------------------------
                 As previously indicated, commenters assert that CO2 and
                criteria air pollutant emissions are relevant to Section 7(a)(2)
                consultation because of the potential impacts of climate change or the
                pollutants themselves on listed species or designated critical habitat.
                However, it is not clear whether their comments are based on the fact
                that the agencies predict increases in CO2 emissions and
                most criteria pollutant emissions under all action alternatives
                compared to the MY 2022-2025 CO2 and augural CAFE standards,
                or the fact that any emissions from passenger cars or light trucks will
                continue under any of the alternatives considered.
                 With regard to the latter, NHTSA does not interpret EPCA/EISA to
                mean that Congress expected the CAFE program to take the U.S. auto
                fleet off of oil entirely--indeed, EISA renders doing so impossible
                because it amended EPCA to prohibit NHTSA from considering the fuel
                economy of dedicated alternative fuel vehicles, including electric
                vehicles, when setting maximum feasible standards. This means that
                standards cannot be set that assume increased usage of full
                electrification for compliance. As a result, no matter the level at
                which NHTSA sets CAFE standards in accordance with EPCA, CO2
                and criteria pollutant emissions will continue. So long as NHTSA's
                obligation to set CAFE standards remains in place, it is reasonable to
                assume that Congress's expectation for EPA, in coordinating with NHTSA,
                is similar.
                 The purpose of Section 7(a)(2) consultation is to ensure that
                Federal agencies are not undertaking, funding, permitting, or
                authorizing actions that are likely to jeopardize the continued
                existence of listed species or destroy or adversely modify designated
                critical habitat. However, no matter what standards the agencies set
                under the CAFE and CO2 programs, Americans will continue to
                drive. Neither NHTSA nor EPA has authority to control vehicle miles
                traveled. As long as there is driving, there will be emissions--whether
                from vehicle tailpipes or from the stationary sources that create the
                energy that the vehicles consume. Moreover, both agencies have
                concluded that significant further electrification of the fleet is not
                practicable at this time due to concerns about consumer acceptance in a
                time of foreseeably low fuel prices. The fact that CO2 and
                criteria pollutant emissions will continue after NHTSA and EPA actions
                on standards cannot, alone, trigger Section 7(a)(2) consultation as the
                agencies lack the discretion or control over these emissions to simply
                regulate them away entirely in this action.\3521\ Consultation is not
                required where an agency lacks discretion to take action that will
                inure to the benefit of listed species.\3522\ Since elimination of oil
                from the fleet is inconsistent with the agencies' statutory authorities
                and the clear intent of Congress, consultation is not triggered under
                this scenario.
                ---------------------------------------------------------------------------
                 \3521\ National Ass'n of Home Builders v. Defenders of Wildlife,
                551 U.S. 644, 673 (2007) (``Applying Chevron, we defer to the
                Agency's reasonable interpretation of ESA [section] 7(a)(2) as
                applying only to `actions in which there is discretionary Federal
                involvement or control.''' (quoting 50 CFR 402.03)).
                 \3522\ Id.; Sierra Club v. Babbitt, 65 F.3d 1502, 1509 (9th Cir.
                1995) (ESA Section 7(a)(2) consultation is not required where an
                agency lacks discretion to influence private conduct in a manner
                that will inure to the benefit of listed species).
                ---------------------------------------------------------------------------
                 Commenters may instead be referring to the trend in CO2
                and criteria air pollutant emissions under the action alternatives
                considered in this rulemaking (e.g., whether and by how much emissions
                increase or decrease). To that point, all of the action alternatives
                considered result in increases in CO2 and most criteria air
                pollutant emissions compared to the standards considered and set forth
                in the 2012 rulemaking. However, the agencies do not believe this is
                the relevant comparison for purposes of determining the applicability
                of Section 7 of the ESA to this action. Model years 2021 through 2026,
                for the most part, have not yet arrived. So it is not appropriate to
                compare the current action to a prior action that has not been
                implemented and which the agencies are reconsidering. When compared to
                standards through MY 2020, under any of the alternatives considered,
                fuel economy will improve and CO2 and most criteria
                pollutant emissions will decrease over time, either as stringency
                increases or from the turnover in the fleet to newer, cleaner vehicles.
                 As detailed above, however, there is no way to meaningfully
                differentiate between the alternatives in terms of outcomes for listed
                species and designated critical habitat. The agencies cannot reasonably
                calculate how incrementally less emissions resulting from more
                stringent standards would benefit those species or habitats; rather, at
                most, the agencies can only posit that more stringent standards
                hypothetically could lead to better outcomes. But where to draw any
                line in terms of impacts to species and habitats is an impossible
                exercise. Yet, as noted above, NHTSA is mandated by Congress to set
                ``maximum feasible'' standards and EPA's mission is to protect public
                health and welfare. Under these circumstances, where the agencies must
                issue standards pursuant to statutory mandate that under any scenario
                will involve emissions, yet they lack the commensurate ability to take
                action that will inure to the benefit of species in any meaningful way,
                Section 7(a)(2) consultation is not required.
                 Finally, regardless of the level of stringency at which the
                agencies set CAFE and CO2 standards, criteria pollutant and
                CO2 emissions from motor vehicles will change to a greater
                or lesser degree because of several independent factors. Because of the
                complex relationships between fuel economy, vehicle sales, driver
                behavior (e.g., VMT and driving location), and technology choices by
                manufacturers, emissions will never uniformly increase or decrease for
                all future model years, across all regulated pollutants, and in all
                locations throughout the country. For example, increased stringency may
                result in greater VMT, resulting in larger downstream emissions of some
                criteria pollutants. On the other hand, decreased stringency may result
                in greater fuel refining, result in larger upstream emissions of some
                pollutants. Because vehicle operation and refinery activity depends
                upon independent market forces, impacts to particular listed species or
                designated critical habitat are dependent upon where vehicle operation
                or increased fuel refining occur, but neither agency can control such
                decisions. Regardless of whether NHTSA and EPA engage in Section
                7(a)(2) consultation, the agencies lack the control necessary to negate
                all emissions increases in whatever years and locations they occur
                (e.g., ensure ideal technology choices by manufacturers, control
                consumer purchasing behavior, or regulate driving locations or VMT), or
                otherwise mitigate impacts associated with these particular emissions.
                But setting stringency is, in fact, what the agencies are statutorily
                obligated to do.
                 For the foregoing reasons, NHTSA and EPA conclude that they lack
                sufficient discretion or control to bring these actions under the
                consultation requirement of the ESA.
                7. Floodplain Management (Executive Order 11988 and DOT Order 5650.2)
                 These Orders require Federal agencies to avoid the long- and short-
                term adverse impacts associated with the
                [[Page 25257]]
                occupancy and modification of floodplains, and to restore and preserve
                the natural and beneficial values served by floodplains. Executive
                Order 11988 also directs agencies to minimize the impact of floods on
                human safety, health, and welfare, and to restore and preserve the
                natural and beneficial values served by floodplains through evaluating
                the potential effects of any actions the agency may take in a
                floodplain and ensuring that its program planning and budget requests
                reflect consideration of flood hazards and floodplain management. DOT
                Order 5650.2 sets forth DOT policies and procedures for implementing
                Executive Order 11988. The DOT Order requires that the agency determine
                if a proposed action is within the limits of a base floodplain, meaning
                it is encroaching on the floodplain, and whether this encroachment is
                significant. If significant, the agency is required to conduct further
                analysis of the proposed action and any practicable alternatives. If a
                practicable alternative avoids floodplain encroachment, then the agency
                is required to implement it.
                 In this rulemaking, the agencies are not occupying, modifying and/
                or encroaching on floodplains. The agencies, therefore, conclude that
                the Orders are not applicable to this action. NHTSA has, however,
                conducted a review of the alternatives on potentially affected
                resources, including floodplains, in its FEIS.
                8. Preservation of the Nation's Wetlands (Executive Order 11990 and DOT
                Order 5660.1a)
                 These Orders require Federal agencies to avoid, to the extent
                possible, undertaking or providing assistance for new construction
                located in wetlands unless the agency head finds that there is no
                practicable alternative to such construction and that the proposed
                action includes all practicable measures to minimize harm to wetlands
                that may result from such use. Executive Order 11990 also directs
                agencies to take action to minimize the destruction, loss, or
                degradation of wetlands in ``conducting Federal activities and programs
                affecting land use, including but not limited to water and related land
                resources planning, regulating, and licensing activities.'' DOT Order
                5660.1a sets forth DOT policy for interpreting Executive Order 11990
                and requires that transportation projects ``located in or having an
                impact on wetlands'' should be conducted to assure protection of the
                Nation's wetlands. If a project does have a significant impact on
                wetlands, an EIS must be prepared.
                 In the NPRM, the agencies noted that they are not undertaking or
                providing assistance for new construction located in wetlands. The
                agencies, therefore, concluded that these Orders do not apply to this
                rulemaking. One commenter disagreed with this conclusion, noting the
                potential land use impacts of the rule and the agencies' obligation to
                consider all factors relevant to the proposal's effect on the survival
                and quality of wetlands.\3523\ The agencies do not believe that it is
                feasible to establish the requisite causal chain between the impacts of
                this action and impacts on wetlands, nor would such impacts be
                reasonably foreseeable as a direct or indirect result of this
                rulemaking. The agencies therefore continue to conclude that these
                Orders do not apply to this rulemaking. Regardless, NHTSA addresses the
                potential effects of the alternatives on resources, including wetlands,
                in its FEIS.
                ---------------------------------------------------------------------------
                 \3523\ Joint Submission from the States of California et al. and
                the Cities of Oakland et al., Docket No. NHTSA-2018-0067-11735, at
                46-47.
                ---------------------------------------------------------------------------
                9. Migratory Bird Treaty Act (MBTA), Bald and Golden Eagle Protection
                Act (BGEPA), Executive Order 13186
                 The MBTA (16 U.S.C. 703-712) provides for the protection of certain
                migratory birds by making it illegal for anyone to ``pursue, hunt,
                take, capture, kill, attempt to take, capture, or kill, possess, offer
                for sale, sell, offer to barter, barter, offer to purchase, purchase,
                deliver for shipment, ship, export, import, cause to be shipped,
                exported, or imported, deliver for transportation, transport or cause
                to be transported, carry or cause to be carried, or receive for
                shipment, transportation, carriage, or export'' any migratory bird
                covered under the statute.\3524\
                ---------------------------------------------------------------------------
                 \3524\ 16 U.S.C. 703(a).
                ---------------------------------------------------------------------------
                 The BGEPA (16 U.S.C. 668-668d) makes it illegal to ``take, possess,
                sell, purchase, barter, offer to sell, purchase or barter, transport,
                export or import'' any bald or golden eagles.\3525\ Executive Order
                13186, ``Responsibilities of Federal Agencies to Protect Migratory
                Birds,'' helps to further the purposes of the MBTA by requiring a
                Federal agency to develop a Memorandum of Understanding (MOU) with the
                Fish and Wildlife Service when it is taking an action that has (or is
                likely to have) a measurable negative impact on migratory bird
                populations.
                ---------------------------------------------------------------------------
                 \3525\ 16 U.S.C. 668(a).
                ---------------------------------------------------------------------------
                 The agencies conclude that the MBTA, BGEPA, and Executive Order
                13186 do not apply to this action because there is no disturbance,
                take, measurable negative impact, or other covered activity involving
                migratory birds or bald or golden eagles involved in this rulemaking.
                10. Department of Transportation Act (Section 4(f))
                 Section 4(f) of the Department of Transportation Act of 1966 (49
                U.S.C. 303), as amended, is designed to preserve publicly owned park
                and recreation lands, waterfowl and wildlife refuges, and historic
                sites. Specifically, Section 4(f) provides that DOT agencies cannot
                approve a transportation program or project that requires the use of
                any publicly owned land from a public park, recreation area, or
                wildlife or waterfowl refuge of national, State, or local significance,
                or any land from a historic site of national, State, or local
                significance, unless a determination is made that:
                 (1) There is no feasible and prudent alternative to the use of
                land, and
                 (2) The program or project includes all possible planning to
                minimize harm to the property resulting from the use.
                 These requirements may be satisfied if the transportation use of a
                Section 4(f) property results in a de minimis impact on the area.
                 NHTSA concludes that Section 4(f) is not applicable to this action
                because this rulemaking is not an approval of a transportation program
                or project that requires the use of any publicly owned land.
                11. Executive Order 12898: ``Federal Actions To Address Environmental
                Justice in Minority Populations and Low-Income Populations''
                 Executive Order 12898 (59 FR 7629 (Feb. 16, 1994)) establishes
                Federal executive policy on environmental justice. It directs Federal
                agencies, to the greatest extent practicable and permitted by law, to
                make environmental justice part of their mission by identifying and
                addressing, as appropriate, disproportionately high and adverse human
                health or environmental effects of their programs, policies, and
                activities on minority and low-income populations in the United States.
                DOT Order 5610.2(a) \3526\ sets forth the Department of
                Transportation's policy to consider environmental justice principles in
                all its programs, policies, and activities.
                ---------------------------------------------------------------------------
                 \3526\ Department of Transportation Updated Environmental
                Justice Order 5610.2(a), 77 FR 27534 (May 10, 2012).
                ---------------------------------------------------------------------------
                 Environmental justice is a principle asserting that all people
                deserve fair treatment and meaningful involvement with respect to
                environmental laws,
                [[Page 25258]]
                regulations, and policies. EPA seeks to provide the same degree of
                protection from environmental health hazards for all people. DOT shares
                this goal and is informed about the potential environmental impacts of
                its rulemakings through the NEPA process. One comment on the NPRM
                claimed that the agencies ``failed to recognize the benefits of the
                existing standards'' for disadvantaged communities. Specifically, the
                commenter claimed that the agencies did not provide an underlying
                analysis of environmental justice issues and thereby failed to meet the
                requirements of E.O. 12898.\3527\ However, the agencies addressed their
                obligations under E.O. 12898 in the preamble to the NPRM and in Section
                7.5 of the DEIS. The agencies received a number of comments regarding
                the analysis it presented. NHTSA responds to those comments in Section
                10.7 of the FEIS, and the agencies have revised their environmental
                justice analysis based on the information contained in those comments.
                The revised analysis is presented here and in the FEIS.
                ---------------------------------------------------------------------------
                 \3527\ CARB, Docket No. NHTSA-2018-0067-11873, at 411-12.
                ---------------------------------------------------------------------------
                 There is evidence that proximity to oil refineries could be
                correlated with incidences of cancer and
                leukemia.3528 3529 3530 Proximity to high-traffic roadways
                could result in adverse cardiovascular and respiratory impacts, among
                other possible impacts.3531 3532 3533 3534 3535 3536 3537
                Climate change affects overall global temperatures, which could, in
                turn, affect the number and severity of outbreaks of vector-borne
                illnesses.3538 3539 In the context of this rulemaking, the
                environmental justice concern is the extent to which minority and low-
                income populations could be more exposed or vulnerable to such
                environmental and health impacts.
                ---------------------------------------------------------------------------
                 \3528\ Pukkala, E. Cancer incidence among Finnish oil refinery
                workers, 1971-1994. Journal of Occupational and Environmental
                Medicine. 40(8):675-79 (1998). doi:10.1023/A:1018474919807.
                 \3529\ Chan, C.-C.; Shie, R.H.; Chang, T.Y.; Tsai, D.H. Workers'
                exposures and potential health risks to air toxics in a
                petrochemical complex assessed by improved methodology.
                International Archives of Occupational and Environmental Health.
                79(2):135-142 (2006). doi:10.1007/s00420-005-0028-9. Online at:
                https://www.researchgate.net/publication/7605242_Workers'_exposures_and_potential_health_risks_to_air_toxics_i
                n_a_petrochemical_complex_assessed_by_improved_methodology.
                 \3530\ Bulka, C.; Nastoupil, L.J.; McClellan, W.; Ambinder, A.;
                Phillips, A.; Ward, K.; Bayakly, A.R.; Switchenko, J.M.; Waller, L.;
                Flowers, C.R. Residence proximity to benzene release sites is
                associated with increased incidence of non-Hodgkin lymphoma. Cancer.
                119(18):3309-17 (2013). doi:10.1002/cncr.28083. Online at: http://onlinelibrary.wiley.com/doi/10.1002/cncr.28083/pdf;jsessionid=1520A90A764A95985316057D7D76A362.f02t02.
                 \3531\ HEI (Health Effects Institute). 2010. Traffic-Related Air
                Pollution: A Critical Review of the Literature on Emissions,
                Exposure and Health Effects. Special Report 17. Health Effects
                Institute: Boston, MA:. HEI Panel on the Health Effects of Traffic-
                Related Air Pollution, 386 pp. Available at: https://www.healtheffects.org/system/files/SR17Traffic%20Review.pdf.
                (Accessed: March 3, 2018).
                 \3532\ Heinrich, J. and H.-E. Wichmann. 2004. Traffic Related
                Pollutants in Europe and their Effect on Allergic Disease. Current
                Opinion in Allergy and Clinical Immunology 4(5):341-348.
                 \3533\ Salam, M.T., T. Islam, and F.D. Gilliland. 2008. Recent
                Evidence for Adverse Effects of Residential Proximity to Traffic
                Sources on Asthma. Current Opinion in Pulmonary Medicine 14(1):3-8.
                doi:10.1097/MCP.0b013e3282f1987a.
                 \3534\ Samet, J.M. 2007. Traffic, Air Pollution, and Health.
                Inhalation Toxicology 19(12):1021-27. doi:10.1080/08958370701533541.
                 \3535\ Adar, S. and J. Kaufman. 2007. Cardiovascular Disease and
                Air Pollutants: Evaluating and Improving Epidemiological Data
                Implicating Traffic Exposure. Inhalation Toxicology 19(S1):135-49.
                doi:10.1080/08958370701496012.
                 \3536\ Wilker, E.H., E. Mostofsky, S.H. Lue, D. Gold, J.
                Schwartz, G.A. Wellenius, and M.A. Mittleman. 2013. Residential
                Proximity to High-Traffic Roadways and Poststroke Mortality. Journal
                of Stroke and Cerebrovascular Diseases 22(8): e366-e372.
                doi:10.1016/j.jstrokecerebrovasdis.2013.03.034. Available at:
                https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4066388/. (Accessed:
                March 6, 2018).
                 \3537\ Hart, J.E., E.B. Rimm, K.M. Rexrode, and F. Laden. 2013.
                Changes in Traffic Exposure and the Risk of Incident Myocardial
                Infarction and All-cause Mortality. Epidemiology 24(5):734-42.
                 \3538\ U.S. Global Change Research Program (GCRP). Global
                Climate Change Impacts in the United States: The Third National
                Climate Assessment. U.S. Global Change Research Program. Melillo,
                J.M, T.C. Richmond, and G.W. Yohe (Eds.). U.S. Government Printing
                Office: Washington, DC 841 pp (2014). doi:10.7930/J0Z31WJ2.
                Available at: http://nca2014.globalchange.gov/report. (Accessed:
                February 27, 2018).
                 \3539\ GCRP. The Impacts of Climate Change on Human Health in
                the United States, A Scientific Assessment (2016). April 2016.
                Available at: https://health2016.globalchange.gov. (Accessed:
                February 28, 2018).
                ---------------------------------------------------------------------------
                 Numerous studies have found that some environmental hazards are
                more prevalent in areas where racial/ethnic minorities and people with
                low socioeconomic status represent a higher proportion of the
                population compared with the general population. In addition, compared
                to non-Hispanic whites, some subpopulations defined by race and
                ethnicity have been shown to have a greater incidence of some health
                conditions during certain life stages. For example, in 2014, about 13
                percent of Black, non-Hispanic and 24 percent of Puerto Rican children
                were estimated to have asthma, compared with 8 percent of white, non-
                Hispanic children.\3540\ The agencies have therefore considered areas
                nationwide that could contain minority and low-income communities who
                would most likely be exposed to the environmental and health impacts of
                oil production, distribution, and consumption or the potential impacts
                of climate change. These include areas where oil production and
                refining occur, areas near roadways, coastal flood-prone areas, and
                urban areas that are subject to the heat island effect.\3541\
                ---------------------------------------------------------------------------
                 \3540\ http://www.cdc.gov/asthma/most_recent_data.htm.
                 \3541\ The heat island effect refers to developed areas having
                higher temperatures than surrounding rural areas.
                ---------------------------------------------------------------------------
                 The following discussion addresses environmental justice
                implications related to air quality and to climate change and carbon
                emissions in the context of this final rulemaking. Emissions of air
                pollutants may be affected by this rulemaking due to changes in fuel
                use and VMT, which are described above. To the degree to which minority
                and low-income populations may be present in proximity to the locations
                described in this section, they may be exposed disproportionately to
                these emissions changes. In addition, the following analysis also
                discusses other potential reasons why minority and low-income
                populations may be susceptible to the health impacts of air pollutants.
                NHTSA also discusses environmental justice in Chapter 7.5 of its FEIS.
                a) Proximity to Oil Production and Refining
                 As stated above, numerous studies have found that some
                environmental hazards are more prevaluent in areas where minority and
                low-income populations represent a higher proportion of the population
                compared with the general population. For example, one study found that
                survey respondents who were black and, to a lesser degree, had lower
                income levels, were significantly more likely to live within 1 mile of
                an industrial facility listed in the EPA's 1987 Toxic Release Inventory
                (TRI) national database.\3542\
                ---------------------------------------------------------------------------
                 \3542\ Mohai, P., P.M. Lantz, J. Morenoff, J.S. House, and R.P.
                Mero. Racial and Socioeconomic Disparities in Residential Proximity
                to Polluting Industrial Facilities: Evidence from the Americans'
                Changing Lives Study. American Journal of Public Health 99(S3):
                S649-S656 (2009). doi:10.2105/AJPH.2007.131383. Available at: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2774179/pdf/S649.pdf.
                (Accessed: March 2, 2018).
                ---------------------------------------------------------------------------
                 A meta-analysis of 49 environmental equity studies concluded that
                evidence of race-based environmental inequities is statistically
                significant (although the average magnitude of these inequities is
                small), while evidence supporting the existence of income-based
                environmental inequities is substantially weaker.\3543\ Considering
                poverty-based class effects, that meta-
                [[Page 25259]]
                analysis found an inverse relationship between environmental risk and
                poverty, concluding that environmental risks are less likely to be
                located in areas of extreme poverty.\3544\ However, individual studies
                may reach contradictory conclusions in relation to race- and income-
                based inequities across a range of environmental risks. Therefore, the
                meta-analysis also sought to examine the reasons why conclusions vary
                across studies of environmental inequity. Possible explanations for why
                studies reach contrary conclusions include variability in the source of
                potential environmental risk that the study considers (e.g., the type
                of facility or the associated level of pollution or risk); variability
                in the methodology applied to aggregate demographic data and to define
                the comparison population; and the degree to which statistical models
                control for other variables that may explain the distribution of
                potential environmental risk.
                ---------------------------------------------------------------------------
                 \3543\ Ringquist, E.J. Evidence of Environmental Inequities: A
                Meta-Analysis. Journal of Policy Analysis and Management 24(2):223-
                47 (2005).
                 \3544\ Ringuist (2005).
                ---------------------------------------------------------------------------
                 To test whether there are disparate impacts from hazardous
                industrial facilities on racial/ethnic minorities, the disadvantaged,
                the working class, and manufacturing workers, one study tested the
                relationship between hazard scores of Philadelphia-area facilities in
                EPA's Risk-Screening Environmental Indicators (RSEI) database and the
                demographics of populations near those facilities using multivariate
                regression.\3545\ This study concluded that racial/ethnic minorities,
                the most socioeconomically disadvantaged, and those employed in
                manufacturing suffer a disparate impact from the highest-hazard
                facilities (primarily manufacturing plants).
                ---------------------------------------------------------------------------
                 \3545\ Sicotte, D. and S. Swanson. Whose Risk in Philadelphia?
                Proximity to Unequally Hazardous Industrial Facilities. Social
                Science Quarterly 88(2):516-534 (2007).
                ---------------------------------------------------------------------------
                 Other commissioned reports and case studies provide additional
                evidence of the presence of low-income and minority populations near
                industrial facilities and of racial or socioeconomic disparities in
                exposure to environmental risk, although these sources were not
                published in peer-reviewed scientific
                journals.3546 3547 3548 3549
                ---------------------------------------------------------------------------
                 \3546\ UCC (United Church of Christ). Toxic Wastes and Race at
                Twenty: 1987--2007. A Report Prepared for the United Church of
                Christ Justice and Witness Ministries. Available at: https://www.nrdc.org/sites/default/files/toxic-wastes-and-race-at-twenty-1987-2007.pdf (2007). (Accessed: April 9, 2018).
                 \3547\ National Association for the Advancement of Colored
                People and Clean Air Task Force. Fumes Across the Fence-line: The
                Health Impacts of Air Pollution from Oil & Gas Facilities on African
                American Communities (2017). Available at: http://www.catf.us/wp-content/uploads/2017/11/CATF_Pub_FumesAcrossTheFenceLine.pdf.
                (Accessed: February 24, 2019).
                 \3548\ Ash, M., J.K. Boyce, G. Chang, M. Pastor, J. Scoggins,
                and J. Tran. Justice in the Air: Tracking Toxic Pollution from
                America's Industries and Companies to our States, Cities, and
                Neighborhoods. Political Economy Research Institute at the
                University of Massachusetts, Amherst and the Program for
                Environmental and Regional Equity at the University of Southern
                California (2009). Available at: https://dornsife.usc.edu/assets/sites/242/docs/justice_in_the_air_web.pdf. (Accessed: February 24,
                2019).
                 \3549\ Kay, J. and C. Katz. Pollution, Poverty and People of
                Color: Living With Industry. Scientific American. Available at:
                https://www.scientificamerican.com/article/pollution-poverty-people-color-living-industry/ (2012). (Accessed: March 4, 2018).
                ---------------------------------------------------------------------------
                 Few studies address disproportionate exposure to environmental risk
                associated with oil refineries specifically. One study found that the
                populations surrounding oil refineries are more often minorities,
                concluding that ``56 percent of people living within three miles of
                [oil] refineries in the United States are minorities--almost double the
                national average.'' \3550\ Another examined whether findings of
                environmental inequity varied between coke production plants and oil
                refineries, both of which are significant sources of air
                pollution.\3551\ This study concluded that census tracts near coke
                plants had a disproportionate share of poor and nonwhite residents, and
                that existing inequities were primarily economic in nature. However,
                the findings for oil refineries did not strongly support an
                environmental inequity hypothesis. A more recent study of environmental
                justice in the oil refinery industry found evidence of environmental
                injustice as a result of unemployment levels in areas around refineries
                and, to a slightly lesser extent, as a result of income
                inequality.\3552\ This study did not test for race-based environmental
                inequities.
                ---------------------------------------------------------------------------
                 \3550\ O'Rourke, D. and S. Connolly. Just Oil? The Distribution
                of Environmental and Social Impacts of Oil Production and
                Consumption. Annual Review of Environment and Resources 28(1):587-
                617 (2003). doi:10.1146/annurev.energy.28.050302.105617.
                 \3551\ Graham, J.D., N.D. Beaulieu, D. Sussman, M. Sadowitz, and
                Y.C. Li. Who Lives Near Coke Plants and Oil Refineries? An
                Exploration of the Environmental Inequity Hypothesis. Risk Analysis
                19(2):171-86 (1999). doi:10.1023/A:1006965325489. Green, R.S., S.
                Smorodinsky, J.J. Kim, R. McLaughlin, and B. Ostro. Proximity of
                California public schools to busy roads. Environmental Health
                Perspectives 112 (1):61-66 (2004). Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1241798/. (Accessed: May 31,
                2018).
                 \3552\ Carpenter, A. and M. Wagner. Environmental Justice in the
                Oil Refinery Industry: A Panel Analysis Across United States
                Counties. Ecological Economics 159:101-109 (2019).
                ---------------------------------------------------------------------------
                 Overall, the body of scientific literature points to
                disproportionate representation of minority and low-income populations
                in proximity to a range of industrial, manufacturing, and hazardous
                waste facilities that are stationary sources of air pollution; although
                results of individual studies may vary. While the scientific literature
                specific to oil refineries is limited, disproportionate exposure of
                minority and low-income populations to air pollution from oil
                refineries is suggested by other broader studies of racial and
                socioeconomic disparities in proximity to industrial facilities
                generally.
                 The potential increase in fuel production and consumption projected
                as a result of this rulemaking (compared to the No Action Alternative)
                could lead to an increase in upstream emissions of criteria and toxic
                air pollutants due to increased extraction, refining, and
                transportation of fuel. As described in Section VII.A.4.c.3.b.i, total
                upstream emissions of criteria and toxic air pollutants in 2035 are
                projected to increase under all action alternatives compared to the No
                Action Alternative, with the exception that total upstream emissions of
                SO2 are projected to decrease under all action alternatives
                under the CAFE program (but not under the CO2 program). As
                noted, a correlation between proximity to oil refineries and the
                prevalence of minority and low-income populations is suggested in the
                scientific literature. To the extent that minority and low-income
                populations live closer to oil refining facilities, these populations
                may be more likely to be adversely affected by these emissions.
                However, the magnitude of the change in emissions relative to the
                baseline is minor and would not be characterized as high and adverse.
                Proximity to High-Traffic Roadways
                 Studies have more consistently demonstrated a disproportionate
                prevalence of minority and low-income populations living near mobile
                sources of pollutants. In certain locations in the United States, for
                example, there is consistent evidence that populations or schools near
                roadways typically include a greater percentage of minority or low-
                income residents.3553 3554 3555 3556 3557 3558 3559 In
                [[Page 25260]]
                California, studies demonstrate that minorities and low-income
                populations are disproportionately likely to live near a major roadway
                or in areas of high traffic density compared to the general
                population.3560 3561 A study of traffic, air pollution, and
                socio-economic status inside and outside the Minneapolis-St. Paul
                metropolitan area similarly found that populations on the lower end of
                the socioeconomic spectrum and minorities are disproportionately
                exposed to traffic and air pollution and at higher risk for adverse
                health outcomes.\3562\ Near-road exposure to vehicle emissions can
                cause or exacerbate health conditions such as
                asthma.3563 3564 3565 3566 One study demonstrated that
                students at schools in Michigan closer to major highways had a higher
                risk of respiratory and neurological disease and were more likely to
                fail to meet state educational standards, after controlling for other
                variables.\3567\ In general, studies such as these demonstrate trends
                in specific locations in the United States that may be indicative of
                broader national trends.
                ---------------------------------------------------------------------------
                 \3553\ Green, R.S., S. Smorodinsky, J.J. Kim, R. McLaughlin, and
                B. Ostro. Proximity of California public schools to busy roads.
                Environmental Health Perspectives 112 (1):61-66 (2004). Available
                at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1241798/. Last
                accessed: May 31, 2018.
                 \3554\ Wu, Y-C.; Batterman, S.A. Proximity of schools in
                Detroit, Michigan to automobile and truck traffic. Journal of
                Exposure Science and Environmental Epidemiology 16(5): 457-470
                (2006). doi:10.1038/sj.jes.7500484. Available at: http://www.nature.com/articles/7500484. Last accessed: May 31, 2018.
                 \3555\ Chakraborty, J., and P.A. Zandbergen. Children at risk:
                measuring racial/ethnic disparities in potential exposure to air
                pollution at school and home. Journal of Epidemiology & Community
                Health 61:1074-1079 (2007). doi: 10.1136/jech.2006.054130.
                 \3556\ Depro, B., and C. Timmins. Mobility and Environmental
                Equity: Do Housing Choices Determine Exposure to Air Pollution?
                North Carolina State University and RTI International, Duke
                University and NBER (2008). Available at: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.586.7164&rep=rep1&type=pdf. (Accessed: May 31,
                2018).
                 \3557\ Marshall, J.D. Environmental inequality: air pollution
                exposures in California's South Coast Air Basin. Atmospheric
                Environment 42(21):5499-5503 (2008).
                 \3558\ Su, J. G., T. Larson, T. Gould, M. Cohen, and M.
                Buzzelli. Transboundary air pollution and environmental justice:
                Vancouver and Seattle compared. GeoJournal 75(6):595-608 (2010).
                doi: 10.1007/s10708-009-9269-6.
                 \3559\ Su, J. G., M. Jarrett, A. de Nazelle, and J. Wolch. Does
                exposure to air pollution in urban parks have socioeconomic, racial
                or ethnic gradients? Environmental Research 111 (3):319-328 (2011).
                doi: 10.1016/j.envres.2011.01.002.
                 \3560\ Carlson, A.E. The Clean Air Act's Blind Spot:
                Microclimates and Hotspot Pollution. 65 UCLA Law Review 1036 (2018).
                 \3561\ Gunier, R.B., A. Hertz, J. Von Behren, and P. Reynolds.
                Traffic density in California: socioeconomic and ethnic differences
                among potentially exposed children. Journal of Exposure Analysis and
                Environmental Epidemiology 13(3):240-46 (2003). doi:10.1038/
                sj.jea.7500276.
                 \3562\ Pratt, G.C., M.L. Vadali, D.L. Kvale, and K.M. Ellickson,
                Traffic, air pollution, minority, and socio-economic status:
                addressing inequities in exposure and risk. International Journal of
                Environmental research and Public Health 12(5):53555372 (2015).
                doi:10.3390/ijerph120505355.
                 \3563\ Carlson (2018).
                 \3564\ Gunier et al. (2003).
                 \3565\ Meng, Y-Y., M. Wilhelm, R.P. Rull, P. English, S. Nathan,
                and B. Ritz. Are frequent asthma symptoms among low-income
                individuals related to heavy traffic near homes, vulnerabilities, or
                both? Annals of Epidemiology 18:343-350 (2008). doi:10.1016/
                j.annepidem.2008.01.006.
                 \3566\ Khreis, H., C. Kelly, J. Tate, R. Parslow, K. Lucas, and
                M. Nieuwenhuijsen. Exposure to traffic-related air pollution and
                risk of development of childhood asthma: A systematic review and
                meta-analysis. Environment International 100:1-31 (2017). https://doi.org/10.1016/j.envint.2016.11.012.
                 \3567\ Kweon, B-S., P. Mohai, S. Lee, and A.M. Sametshaw. 2016.
                Proximity of Public Schools to Major Highways and Industrial
                Facilities, and Students' School Performance and Health Hazards.
                Environment and Planning B: Urban Analytics and City Science
                45(2):312-329. doi.org/10.1177/0265813516673060.
                ---------------------------------------------------------------------------
                 Fewer studies have been conducted at the national level, yet those
                that do exist also demonstrate a correlation between minority and low-
                income status and proximity to roadways.3568 3569 For
                example, one study found that greater traffic volumes and densities at
                the national level are associated with larger shares of minority and
                low-income populations living in the vicinity.\3570\ Another study
                found that schools with minority and underprivileged \3571\ children
                were disproportionately located within 250 meters of a major
                roadway.\3572\
                ---------------------------------------------------------------------------
                 \3568\ Tian, N., J. Xue, and T. M. Barzyk. Evaluating
                socioeconomic and racial differences in traffic-related metrics in
                the United States using a GIS approach. Journal of Exposure Science
                and Environmental Epidemiology 23 (2):215 (2013). doi: 10.1038/
                jes.2012.83. Available at: http://www.nature.com/articles/jes201283.
                (Accessed: May 31, 2018).
                 \3569\ Boehmer, T.K., S.L. Foster, J.R. Henry, E.L. Woghiren-
                Akinnifesi, and F.Y. Yip. Residential Proximity to Major Highways--
                United States, 2010. Morbidity and Mortality Weekly Report 62(3):46-
                50 (2013). Available at: http://www.cdc.gov/mmwr/preview/mmwrhtml/su6203a8.htm. (Accessed: February 26, 2018).
                 \3570\ Rowangould, G.M. A Census of the US Near-roadway
                Population: Public Health and Environmental Justice Considerations.
                Transportation Research Part D: Transport and Environment 25:59-67
                (2013). doi:10.1016/j.trd.2013.08.003.
                 \3571\ Public schools were determined to serve predominantly
                underprivileged students if they were eligible for Title I programs
                (federal programs that provide funds to school districts and schools
                with high numbers or high percentages of children who are
                disadvantaged) or had a majority of students who were eligible for
                free/reduced-price meals under the National School Lunch and
                Breakfast Programs.
                 \3572\ Kingsley, S.L., M.N. Eliot, L. Carlson, J. Finn, D.L.
                MacIntosh, H.H. Suh, and G.A. Wellenius. Proximity of US Schools to
                Major Roadways: A Nationwide Assessment. Journal of Exposure Science
                and Environmental Epidemiology 24(3):253-59 (2014). doi:10.1038/
                jes.2014.5.
                ---------------------------------------------------------------------------
                 As detailed in Section 10.3.8 of the PRIA and Section X.E.11.a.2 of
                the FRIA, NHTSA and EPA analyzed two national databases that allowed
                evaluation of whether homes and schools were located near a major road
                and whether disparities in exposure may be occurring in these
                environments. The American Housing Survey (AHS) includes descriptive
                statistics of over 70,000 housing units across the nation. The study
                survey is conducted every two years by the U.S. Census Bureau. The
                second database the agencies analyzed was the U.S. Department of
                Education's Common Core of Data, which includes enrollment and location
                information for schools across the U.S.
                 In analyzing the 2009 AHS, the focus was on whether or not a
                housing unit was located within 300 feet of a ``4-or-more lane highway,
                railroad, or airport.'' \3573\ Whether there were differences between
                households in such locations compared with those in locations farther
                from these transportation facilities was analyzed.\3574\ Other
                variables, such as land use category, region of country, and housing
                type, were included. Homes with a nonwhite householder were found to be
                22 to 34 percent more likely to be located within 300 feet of these
                large transportation facilities than homes with white householders.
                Homes with a Hispanic householder were 17 to 33 percent more likely to
                be located within 300 feet of these large transportation facilities
                than homes with non-Hispanic householders. Households near large
                transportation facilities were, on average, lower in income and
                educational attainment, more likely to be a rental property, and more
                likely to be located in an urban area compared with households more
                distant from transportation facilities.
                ---------------------------------------------------------------------------
                 \3573\ This variable primarily represents roadway proximity.
                According to the Central Intelligence Agency's World Factbook, in
                2010, the United States had 6,506,204 km of roadways, 224,792 km of
                railways, and 15,079 airports. Highways thus represent the
                overwhelming majority of transportation facilities described by this
                factor in the AHS.
                 \3574\ Bailey, C. (2011) Demographic and Social Patterns in
                Housing Units Near Large Highways and other Transportation Sources.
                Memorandum to docket.
                ---------------------------------------------------------------------------
                 In examining schools near major roadways, the Common Core of Data
                (CCD) from the U.S. Department of Education, which includes information
                on all public elementary and secondary schools and school districts
                nationwide, was examined.\3575\ To determine school proximities to
                major roadways, a geographic information system (GIS) to map each
                school and roadways based on the U.S. Census's TIGER roadway file was
                used.\3576\ Minority students were found to be overrepresented at
                schools within 200 meters of the largest roadways, and schools within
                200 meters of the largest roadways also had higher than expected
                numbers of
                [[Page 25261]]
                students eligible for free or reduced-price lunches. For example, Black
                students represent 22 percent of students at schools located within 200
                meters of a primary road, whereas Black students represent 17 percent
                of students in all U.S. schools. Hispanic students represent 30 percent
                of students at schools located within 200 meters of a primary road,
                whereas Hispanic students represent 22 percent of students in all U.S.
                schools.
                ---------------------------------------------------------------------------
                 \3575\ http://nces.ed.gov/ccd/.
                 \3576\ Pedde, M.; Bailey, C. Identification of Schools within
                200 Meters of U.S. Primary and Secondary Roads. Memorandum to the
                docket (2011).
                ---------------------------------------------------------------------------
                 Overall, there is substantial evidence that the population who
                lives or attends school near major roadways are more likely to be
                minority or low income. As described in Section VII.A.4.c.3.b.i, total
                downstream (tailpipe) emissions of criteria and toxic air pollutants
                for cars and light trucks in 2035 are projected to remain relatively
                unchanged or decrease under all action alternatives compared to the No
                Action Alternative, with the following exceptions: total downstream
                emissions of SO2 would increase under all action
                alternatives under both the CAFE and CO2 programs; total
                downstream emissions of acrolein would increase under Alternatives 5,
                6, and 7 under the CAFE program (but not under the CO2
                program); and total downstream emissions of acetaldehyde and butadiene
                would increase under Alternatives 6 and 7 under the CAFE program (but
                not under the CO2 program). To the extent minority and low-
                income populations disproportionately live or attend schools near major
                roadways, these populations may be more likely to be affected by these
                emissions. However, because some pollutant emissions are expected to
                decrease and others are expected to increase, health impacts are mixed.
                Overall, as the magnitude of the emissions changes is anticipated to be
                minor compared to total tailpipe emissions for these vehicles, the
                impacts to minority or low-income populations are not considered high
                and adverse.
                 The agencies used the standards that were discussed in the 2012
                rulemaking as the baseline for this rulemaking. Therefore, the agencies
                project increases in certain air pollutants for purposes of this
                analysis. However, as discussed above, one impact of the standards
                finalized in this rulemaking is to reduce the up-front cost of new and
                used vehicles. Low income populations may benefit most from the
                reduction in cost of acquiring newer vehicles, which generally are more
                fuel efficient and have lower air pollutant emissions than older
                vehicles. This cost reduction may have the effect of encouraging the
                quicker adoption of cleaner vehicles in low income communities, which
                could result in air quality and health benefits for those who live or
                attend school in proximity to the roadways where they are operated. To
                the degree to which minority populations may also live in proximity to
                these roadways, they would also experience benefits, thereby mitigating
                the disparity in racial, ethnic, and economically based exposures.
                c) Other Vulnerabilities to Climate Change and Health Impacts of Air
                Pollutants
                 Some areas most vulnerable to climate change tend to have a higher
                concentration of minority and low-income populations, potentially
                putting these communities at higher risk from climate variability and
                climate-related extreme weather events.\3577\ For example, urban areas
                tend to have pronounced social inequities that could result in
                disproportionately larger minority and low-income populations than
                those in the surrounding nonurban areas.\3578\ Urban areas are also
                subject to the most substantial temperature increases from climate
                change because of the urban heat island
                effect.3579 3580 3581 Taken together, these tendencies
                demonstrate a potential for disproportionate impacts on minority and
                low-income populations in urban areas. Low-income populations in
                coastal urban areas, which are vulnerable to increases in flooding as a
                result of projected sea-level rise, larger storm surges, and human
                settlement in floodplains, could also be disproportionately affected by
                climate change because they are less likely to have the means to
                evacuate quickly in the event of a natural disaster and, therefore, are
                at greater risk of injury and loss of life.3582 3583
                ---------------------------------------------------------------------------
                 \3577\ U.S. Global Change Research Program (GCRP). Global
                Climate Change Impacts in the United States: The Third National
                Climate Assessment. U.S. Global Change Research Program. Melillo,
                J.M, T.C. Richmond, and G.W. Yohe (Eds.)]. U.S. Government Printing
                Office: Washington, DC 841 pp (2014). doi:10.7930/J0Z31WJ2.
                Available at: http://nca2014.globalchange.gov/report. (Accessed:
                February 27, 2018).
                 \3578\ GCRP (2014).
                 \3579\ GCRP (2014).
                 \3580\ Knowlton, K., B. Lynn, R.A. Goldberg, C. Rosenzweig, C.
                Hogrefe, J.K. Rosenthal, and P.L. Kinney. Projecting Heat-related
                Mortality Impacts under a Changing Climate in the New York City
                Region. American Journal of Public Health 97(11):2028-34 (2007).
                doi:10.2105/AJPH.2006.102947. Available in: http://ajph.aphapublications.org/cgi/content/full/97/11/2028. Last
                accessed: March 4, 2018.
                 \3581\ EPA. Heat Island Effect. U.S. Environmental Protection
                Agency (2017). Last revised: February 20, 2018. Available at:
                https://www.epa.gov/heat-islands. (Accessed: February 28, 2018.).
                 \3582\ GCRP. Global Climate Impacts in the United States (2009).
                Cambridge, United Kingdom and New York, NY, USA. Karl, T.R., J.M.
                Melillo, and T.C. Peterson (Eds.). Cambridge University Press:
                Cambridge, UK. pp. 196.
                 \3583\ GCRP (2014).
                ---------------------------------------------------------------------------
                 Independent of their proximity to pollution sources or climate
                change, locations of potentially high impact, minority and low-income
                populations could be more vulnerable to the health impacts of
                pollutants and climate change. Reports from the U.S. Department of
                Health and Human Services have stated that minority and low-income
                populations tend to have less access to health care services, and the
                services received are more likely to suffer with respect to
                quality.3584 3585 3586 Other studies show that low
                socioeconomic position can modify the health effects of air pollution,
                with higher effects observed in groups with lower socioeconomic
                position.3587 3588 Possible explanations for this
                observation include that low socioeconomic position groups may be
                differentially exposed to air pollution or may be differentially
                vulnerable to effects of exposure.\3589\
                ---------------------------------------------------------------------------
                 \3584\ U.S. Department of Health and Human Services (HHS).
                National Healthcare Disparities Report. U.S. Department of Health
                and Human Service. Rockville, MD, Agency for Healthcare Research and
                Quality (2003). Available at: http://archive.ahrq.gov/qual/nhdr03/nhdr03.htm. (Accessed: March 3, 2018).
                 \3585\ HHS. Minority Health: Recent Findings. Agency for
                Healthcare Research Quality (2013). Last revised: February 2013.
                Available at: https://www.ahrq.gov/research/findings/factsheets/minority/minorfind/index.html. (Accessed: March 3, 2018).
                 \3586\ HHS. 2016 National Healthcare Disparities Report. U.S.
                Department of Health and Human Service (2017). Rockville, MD. Agency
                for Healthcare Research and Quality. Available at: https://www.ahrq.gov/research/findings/nhqrdr/nhqdr16/summary.html.
                (Accessed: September 20, 2017).
                 \3587\ O'Neill, M.S., M. Jerrett, I. Kawachi, J.I. Levy, A.J.
                Cohen, N. Gouveia, P. Wilkinson, T. Fletcher, L. Cifuentes, and J.
                Schwartz. Health, Wealth, and Air Pollution: Advancing Theory and
                Methods. Environmental Health Perspectives 111(16):1861-70 (2003).
                doi: 10.1289/ehp.6334. Available at: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1241758/pdf/ehp0111-001861.pdf. (Accessed: February
                24, 2019).
                 \3588\ Finkelstein, M.M.; Jerrett, M.; DeLuca, P.; Finkelstein,
                N.; Verma, D.K.; Chapman, K.; Sears, M.R. Relation between income,
                air pollution and mortality: a cohort study. Canadian Med Assn J
                169: 397-402 (2003).
                 \3589\ O'Neill et al. (2003).
                ---------------------------------------------------------------------------
                 In terms of climate change, increases in heat-related morbidity and
                mortality because of higher overall and extreme temperatures are likely
                to affect minority and low-income populations disproportionately,
                partially because of limited access to air conditioning and high energy
                costs.3590 3591 3592 3593 Native
                [[Page 25262]]
                American tribes and Alaskan Native villages are also more susceptible
                to the impacts of climate change, as these groups often
                disproportionately rely on natural resources for livelihoods,
                medicines, and cultural and spiritual purposes.\3594\ Moreover, coastal
                tribal communities may have to relocate because of sea-level rise,
                erosion, and permafrost thaw.\3595\ NHTSA's FEIS provides additional
                discussion of health and societal impacts of climate change on
                indigenous communities in Section 8.6.5.2, Sectoral Impacts of Climate
                Change, under Human Health and Human Security.
                ---------------------------------------------------------------------------
                 \3590\ EPA. 2009. Technical Support Document for Endangerment
                and Cause or Contribute Findings for Greenhouse Gases under Section
                202(a) of the Clean Air Act. December 7, 2009. U.S. Environmental
                Protection Agency, Office of Atmospheric Programs, Climate Change
                Division: Washington, DC Available at: https://www.epa.gov/sites/production/files/2016-08/document/endangerment_tsd.pdf. (Accessed:
                February 28, 2018).
                 \3591\ O'Neill, M.S., A. Zanobetti, and J. Schwartz. Disparities
                by Race in Heat-Related Mortality in Four US Cities: The Role of Air
                Conditioning Prevalence. Journal of Urban Health 82(2):191-97
                (2005). doi:10.1093/jurban/jti043. Available at: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3456567/pdf/11524_2006_Article_375.pdf. (Accessed: March 4, 2018).
                 \3592\ GCRP (2014).
                 \3593\ Harlan, S.L. and D.M. Ruddell. Climate Change and Health
                in Cities: Impacts of Heat and Air Pollution and Potential Co-
                Benefits from Mitigation and Adaptation. Current Opinion in
                Environmental Sustainability 3(3):126-34 (2011). doi: 10.1016/
                j.cosust.2011.01.001.
                 \3594\ National Tribal Air Association. 2009. Impacts of climate
                change on Tribes in the United States. Submitted December 11, 2009
                to Assistant Administrator Gina McCarthy, USEPA, Office of Air and
                Radiation. Available at: http://www.epa.gov/air/tribal/pdfs/Impacts%20of%20Climate%20Change%20on%20Tribes%20in%20the%20United%20States.pdf. Last accessed: February 24, 2019.
                 \3595\ Maldonado, J., C. Shearer, R. Bronen, K. Peterson, and H.
                Lazrus. The Impact of Climate Change on Tribal Communities in the
                US: Displacement, Relocation, and Human Rights. Climatic Change
                120(3):601-14 (2013).
                ---------------------------------------------------------------------------
                 Together, this information indicates that the same set of potential
                environmental effects (e.g., air pollutants, heat increases, and sea-
                level rise) may disproportionately affect minority and low-income
                populations because of socioeconomic circumstances or histories of
                discrimination and inequity.
                 As described in Chapter 5 of NHTSA's FEIS, the action alternatives
                are projected to increase CO2 emissions from passenger cars
                and light trucks by 4 to 10 percent by 2100 compared to the No Action
                Alternative. Impacts of climate change could disproportionately affect
                minority and low-income populations in urban areas that are subject to
                the most substantial temperature increases from climate change. These
                impacts are largely because of the urban heat island effect.
                Additionally, minority and low-income populations that live in flood-
                prone coastal areas could be disproportionately affected. However, the
                contribution of the action alternatives to climate change impacts would
                be very minor rather than high and adverse. Compared to the annual U.S.
                CO2 emissions of 7,193 MMTCO2e from all sources
                by the end of the century projected by the GCAM Reference scenario, the
                action alternatives are projected to increase annual U.S.
                CO2 emissions by 0.4 to 1.2 percent in 2100. Compared to
                annual global CO2 emissions, the action alternatives would
                represent an even smaller percentage increase and ultimately, by 2100,
                are projected to result in percentage increases in global mean surface
                temperature, atmospheric CO2 concentrations, and sea level,
                and decreases in ocean pH, ranging from 0.09 percent to less than 0.01
                percent. Any impacts of this rulemaking on low-income and minority
                communities would be attenuated by a lengthy causal chain; but if one
                could attempt to draw those links, the changes to climate values would
                be very small and incremental compared to the expected changes
                associated with the emissions trajectories in the GCAM Reference
                scenario.
                 As reported in Section VII.A.4.c.3.c above, adverse health impacts
                over the lifetimes of vehicles through MY 2029 are projected to
                increase nationwide under each of the action alternatives (except
                Alternative 6 and Alternative 7 under the CAFE program, which show
                decreases) compared to the No Action Alternative. Increases in these
                pollutant emissions, however, would be primarily the result of
                increases in upstream emissions (emissions near refineries, power
                plants, and extraction sites), while downstream emissions (tailpipe
                emissions near roadways) are anticipated to decrease or increase by
                smaller amounts. The health impacts reported in that section occur over
                a long period of time, would be incremental in magnitude, and would not
                be characterized as high. Those impacts would also be borne nationwide,
                so impacts to minority and low-income populations would be smaller.
                d) Conclusion
                 Based on the foregoing, the agencies have determined that this
                rulemaking (and alternatives considered) would not result in
                disproportionately high and adverse human health or environmental
                effects on minority or low-income populations. This rulemaking would
                set standards nationwide, and although minority and low-income
                populations may experience some disproportionate effects, in particular
                locations, the overall impacts on human health and the environment
                would not be ``high and adverse'' under E.O. 12898.
                 Furthermore, the agencies note that there are no mitigation
                measures or alternatives available as part of this action that could
                fulfill the respective statutory missions of the agencies and that
                would address the considerations discussed in Section VIII (e.g.,
                economic practicability) or avoid or reduce any disproportionate
                effects in particular locations experienced by minority and low-income
                populations. The impacts described in this analysis would result from
                air pollutant and CO2 emissions that may occur from the
                levels of stringency selected by the agencies. However, for the reasons
                described in Section VIII, the agencies cannot select a higher level of
                stringency. While the agencies have considered the potential impacts
                described in this analysis, there is a substantial need, based on the
                overall public interest, to address the costs associated with the
                standards discussed in the 2012 rulemaking. More stringent alternatives
                would have severe adverse social and economic costs, as described in
                Section VIII, and necessitate the level of standards finalized in this
                rulemaking.
                12. Executive Order 13045: ``Protection of Children From Environmental
                Health Risks and Safety Risks''
                 This action is subject to E.O. 13045 (62 FR 19885, April 23, 1997)
                because it is an economically significant regulatory action as defined
                by E.O. 12866, and the agencies have reason to believe that the
                environmental health or safety risks related to this action may have a
                disproportionate effect on children. Specifically, children are more
                vulnerable to adverse health effects related to mobile source
                emissions, as well as to the potential long-term impacts of climate
                change. Pursuant to E.O. 13045, NHTSA and EPA must prepare an
                evaluation of the environmental health or safety effects of the planned
                regulation on children and an explanation of why the planned regulation
                is preferable to other potentially effective and reasonably feasible
                alternatives considered by the agencies. Further, this analysis may be
                included as part of any other required analysis.
                 This preamble and NHTSA's Final EIS discuss air quality, climate
                change, and their related environmental and health effects, noting
                where these would disproportionately affect children. The EPA
                Administrator has also discussed the impact of climate-related health
                effects on children in the Endangerment and Cause or Contribute
                Findings for
                [[Page 25263]]
                Greenhouse Gases Under Section 202(a) of the Clean Air Act (74 FR
                66496, December 15, 2009). In addition, this preamble explains why the
                agencies' final standards are preferable to other alternatives
                considered. Together, this preamble and NHTSA's Final EIS satisfy the
                agencies' responsibilities under E.O. 13045.
                F. Regulatory Flexibility Act
                 Pursuant to the Regulatory Flexibility Act (5 U.S.C. 601 et seq.,
                as amended by the Small Business Regulatory Enforcement Fairness Act
                (SBREFA) of 1996), whenever an agency is required to publish a notice
                of proposed rulemaking or final rule, it must prepare and make
                available for public comment a regulatory flexibility analysis that
                describes the effect of the rule on small entities (i.e., small
                businesses, small organizations, and small governmental jurisdictions).
                No regulatory flexibility analysis is required if the head of an agency
                certifies the rule will not have a significant economic impact on a
                substantial number of small entities. SBREFA amended the Regulatory
                Flexibility Act to require Federal agencies to provide a statement of
                the factual basis for certifying that a rule will not have a
                significant economic impact on a substantial number of small entities.
                 Two comments argued that the agencies should prepare a regulatory
                flexibility analysis and convene a small business review panel to
                assess the impacts in accordance with the Regulatory Flexibility Act, 5
                U.S.C. 601 et seq., as amended by SBREFA.\3596\ The agencies considered
                these comments and the impacts of this rule under the Regulatory
                Flexibility Act and certify that this rule will not have a significant
                economic impact on a substantial number of small entities. The
                following is the agencies' statement providing the factual basis for
                this certification pursuant to 5 U.S.C. 605(b).
                ---------------------------------------------------------------------------
                 \3596\ See National Coalition for Advanced Transportation (NCAT)
                Comment, Docket No. NHTSA-2018-0067-11969, at 64-65; Workhorse
                Group, Inc. Comment, Docket No. NHTSA-2018-0067-12215, at 1-2.
                ---------------------------------------------------------------------------
                 Small businesses are defined based on the North American Industry
                Classification System (NAICS) code.\3597\ One of the criteria for
                determining size is the number of employees in the firm. For
                establishments primarily engaged in manufacturing or assembling
                automobiles, as well as light duty trucks, the firm must have less than
                1,500 employees to be classified as a small business. This rule would
                affect motor vehicle manufacturers. As shown in Table X-1, the agencies
                have identified 15 small manufacturers of passenger cars, light trucks,
                and SUVs of electric, hybrid, and internal combustion engines.\3598\
                The agencies acknowledge that some newer manufacturers may not be
                listed. However, those new manufacturers tend to have transportation
                products that are not part of the light-duty vehicle fleet and have yet
                to start production of light-duty vehicles. Moreover, NHTSA does not
                believe that there are a ``substantial number'' of these newer
                companies.\3599\
                ---------------------------------------------------------------------------
                 \3597\ Classified in NAICS under Subsector 336--Transportation
                Equipment Manufacturing for Automobile Manufacturing (336111), Light
                Truck (336112), and Heavy Duty Truck Manufacturing (336120). https://www.sba.gov/document/support-table-size-standards.
                 \3598\ Two comments pointed out that Workhorse Group Inc. was
                not listed as a small domestic vehicle manufacturer in Table XII-1
                of the proposal. See National Coalition for Advanced Transportation
                (NCAT) Comment, Docket No. NHTSA-2018-0067-11969, at 64-65;
                Workhorse Group, Inc. Comment, Docket No. NHTSA-2018-0067-12215, at
                1-2. Workhorse Group has been added to the table here, but neither
                its addition nor the existence of a small number of other new small
                manufacturers does not alter the conclusion that this rule will not
                have a significant economic impact on a substantial number of small
                entities.
                 \3599\ 5 U.S.C. 605(b).
                 [GRAPHIC] [TIFF OMITTED] TR30AP20.758
                
                [[Page 25264]]
                 NHTSA believes that the rulemaking would not have a significant
                economic impact on the small vehicle manufacturers because under 49 CFR
                part 525, passenger car manufacturers making less than 10,000 vehicles
                per year can petition NHTSA to have alternative standards set for those
                manufacturers. These manufacturers do not currently meet the 27.5 mpg
                standard and must already petition the agency for relief. If the
                standard is raised, it has no meaningful impact on these
                manufacturers--they still must go through the same process and petition
                for relief. Given there already is a mechanism for relieving burden on
                small businesses, which is the purpose of the Regulatory Flexibility
                Act, a regulatory flexibility analysis was not prepared.
                ---------------------------------------------------------------------------
                 \3600\ Estimated number of employees as of 2018, source:
                Linkedin.com.
                 \3601\ Rough estimate of light duty vehicle production for model
                year 2017.
                ---------------------------------------------------------------------------
                 Two comments argued that small manufacturers of electric vehicles
                would face a significant economic impact because their ability to earn
                credits would be ``substantially diminished.'' \3602\ The method for
                earning credits applies equally across manufacturers and does not place
                small entities at a significant competitive disadvantage. In any event,
                even if the rule had a ``significant economic impact'' on these small
                EV manufacturers, the amount of these companies is not ``a substantial
                number.'' \3603\ For these reasons, their existence does not alter the
                agencies' analysis of the applicability of the Regulatory Flexibility
                Act. EPA believes this rulemaking would not have a significant economic
                impact on a substantial number of small entities under the Regulatory
                Flexibility Act, as amended by the Small Business Regulatory
                Enforcement Fairness Act. EPA is exempting from the CO2
                standards any manufacturer, domestic or foreign, meeting SBA's size
                definitions of small business as described in 13 CFR 121.201. EPA
                adopted the same type of exemption for small businesses in the 2017 and
                later rulemaking. EPA estimates that small entities comprise less than
                0.1 percent of total annual vehicle sales and exempting them will have
                a negligible impact on the CO2 emissions reductions from the
                standards. Because EPA is exempting small businesses from the
                CO2 standards, the agency certifies that the rule will not
                have a significant economic impact on a substantial number of small
                entities. Therefore, EPA has not conducted a Regulatory Flexibility
                Analysis or a SBREFA SBAR Panel for the rule.
                ---------------------------------------------------------------------------
                 \3602\ National Coalition for Advanced Transportation (NCAT)
                Comment, Docket No. NHTSA-2018-0067-11969, at 65; Workhorse Group,
                Inc. Comment, Docket No. NHTSA-2018-0067-12215, at 2.
                 \3603\ 5 U.S.C. 605.
                ---------------------------------------------------------------------------
                 EPA regulations allow small businesses voluntarily to waive their
                small business exemption and optionally to certify to the
                CO2 standards. This option allows small entity manufacturers
                to earn CO2 credits under the CO2 program, if
                their actual fleetwide CO2 performance is better than their
                fleetwide CO2 target standard. However, the exemption waiver
                is optional for small entities and thus the agency believes that
                manufacturers opt into the CO2 program if it is economically
                advantageous for them to do so, for example in order to generate and
                sell CO2 credits. Therefore, EPA believes this voluntary
                option does not affect EPA's determination that the standards will
                impose no significant adverse impact on small entities.
                G. Executive Order 13132 (Federalism)
                 Executive Order 13132 requires Federal agencies to develop an
                accountable process to ensure ``meaningful and timely input by State
                and local officials in the development of regulatory policies that have
                federalism implications.'' The Order defines the term ``[p]olicies that
                have federalism implications'' to include regulations that have
                ``substantial direct effects on the States, on the relationship between
                the national government and the States, or on the distribution of power
                and responsibilities among the various levels of government.'' Under
                the Order, agencies may not issue a regulation that has federalism
                implications, that imposes substantial direct compliance costs, unless
                the Federal government provides the funds necessary to pay the direct
                compliance costs incurred by State and local governments, or the
                agencies consult with State and local officials early in the process of
                developing the proposed regulation. The agencies complied with the
                Order's requirements.
                 NHTSA also addressed the federalism implications of its proposal in
                The Safer Affordable Fuel-Efficient Vehicles Rule Part One: One
                National Program final rulemaking.\3604\
                ---------------------------------------------------------------------------
                 \3604\ 84 FR 51310 (Sep. 27, 2019).
                ---------------------------------------------------------------------------
                H. Executive Order 12988 (Civil Justice Reform)
                 Pursuant to Executive Order 12988, ``Civil Justice Reform,'' \3605\
                NHTSA has considered whether this rulemaking would have any retroactive
                effect. This proposed rule does not have any retroactive effect.
                ---------------------------------------------------------------------------
                 \3605\ 61 FR 4729 (Feb. 7, 1996).
                ---------------------------------------------------------------------------
                I. Executive Order 13175 (Consultation and Coordination With Indian
                Tribal Governments)
                 This final rule does not have tribal implications, as specified in
                Executive Order 13175 (65 FR 67249, November 9, 2000). This rule will
                be implemented at the Federal level and impose compliance costs only on
                vehicle manufacturers. Thus, Executive Order 13175 does not apply to
                this rule. Some comments complained that the agencies have not
                consulted or coordinated with Native American communities and Indian
                Tribes in promulgating this rule.\3606\ Executive Order 13175 requires
                consultation with Tribal officials when agencies are developing
                policies that have ``substantial direct effects'' on Tribes and Tribal
                interests.\3607\ Even accepting the comments' description of the
                effects of the rule, they have identified only indirect effects of the
                standards on Tribal interests.\3608\
                ---------------------------------------------------------------------------
                 \3606\ See, e.g., CARB Comment, Docket No. NHTSA-2018-0067-
                11873, at 412; National Tribal Air Association Comment, Docket No.
                NHTSA-2018-0067-11948, at 4; Keweenaw Bay Indian Community Comment,
                Docket No. EPA-HQ-OAR-2018-0283-3325, at 1-2; Fond du Lac Band of
                Lake Superior Chippewa Comment, Docket No. EPA-HQ-OAR-2018-0283-
                4030, at 3; Sac and Fox Nation, Docket No. EPA-HQ-OAR-2018-0283-
                4159, at 4-5; The Leech Lake Band of Ojibwe Comment, Docket No. EPA-
                HQ-OAR-2018-0283-5931, at 4-5.
                 \3607\ 65 FR 67249, 67249 (Nov. 6, 2000).
                 \3608\ See, e.g., National Tribal Air Association Comment,
                Docket No. NHTSA-2018-0067-11948, at 4.
                ---------------------------------------------------------------------------
                J. Unfunded Mandates Reform Act
                 Section 202 of the Unfunded Mandates Reform Act of 1995 (UMRA)
                requires Federal agencies to prepare a written assessment of the costs,
                benefits, and other effects of a proposed or final rule that includes a
                Federal mandate likely to result in the expenditure by State, local, or
                Tribal governments, in the aggregate, or by the private sector, of more
                than $100 million in any one year (adjusted for inflation with base
                year of 1995). Adjusting this amount by the implicit gross domestic
                product price deflator for 2016 results in $148 million (111.416/75.324
                = 1.48).\3609\ Before promulgating a rule for which a written statement
                is needed, section 205 of UMRA generally requires NHTSA and EPA to
                identify and consider a reasonable number of regulatory
                [[Page 25265]]
                alternatives and adopt the least costly, most cost-effective, or least
                burdensome alternative that achieves the objective of the rule. The
                provisions of section 205 do not apply when they are inconsistent with
                applicable law. Moreover, section 205 allows NHTSA and EPA to adopt an
                alternative other than the least costly, most cost-effective, or least
                burdensome alternative if the agency publishes with the rule an
                explanation of why that alternative was not adopted.
                ---------------------------------------------------------------------------
                 \3609\ Bureau of Economic Analysis, National Income and Product
                Accounts (NIPA), Table 1.1.9 Implicit Price Deflators for Gross
                Domestic Product. https://bea.gov/iTable/index_nipa.cfm.
                ---------------------------------------------------------------------------
                 This rule will not result in the expenditure by State, local, or
                Tribal governments, in the aggregate, of more than $148 million
                annually, but it will result in the expenditure of that magnitude by
                vehicle manufacturers and/or their suppliers. In developing this rule,
                NHTSA and EPA considered a variety of alternative average fuel economy
                standards lower and higher than those previously proposed. The fuel
                economy standards for MYs 2021-2026 are the least costly, most cost-
                effective, and least burdensome alternative that achieve the objectives
                of the rule.
                K. Regulation Identifier Number
                 The Department of Transportation assigns a regulation identifier
                number (RIN) to each regulatory action listed in the Unified Agenda of
                Federal Regulations. The Regulatory Information Service Center
                publishes the Unified Agenda in April and October of each year. The RIN
                contained in the heading at the beginning of this document may be used
                to find this action in the Unified Agenda.
                L. National Technology Transfer and Advancement Act
                 Section 12(d) of the National Technology Transfer and Advancement
                Act (NTTAA) requires NHTSA and EPA to evaluate and use existing
                voluntary consensus standards in its regulatory activities unless doing
                so would be inconsistent with applicable law (e.g., the statutory
                provisions regarding NHTSA's vehicle safety authority, or EPA's testing
                authority) or otherwise impractical.\3610\
                ---------------------------------------------------------------------------
                 \3610\ 15 U.S.C. 272.
                ---------------------------------------------------------------------------
                 Voluntary consensus standards are technical standards developed or
                adopted by voluntary consensus standards bodies. Technical standards
                are defined by the NTTAA as ``performance-based or design-specific
                technical specification and related management systems practices.''
                They pertain to ``products and processes, such as size, strength, or
                technical performance of a product, process or material.''
                 Examples of organizations generally regarded as voluntary consensus
                standards bodies include the American Society for Testing and Materials
                (ASTM), the Society of Automotive Engineers (SAE), and the American
                National Standards Institute (ANSI). If the agencies do not use
                available and potentially applicable voluntary consensus standards,
                they are required by the Act to provide Congress, through OMB, an
                explanation of the reasons for not using such standards.
                 For CO2 emissions, EPA will collect data over the same
                tests that are used for the MY 2012-2016 CO2 standards and
                for the CAFE program. This unified data collection will minimize the
                amount of testing done by manufacturers because manufacturers are
                already required to run these tests. For A/C credits, EPA will use a
                consensus methodology developed by the Society of Automotive Engineers
                (SAE) and also a new A/C test. EPA knows of no consensus standard
                available for the A/C test.
                 There are currently no voluntary consensus standards that NHTSA
                administers relevant to today's CAFE standards.
                M. Department of Energy Review
                 In accordance with 49 U.S.C. 32902(j)(2), NHTSA submitted this rule
                to the Department of Energy for review.
                N. Paperwork Reduction Act
                 The Paperwork Reduction Act (PRA) of 1995, Public Law 104-13,\3611\
                gives OMB authority to regulate matters regarding the collection,
                management, storage, and dissemination of certain information by and
                for the Federal government. It seeks to reduce the total amount of
                paperwork handled by the government and the public. NHTSA strives to
                reduce the public's information collection burden hours each fiscal
                year by streamlining external and internal processes.
                ---------------------------------------------------------------------------
                 \3611\ Codified at 44 U.S.C. 3501 et seq.
                ---------------------------------------------------------------------------
                 To this end, NHTSA will continue to collect information to ensure
                compliance with its CAFE program. NHTSA will reinstate its previously-
                approved collection of information for Corporate Average Fuel Economy
                (CAFE) reports specified in 49 CFR part 537 (OMB control number 2127-
                0019), add the additional burden for reporting changes adopted in the
                October 15, 2012 final rule that recently came into effect (see 77 FR
                62623), and account for the change in burden in this rule as well as
                for other CAFE reporting provisions required by Congress and NHTSA.
                NHTSA is also changing the name of this collection to represent more
                accurately the breadth of all CAFE regulatory reporting. Although NHTSA
                is adding additional burden hours to its CAFE report requirement in 49
                CFR 537, the agency believes there will be a reduction in the overall
                paperwork burden due to the standardization of data and the streamlined
                process.
                 In compliance with the PRA, the information collection request
                (ICR) abstracted below was forwarded to OMB for review and comment. The
                ICR describes the nature of the information collection and its expected
                burden.
                 Title: Corporate Average Fuel Economy.
                 Type of Request: Reinstatement and amendment of a previously
                approved collection.
                 OMB Control Number: 2127-0019.
                 Form Numbers: NHTSA Form 1474 (CAFE Projections Reporting Template)
                and NHTSA Form 1475 (CAFE Credit Template).
                 Requested Expiration Date of Approval: Three years from date of
                approval.
                 Summary of the collection of information: As part of this
                rulemaking, NHTSA is reinstating and modifying its previously-approved
                collection for CAFE-related collections of information. NHTSA and EPA
                have coordinated their compliance and reporting requirements in an
                effort not to impose duplicative burdens on regulated entities. This
                information collection contains three different components: Burden
                related to NHTSA's CAFE reporting requirements; burden related to CAFE
                compliance, but not via reporting requirements; and information
                gathered by NHTSA to help inform CAFE analyses. All templates
                referenced in this section will be available in the rulemaking docket
                and the NHTSA public information center.\3612\
                ---------------------------------------------------------------------------
                 \3612\ https://one.nhtsa.gov/cafe_pic/CAFE_PIC_Home.htm.
                ---------------------------------------------------------------------------
                CAFE Compliance Reports
                 NHTSA is reinstating \3613\ its collection related to the reporting
                requirements in 49 U.S.C. 32907, ``Reports and tests of
                manufacturers.'' In that section, manufacturers are statutorily
                required to submit CAFE compliance reports to the Secretary of
                Transportation.\3614\ The reports must state if a manufacturer will
                comply with its applicable fuel economy standard(s), describe what
                actions the manufacturer
                [[Page 25266]]
                intends to take to comply with the standard(s), and include other
                information as required by NHTSA. Manufacturers are required to submit
                two CAFE compliance reports--a pre-model year report (PMY) and a mid-
                model year (MMY) report--each year. In the event a manufacturer needs
                to correct previously-submitted information, a manufacturer may need to
                file additional reports.\3615\
                ---------------------------------------------------------------------------
                 \3613\ This collection expired on April 30, 2016.
                 \3614\ 49 U.S.C. 32907 (delegated to the NHTSA Administrator at
                49 CFR 1.95). Because of this delegation, for purposes of
                discussion, statutory references to the Secretary of Transportation
                in this section will be discussed in terms of NHTSA or the NHTSA
                Administrator.
                 \3615\ Specifically, a manufacturer shall submit a report
                containing the information during the 30 days before the beginning
                of each model year, and during the 30 days beginning the 180th day
                of the model year. When a manufacturer decides that actions reported
                are not sufficient to ensure compliance with that standard, the
                manufacturer shall report additional actions it intends to take to
                comply with the standard and include a statement about whether those
                actions are sufficient to ensure compliance.
                ---------------------------------------------------------------------------
                 To implement this statute, NHTSA issued 49 CFR part 537,
                ``Automotive Fuel Economy Reports,'' which adds additional definition
                to the terms of section 32907. The first report, the PMY report must be
                submitted to NHTSA before December 31 of the calendar year prior to the
                corresponding model year and contain manufacturers' projected
                information for that upcoming model year. The second report, the MMY
                report must be submitted by July 31 of the given model year and contain
                updated information from manufacturers based on actual and projected
                information known midway through the model year. Finally, the last
                report, a supplementary report, is required to be submitted anytime a
                manufacturer needs to correct information previously submitted to
                NHTSA.
                 Compliance reports must include information on passenger and non-
                passenger automobiles (trucks) describing the projected and actual fuel
                economy standards, fuel economy performance values, production sales
                volumes and information on vehicle design features (e.g., engine
                displacement and transmission class) and other vehicle attribute
                characteristics (e.g., track width, wheel base, and other light truck
                off-road features). Manufacturers submit confidential and non-
                confidential versions of these reports to NHTSA. Confidential reports
                differ by including estimated or actual production sales information,
                which is withheld from public disclosure to protect each manufacturer's
                competitive sales strategies. NHTSA uses the reports as the basis for
                vehicle auditing and testing, which helps manufacturers correct
                reporting errors prior to the end of the model year and facilitate
                acceptance of their final CAFE report by the Environmental Protection
                Agency (EPA). The reports also help the agency, as well as the
                manufacturers who prepare them, anticipate potential compliance issues
                as early as possible, and help manufacturers plan their compliance
                strategies.
                 Further, NHTSA is modifying this collection to account for
                additional information manufacturers are required to include in their
                reports. In the CAFE standards previously promulgated for MY 2017 and
                beyond,\3616\ NHTSA allowed for manufacturers to gain additional fuel
                economy benefits by installing certain technologies on their vehicles
                beginning with MY 2017.\3617\ These technologies include air-
                conditioning systems with increased efficiency, off-cycle technologies
                whose benefits are not adequately captured on the Federal Test
                Procedure and/or the Highway Fuel Economy Test,\3618\ and hybrid
                electric technologies installed on full-size pickup trucks. Prior to MY
                2017, manufacturers were unable to earn a fuel economy benefit for
                these technologies, so NHTSA's reporting requirements did not include
                an opportunity to report them. Now, manufacturers must provide
                information on these technologies in their CAFE reports. NHTSA requires
                manufacturers to provide detailed information on the model types using
                these technologies to gain fuel economy benefits. These details are
                necessary to facilitate NHTSA's technical analyses and to ensure the
                agency can perform random enforcement audits when necessary.
                ---------------------------------------------------------------------------
                 \3616\ 77 FR 62623 (Oct. 15, 2012).
                 \3617\ These technologies were not included in the burden for
                part 537 at the time as the additional reporting requirements would
                not take effect until years later.
                 \3618\ E.g., engine idle stop-start systems, active transmission
                warmup systems, etc.
                ---------------------------------------------------------------------------
                 In addition to a list of all fuel consumption improvement
                technologies utilized in their fleet, 49 CFR 537 requires manufacturers
                to report the make, model type, compliance category, and production
                volume of each vehicle equipped with each technology and the associated
                fuel consumption improvement value (FCIV). NHTSA is adding the
                reporting and enforcement burden hours and cost for these new
                incentives to this collection. Manufacturers can also petition the EPA
                and NHTSA, in accordance with 40 CFR 86.1868-12 or 40 CFR 86.1869-12,
                to gain additional credits based upon the improved performance of any
                of the new incentivized technologies allowed starting in model year
                2017. EPA approves these petitions in collaboration with NHTSA and any
                adjustments are taken into account for both programs. As a part the
                agencies' coordination, NHTSA provides EPA with an evaluation of each
                new technology to ensure its direct impact on fuel economy and an
                assessment on the suitability of each technology for use in increasing
                a manufacturer's fuel economy performance. Furthermore, at times, NHTSA
                may independently request additional information from a manufacturer to
                support its evaluations. This information along with any research
                conclusions shared with EPA and NHTSA in the petitions is required to
                be submitted in manufacturer's CAFE reports.
                 NHTSA is also changing the burden hours for its CAFE reporting
                requirements in 49 CFR part 537 by adjusting the total amount of time
                spent collecting the required reporting information through the use of
                a standardized reporting template to streamline the collection process.
                The standardized template will be used by manufacturers to collect all
                the required CAFE information under 49 CFR 537.7(b) and (c) and
                provides a format which ensures accuracy, completeness, and better
                alignment with the final data provided to EPA.
                2. Other CAFE Compliance Collections
                 NHTSA is adopting a new standardized template for manufacturers
                buying CAFE credits and for manufacturers submitting credit
                transactions in accordance with 49 CFR part 536. In 49 CFR part
                536.5(d), NHTSA is required to assess compliance with fuel economy
                standards each year, utilizing the certified and reported CAFE data
                provided by the EPA for enforcement of the CAFE program pursuant to 49
                U.S.C. 32904(e). Credit values are calculated based on the CAFE data
                from the EPA. If a manufacturer's vehicles in a particular compliance
                category performs better than its required fuel economy standard, NHTSA
                adds credits to the manufacturer's account for that compliance
                category. If a manufacturer's vehicles in a particular compliance
                category perform worse than the required fuel economy standard, NHTSA
                will add a credit deficit to the manufacturer's account and will
                provide written notification to the manufacturer concerning its failure
                to comply. The manufacturer will be required to confirm the shortfall
                and must either: Submit a plan indicating how it will allocate existing
                credits or earn, transfer, and/or acquire credits or pay the equivalent
                civil penalty. The manufacturer must submit a plan or
                [[Page 25267]]
                payment within 60 days of receiving notification from NHTSA.
                 Manufacturers should use the credit transaction template any time a
                credit transaction request is sent to NHTSA. For example, manufacturers
                that purchase credits and want to apply them to their credit accounts
                will use the credit transaction template. The template NHTSA is
                adopting is a simple spreadsheet that credit entities fill out. When
                completed, credit entities will have an organized list of credit
                transactions and will be able to click a button on the spreadsheet to
                generate a joint transaction letter for trading parties to sign and
                submit to NHTSA, along with the spreadsheet. Entities trading credits
                are also required to provide to NHTSA all the confidential information
                associated with the monetary and non-monetary price of credit trades.
                NHTSA believes these changes will significantly reduce the burden on
                manufacturers in managing their CAFE credit accounts and provide better
                oversight of the CAFE credit program for NHTSA.
                 Finally, NHTSA is accounting for the additional burden due to
                existing CAFE program elements. In 49 CFR part 525, small volume
                manufacturers submit petitions to NHTSA for exemption from an
                applicable average fuel economy standard and to request to comply with
                a less stringent alternative average fuel economy standard. In 49 CFR
                part 534, manufacturers are required to submit information to NHTSA
                when establishing a corporate controlled relationship with another
                manufacturer. A controlled relationship exists between manufacturers
                that control, are controlled by, or are under common control with, one
                or more other manufacturers. Accordingly, manufacturers that have
                entered into written contracts transferring rights and responsibilities
                to other manufacturers in controlled relationships for CAFE purposes
                are required to provide reports to NHTSA. There are additional
                reporting requirements for manufacturers submitting carry back plans
                and when manufacturers split apart from controlled relationships and
                must designate how credits are to be allocated between the
                parties.\3619\ Manufacturers with credit deficits at the end of the
                model year, can carry back future earned credits up to three model
                years in advance of the deficit to resolve a current shortfall. The
                carryback plan proving the existence of a manufacturer's future earned
                credits must be submitted and approved by NHTSA, pursuant to 49 U.S.C.
                32903(b).
                ---------------------------------------------------------------------------
                 \3619\ See 49 CFR part 536.
                ---------------------------------------------------------------------------
                3. Analysis Fleet Composition
                 As discussed in Section VI.B, in setting CAFE standards, NHTSA
                creates an analysis fleet from which to model potential future economy
                improvements. To compose this fleet, the agency uses a mixture of
                compliance data and information from other sources to replicate more
                closely the fleet from a recent model year. While refining the analysis
                fleet, NHTSA occasionally asks manufacturers for information that is
                similar to information submitted as part of EPA's final model year
                report (e.g., final model year vehicle volumes). Periodically, NHTSA
                may ask manufacturers for more detailed information than what is
                required for compliance (e.g., what engines are shared across vehicle
                models). Often, NHTSA requests this information from manufacturers
                after manufacturers have submitted their final model year reports to
                EPA, but before EPA processes and releases final model year reports.
                 Information like this, which is used to verify and supplement the
                data used to create the analysis fleet, is tremendously valuable to
                generating an accurate analysis fleet, and setting maximum feasible
                standards. The more accurate the analysis fleet is, the more accurate
                the modeling of what technologies could be applied will be. Therefore,
                NHTSA is accounting for the burden on manufacturers to provide the
                agency with this additional information. In almost all instances,
                manufacturers already have the information NHTSA seeks, but it might
                need to be reformatted or recompiled. Because of this, NHTSA believes
                the burden to provide this information will often be minimal.
                 Affected Public: Respondents are manufacturers of engines and
                vehicles within the North American Industry Classification System
                (NAICS) and use the coding structure as defined by NAICS including
                codes 33611, 336111, 336112, 33631, 33631, 33632, 336320, 33635, and
                336350 for motor vehicle and parts manufacturing.
                 Respondent's obligation to respond: Regulated entities are required
                to respond to inquiries covered by this collection. 49 U.S.C. 32907. 49
                CFR part 525, 534, 536, and 537.
                 Frequency of response: Variable, based on compliance obligation.
                Please see PRA supporting documentation in the docket for more detailed
                information.
                 Average burden time per response: Variable, based on compliance
                obligation. Please see PRA supporting documentation in the docket for
                more detailed information.
                 Number of respondents: 23.
                4. Estimated Total Annual Burden Hours and Costs:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.759
                [GRAPHIC] [TIFF OMITTED] TR30AP20.760
                [[Page 25268]]
                O. Privacy Act
                 In accordance with 5 U.S.C. 553(c), the agencies solicited comments
                from the public to inform the rulemaking process better. These comments
                are posted, without edit, to www.regulations.gov, as described in DOT's
                system of records notice, DOT/ALL-14 FDMS, accessible through
                www.transportation.gov/privacy. In order to facilitate comment tracking
                and response, the agencies encouraged commenters to provide their
                names, or the names of their organizations; however, submission of
                names is completely optional.
                List of Subjects
                40 CFR Part 86
                 Administrative practice and procedure, Confidential business
                information, Incorporation by reference, Labeling, Motor vehicle
                pollution, Reporting and recordkeeping requirements.
                40 CFR Part 600
                 Administrative practice and procedure, Electric power, Fuel
                economy, Labeling, Reporting and recordkeeping requirements.
                49 CFR Parts 523, 531, and 533
                 Fuel economy.
                49 CFR Parts 536 and 537
                 Fuel economy, Reporting and recordkeeping requirements.
                Environmental Protection Agency
                40 CFR Chapter I
                 For the reasons set forth in the preamble, the Environmental
                Protection Agency is amending part 86 of title 40, Chapter I of the
                Code of Federal Regulations as follows:
                PART 86--CONTROL OF EMISSIONS FROM NEW AND IN-USE HIGHWAY VEHICLES
                AND ENGINES
                0
                1. The authority citation for part 86 continues to read as follows:
                 Authority: 42 U.S.C. 7401-7671q.
                0
                2. Section 86.1818-12 is amended by revising paragraphs (c)(2)(i)(A)
                through (C) and (c)(3)(i)(A), (B), and (D), to read as follows:
                Sec. 86.1818-12 Greenhouse gas emission standards for light-duty
                vehicles, light-duty trucks, and medium-duty passenger vehicles.
                * * * * *
                 (c) * * *
                 (2) * * *
                 (i) * * *
                 (A) For passenger automobiles with a footprint of less than or
                equal to 41 square feet, the gram/mile CO2 target value
                shall be selected for the appropriate model year from the following
                table:
                ------------------------------------------------------------------------
                 CO2 target
                 Model year value (grams/
                 mile)
                ------------------------------------------------------------------------
                2012.................................................... 244.0
                2013.................................................... 237.0
                2014.................................................... 228.0
                2015.................................................... 217.0
                2016.................................................... 206.0
                2017.................................................... 195.0
                2018.................................................... 185.0
                2019.................................................... 175.0
                2020.................................................... 166.0
                2021.................................................... 161.8
                2022.................................................... 159.0
                2023.................................................... 156.4
                2024.................................................... 153.7
                2025.................................................... 151.2
                2026 and later.......................................... 148.6
                ------------------------------------------------------------------------
                 (B) For passenger automobiles with a footprint of greater than 56
                square feet, the gram/mile CO2 target value shall be
                selected for the appropriate model year from the following table:
                ------------------------------------------------------------------------
                 CO2 target
                 Model year value (grams/
                 mile)
                ------------------------------------------------------------------------
                2012.................................................... 315.0
                2013.................................................... 307.0
                2014.................................................... 299.0
                2015.................................................... 288.0
                2016.................................................... 277.0
                2017.................................................... 263.0
                2018.................................................... 250.0
                2019.................................................... 238.0
                2020.................................................... 226.0
                2021.................................................... 220.9
                2022.................................................... 217.3
                2023.................................................... 213.7
                2024.................................................... 210.2
                2025.................................................... 206.8
                2026 and later.......................................... 203.4
                ------------------------------------------------------------------------
                 (C) For passenger automobiles with a footprint that is greater than
                41 square feet and less than or equal to 56 square feet, the gram/mile
                CO2 target value shall be calculated using the following
                equation and rounded to the nearest 0.1 grams/mile, except that for any
                vehicle footprint the maximum CO2 target value shall be the
                value specified for the same model year in paragraph (c)(2)(i)(B) of
                this section:
                Target CO2 = [a x f] + b
                Where: f is the vehicle footprint, as defined in Sec. 86.1803; and
                a and b are selected from the following table for the appropriate
                model year:
                ------------------------------------------------------------------------
                 Model year a b
                ------------------------------------------------------------------------
                2012.................................... 4.72 50.5
                2013.................................... 4.72 43.3
                2014.................................... 4.72 34.8
                2015.................................... 4.72 23.4
                2016.................................... 4.72 12.7
                2017.................................... 4.53 8.9
                2018.................................... 4.35 6.5
                2019.................................... 4.17 4.2
                2020.................................... 4.01 1.9
                2021.................................... 3.94 0.2
                2022.................................... 3.88 -0.1
                2023.................................... 3.82 -0.4
                2024.................................... 3.77 -0.6
                2025.................................... 3.71 -0.9
                2026 and later.......................... 3.65 -1.2
                ------------------------------------------------------------------------
                * * * * *
                 (3) * * *
                 (i) * * *
                 (A) For light trucks with a footprint of less than or equal to 41
                square feet, the gram/mile CO2 target value shall be
                selected for the appropriate model year from the following table:
                [[Page 25269]]
                ------------------------------------------------------------------------
                 CO2 target
                 Model year value (grams/
                 mile)
                ------------------------------------------------------------------------
                2012.................................................... 294.0
                2013.................................................... 284.0
                2014.................................................... 275.0
                2015.................................................... 261.0
                2016.................................................... 247.0
                2017.................................................... 238.0
                2018.................................................... 227.0
                2019.................................................... 220.0
                2020.................................................... 212.0
                2021.................................................... 206.5
                2022.................................................... 203.0
                2023.................................................... 199.6
                2024.................................................... 196.2
                2025.................................................... 193.2
                2026 and later.......................................... 189.9
                ------------------------------------------------------------------------
                 (B) For light trucks with a footprint that is greater than 41
                square feet and less than or equal to the maximum footprint value
                specified in the table below for each model year, the gram/mile
                CO2 target value shall be calculated using the following
                equation and rounded to the nearest 0.1 grams/mile, except that for any
                vehicle footprint the maximum CO2 target value shall be the
                value specified for the same model year in paragraph (c)(3)(i)(D) of
                this section:
                Target CO2 = (a x f) + b
                Where:
                 f is the footprint, as defined in Sec. 86.1803; and a and b are
                selected from the following table for the appropriate model year:
                ----------------------------------------------------------------------------------------------------------------
                 Maximum
                 Model year footprint a b
                ----------------------------------------------------------------------------------------------------------------
                2012............................................................ 66.0 4.04 128.6
                2013............................................................ 66.0 4.04 118.7
                2014............................................................ 66.0 4.04 109.4
                2015............................................................ 66.0 4.04 95.1
                2016............................................................ 66.0 4.04 81.1
                2017............................................................ 50.7 4.87 38.3
                2018............................................................ 60.2 4.76 31.6
                2019............................................................ 66.4 4.68 27.7
                2020............................................................ 68.3 4.57 24.6
                2021............................................................ 68.3 4.51 21.5
                2022............................................................ 68.3 4.44 20.6
                2023............................................................ 68.3 4.37 20.2
                2024............................................................ 68.3 4.31 19.6
                2025............................................................ 68.3 4.23 19.6
                2026 and later.................................................. 68.3 4.17 19.0
                ----------------------------------------------------------------------------------------------------------------
                * * * * *
                 (D) For light trucks with a footprint greater than the minimum
                value specified in the table below for each model year, the gram/mile
                CO2 target value shall be selected for the appropriate model
                year from the following table:
                ------------------------------------------------------------------------
                 CO2 target
                 Model year Minimum value (grams/
                 footprint mile)
                ------------------------------------------------------------------------
                2012.................................... 66.0 395.0
                2013.................................... 66.0 385.0
                2014.................................... 66.0 376.0
                2015.................................... 66.0 362.0
                2016.................................... 66.0 348.0
                2017.................................... 66.0 347.0
                2018.................................... 66.0 342.0
                2019.................................... 66.4 339.0
                2020.................................... 68.3 337.0
                2021.................................... 68.3 329.4
                2022.................................... 68.3 324.1
                2023.................................... 68.3 318.9
                2024.................................... 68.3 313.7
                2025.................................... 68.3 308.7
                2026 and later.......................... 68.3 303.7
                ------------------------------------------------------------------------
                * * * * *
                0
                3. Section 86.1866-12 is amended by revising paragraph (a)(2), removing
                paragraph (a)(3), and revising (b) introductory text, (b)(1), and
                (b)(2)(i) to read as follows:
                Sec. 86.1866-12 CO2 credits for advanced technology vehicles.
                * * * * *
                 (a) * * *
                 (2) Model years 2017 through 2026: For electric vehicles, plug-in
                hybrid electric vehicles, and fuel cell vehicles produced for U.S.
                sale, where ``U.S.'' means the states and territories of the United
                States, in the 2017 through 2026 model years, such use of zero (0)
                grams/mile CO2 is unrestricted.
                 (b) For electric vehicles, plug-in hybrid electric vehicles, fuel
                cell vehicles, dedicated natural gas vehicles, and dual-fuel natural
                gas vehicles as those terms are defined in Sec. 86.1803-01, that are
                certified and produced for U.S. sale in the specified model years and
                that meet the additional specifications in this section, the
                manufacturer may use the production multipliers in this paragraph (b)
                when determining additional credits for advanced technology vehicles.
                Full size pickup trucks eligible for and using a production multiplier
                are not eligible for the performance-based credits described in Sec.
                86.1870-12(b).
                [[Page 25270]]
                 (1) The production multipliers, by model year, for model year 2017
                through 2021 electric vehicles and fuel cell vehicles are as follows:
                ------------------------------------------------------------------------
                 Production
                 Model year multiplier
                ------------------------------------------------------------------------
                2017.................................................... 2.0
                2018.................................................... 2.0
                2019.................................................... 2.0
                2020.................................................... 1.75
                2021.................................................... 1.5
                ------------------------------------------------------------------------
                 (2)(i) The production multipliers, by model year, for model year
                2017 through 2021 plug-in hybrid electric vehicles and model year 2017
                through 2026 dedicated natural gas vehicles and dual-fuel natural gas
                vehicles are as follows:
                ------------------------------------------------------------------------
                 Production
                 Model year multiplier
                ------------------------------------------------------------------------
                2017.................................................... 1.6
                2018.................................................... 1.6
                2019.................................................... 1.6
                2020.................................................... 1.45
                2021.................................................... 1.3
                2022-2026 (dedicated and dual fuel natural gas vehicles 2.0
                 only)..................................................
                ------------------------------------------------------------------------
                * * * * *
                0
                4. Section 86.1868-12 is amended by adding an entry to the end of the
                table in paragraph (a)(2) and by adding paragraph (h)(7) to read as
                follows:
                Sec. 86.1868-12 CO2 credits for improving the efficiency of air
                conditioning systems.
                * * * * *
                 (a) * * *
                 (2) * * *
                ------------------------------------------------------------------------
                 Passenger
                 Air conditioning technology automobiles Light trucks
                 (g/mi) (g/mi)
                ------------------------------------------------------------------------
                
                 * * * * * * *
                Advanced technology air conditioning 1.1 1.1
                 compressor with improved efficiency
                 relative to fixed-displacement
                 compressors achieved through the
                 addition of a variable crankcase
                 suction valve..........................
                ------------------------------------------------------------------------
                * * * * *
                 (h) * * *
                 (7) Advanced technology air conditioning compressor means an air
                conditioning compressor with improved efficiency relative to fixed-
                displacement compressors. Efficiency gains are derived from improved
                internal valve systems that optimize the internal refrigerant flow
                across the range of compressor operator conditions through the addition
                of a variable crankcase suction valve.
                0
                5. Section 86.1869-12 is amended by revising paragraph (a), by adding
                paragraphs (b)(1)(ix), (b)(1)(x), (b)(4)(xiii) and (b)(4)(xiv), and by
                revising paragraph (d)(2) to read as follows:
                Sec. 86.1869-12 CO2 credits for off-cycle CO2 reducing technologies.
                * * * * *
                 (a) Manufacturers may generate credits for CO2-reducing
                technologies where the CO2 reduction benefit of the
                technology is not adequately captured on the Federal Test Procedure
                and/or the Highway Fuel Economy Test such that the technology would not
                be otherwise installed for purposes of reducing emissions (directly or
                indirectly) over those test cycles for compliance with the GHG
                standards. These technologies must have a measurable, demonstrable, and
                verifiable real-world CO2 reduction that occurs outside the
                conditions of the Federal Test Procedure and the Highway Fuel Economy
                Test. These optional credits are referred to as ``off-cycle'' credits.
                The technologies must not be integral or inherent to the basic vehicle
                design, such as engine, transmission, mass reduction, passive
                aerodynamic design, and tire technologies. Technologies installed for
                non-off-cycle emissions related reasons are also not eligible as they
                would be considered part of the baseline vehicle design. The technology
                must not be inherent to the design of occupant comfort and
                entertainment features except for technologies related to reducing
                passenger air conditioning demand and improving air conditioning system
                efficiency. Notwithstanding the provisions of this paragraph (a), off-
                cycle menu technologies included in paragraph (b) of this section
                remain eligible for credits. Off-cycle technologies used to generate
                emission credits are considered emission-related components subject to
                applicable requirements and must be demonstrated to be effective for
                the full useful life of the vehicle. Unless the manufacturer
                demonstrates that the technology is not subject to in-use
                deterioration, the manufacturer must account for the deterioration in
                their analysis. Durability evaluations of off-cycle technologies may
                occur at any time throughout a model year, provided that the results
                can be factored into the data provided in the model year report. Off-
                cycle credits may not be approved for crash-avoidance technologies,
                safety critical systems or systems affecting safety-critical functions,
                or technologies designed for the purpose of reducing the frequency of
                vehicle crashes. Off-cycle credits may not be earned for technologies
                installed on a motor vehicle to attain compliance with any vehicle
                safety standard or any regulation set forth in Title 49 of the Code of
                Federal Regulations. The manufacturer must use one of the three options
                specified in this section to determine the CO2 gram per mile
                credit applicable to an off-cycle technology. Note that the option
                provided in paragraph (b) of this section applies only to the 2014 and
                later model years. The manufacturer should notify EPA in their pre-
                model year report of their intention to generate any credits under this
                section.
                 (b) * * *
                 (1) * * *
                 (ix) High efficiency alternator. The credit for a high efficiency
                alternator for passenger automobiles and light trucks shall be
                calculated using the following equation, and rounded to the nearest 0.1
                grams/mile:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.761
                [[Page 25271]]
                Where:
                VDAHEA is the ratio of the alternator output power to the
                power supplied to the alternator, as measured using the Verband der
                Automobilindustrie (VDA) efficiency measurement methodology and
                expressed as a whole number percent from 68 to 100.
                * * * * *
                 (4) * * *
                 (xiii) High efficiency alternator means an alternator where the
                ratio of the alternator output power to the power supplied to the
                alternator is greater than 67 percent, as measured using the Verband
                der Automobilindustrie (VDA) efficiency measurement methodology.
                * * * * *
                 (d) * * *
                 (2) Notice and opportunity for public comment. (i) The
                Administrator will publish a notice of availability in the Federal
                Register notifying the public of a manufacturer's proposed alternative
                off-cycle credit calculation methodology. The notice will include
                details regarding the proposed methodology but will not include any
                Confidential Business Information. The notice will include instructions
                on how to comment on the methodology. The Administrator will take
                public comments into consideration in the final determination and will
                notify the public of the final determination. Credits may not be
                accrued using an approved methodology until the first model year for
                which the Administrator has issued a final approval.
                 (ii) The Administrator may waive these notice and comment
                requirements for technologies for which EPA has previously approved a
                methodology for determining credits. To qualify for this waiver, the
                new application must be substantially identical in form, content, and
                methodology to the application for a previously approved methodology,
                and must include the following:
                 (A) A cite to the appropriate previously approved methodology,
                including the appropriate Federal Register Notice and any subsequent
                EPA documentation of the Administrator's decision;
                 (B) All necessary manufacturer- and vehicle-specific test data,
                modeling, and credit calculations; and,
                 (C) Any other vehicle- or technology-specific details required
                pursuant to the previously approved methodology to assess and support
                an appropriate credit value.
                 (iii) A waiver of the notice and comment requirements does not
                imply a determination that a specific credit value for a given
                technology is appropriate, and nor does it imply a waiver from the
                requirements in paragraphs (d)(1) and (e) of this section.
                 (iv) The Administrator retains the option to require a notice and
                opportunity for public comment in cases where a new application
                deviates in significant respects from a previously approved methodology
                or raises novel substantive issues.
                * * * * *
                0
                6. Section 86.1870-12 is amended by revising paragraphs (a)(2) and
                (b)(2) to read as follows:
                Sec. 86.1870-12 CO2 credits for qualifying full-size light pickup
                trucks.
                * * * * *
                 (a) * * *
                 (2) Full size pickup trucks that are strong hybrid electric
                vehicles and that are produced in the 2017 through 2021 model years are
                eligible for a credit of 20 grams/mile. To receive this credit in a
                model year, the manufacturer must produce a quantity of strong hybrid
                electric full size pickup trucks such that the proportion of production
                of such vehicles, when compared to the manufacturer's total production
                of full size pickup trucks, is not less than 10 percent in that model
                year.
                * * * * *
                 (b) * * *
                 (2) Full size pickup trucks that are produced in the 2017 through
                2021 model years and that achieve carbon-related exhaust emissions less
                than or equal to the applicable target value determined in Sec.
                86.1818-12(c)(3) multiplied by 0.80 (rounded to the nearest gram/mile)
                in a model year are eligible for a credit of 20 grams/mile. A pickup
                truck that qualifies for this credit in a model year may claim this
                credit for a maximum of four subsequent model years (a total of five
                consecutive model years) if the carbon-related exhaust emissions of
                that pickup truck do not increase relative to the emissions in the
                model year in which the pickup truck first qualified for the credit.
                This credit may not be claimed in any model year after 2021. To qualify
                for this credit in a model year, the manufacturer must produce a
                quantity of full size pickup trucks that meet the emission requirements
                of this paragraph (b)(2) such that the proportion of production of such
                vehicles, when compared to the manufacturer's total production of full
                size pickup trucks, is not less than 10 percent in that model year.
                * * * * *
                PART 600--FUEL ECONOMY AND GREENHOUSE GAS EXHAUST EMISSIONS OF
                MOTOR VEHICLES
                0
                7. The authority citation for part 600 continues to read as follows:
                 Authority: 49 U.S.C. 32901--23919q, Pub. L. 109-58.
                0
                8. Section 600.113-12 is amended by revising paragraphs (n)
                introductory text, (n)(1), and (n)(3) to read as follows:
                Sec. 600.113-12 Fuel economy, CO2 emissions, and carbon-related
                exhaust emission calculations for FTP, HFET, US06, SC03 and cold
                temperature FTP tests.
                * * * * *
                 (n) Manufacturers shall determine CO2 emissions and
                carbon-related exhaust emissions for electric vehicles, fuel cell
                vehicles, and plug-in hybrid electric vehicles according to the
                provisions of this paragraph (n). Subject to the limitations on the
                number of vehicles produced and delivered for sale as described in
                Sec. 86.1866 of this chapter, the manufacturer may be allowed to use a
                value of 0 grams/mile to represent the emissions of fuel cell vehicles
                and the proportion of electric operation of a electric vehicles and
                plug-in hybrid electric vehicles that is derived from electricity that
                is generated from sources that are not onboard the vehicle, as
                described in paragraphs (n)(1) through (3) of this section. For
                purposes of labeling under this part, the CO2 emissions for
                electric vehicles shall be 0 grams per mile. Similarly, for purposes of
                labeling under this part, the CO2 emissions for plug-in
                hybrid electric vehicles shall be 0 grams per mile for the proportion
                of electric operation that is derived from electricity that is
                generated from sources that are not onboard the vehicle. For all 2027
                and later model year electric vehicles, fuel cell vehicles, and plug-in
                hybrid electric vehicles, the provisions of this paragraph (n) shall be
                used to determine the non-zero value for CREE for purposes of meeting
                the greenhouse gas emission standards described in Sec. 86.1818 of
                this chapter.
                 (1) For electric vehicles, but not including fuel cell vehicles,
                the carbon-related exhaust emissions in grams per mile is to be
                calculated using the following equation and rounded to the nearest one
                gram per mile:
                CREE = CREEUP - CREEGAS
                Where:
                CREE means the carbon-related exhaust emission value as defined in
                Sec. 600.002, which may be set equal to zero for eligible 2012
                through 2026 model year electric vehicles as described in Sec.
                86.1866-12(a) of this chapter.
                [[Page 25272]]
                [GRAPHIC] [TIFF OMITTED] TR30AP20.762
                [GRAPHIC] [TIFF OMITTED] TR30AP20.763
                Where:
                EC = The vehicle energy consumption in watt-hours per mile, for
                combined FTP/HFET operation, determined according to procedures
                established by the Administrator under Sec. 600.116-12.
                GRIDLOSS = 0.935 (to account for grid transmission losses).
                AVGUSUP = 0.534 (the nationwide average electricity greenhouse gas
                emission rate at the powerplant, in grams per watt-hour).
                2478 is the estimated grams of upstream greenhouse gas emissions per
                gallon of gasoline.
                8887 is the estimated grams of CO2 per gallon of
                gasoline.
                TargetCO2 = The CO2 Target Value for the fuel
                cell or electric vehicle determined according to Sec. 86.1818 of
                this chapter for the appropriate model year.
                * * * * *
                 (3) For 2012 and later model year fuel cell vehicles, the carbon-
                related exhaust emissions in grams per mile shall be calculated using
                the method specified in paragraph (n)(1) of this section, except that
                CREEUP shall be determined according to procedures established by the
                Administrator under Sec. 600.111-08(f). As described in Sec. 86.1866
                of this chapter, the value of CREE may be set equal to zero for 2012
                through 2026 model year fuel cell vehicles.
                * * * * *
                0
                9. Section 600.510-12 is amended by revising paragraphs (c)(2)(vi)
                introductory text, adding paragraph (c)(2)(vii) introductory text,
                revising the introductory text of paragraphs (c)(2)(vii)(B), (j)(2)(v),
                (vii)(A) and (vii)(B) to read as follows:
                Sec. 600.510-12 Calculation of average fuel economy and average
                carbon-related exhaust emissions.
                * * * * *
                 (c) * * *
                 (2) * * *
                 (vi) For natural gas dual fuel model types, for model years 1993
                through 2016, and optionally for 2021 and later model years, the
                harmonic average of the following two terms; the result rounded to the
                nearest 0.1 mpg:
                * * * * *
                 (vii) This paragraph (c)(2)(vii) applies to model year 2017 through
                2020 natural gas dual fuel model types. Model year 2021 and later
                natural gas dual fuel model types may use the provisions of paragraph
                (c)(2)(vi) of this section or this paragraph (c)(2)(vii).
                * * * * *
                 (B) Model year 2017 through 2020 natural gas dual fuel model types
                must meet the following criteria to qualify for use of a Utility Factor
                greater than 0.5:
                * * * * *
                 (j) * * *
                 (2) * * *
                 (v) For natural gas dual fuel model types, for model years 2012
                through 2015, and optionally for 2021 and later model years, the
                arithmetic average of the following two terms; the result rounded to
                the nearest gram per mile:
                * * * * *
                 (vii)(A) This paragraph (j)(2)(vii) applies to model year 2016
                through 2020 natural gas dual fuel model types. Model year 2021 and
                later natural gas dual fuel model types may use the provisions of
                paragraph (j)(2)(v) of this section or this paragraph (j)(2)(vii).
                * * * * *
                 (B) Model year 2016 through 2020 natural gas dual fuel model types
                must meet the following criteria to qualify for use of a Utility Factor
                greater than 0.5:
                * * * * *
                National Highway Transportation Administration
                Chapter V
                 For the reasons discussed in the preamble, the National Highway
                Traffic Safety Administration amends 49 CFR chapter V as follows:
                PART 523--VEHICLE CLASSIFICATION
                0
                10. The authority citation for part 523 continues to read as follows:
                 Authority: 49 U.S.C 32901; delegation of authority at 49 CFR
                1.95.
                0
                11. Amend Sec. 523.2 by revising the definitions of ``Curb weight''
                and ``Full-size pickup truck'' to read as follows:
                Sec. 523.2 Definitions.
                * * * * *
                 Curb weight has the meaning given in 40 CFR 86.1803-01.
                * * * * *
                 Full-size pickup truck means a light truck or medium duty passenger
                vehicle that meets the specifications in 40 CFR 86.1803-01.
                * * * * *
                PART 531--PASSENGER AUTOMOBILE AVERAGE FUEL ECONOMY STANDARDS
                0
                12. The authority citation for part 531 is revised to read as follows:
                 Authority: 49 U.S.C. 32902; delegation of authority at 49 CFR
                1.95.
                0
                13. Amend Sec. 531.5 by revising the introductory text of paragraph
                (c), Table III to paragraph (c), and paragraph (d), removing paragraph
                (e), and redesignating paragraph (f) as paragraph (e) to read as
                follows:
                Sec. 531.5 Fuel economy standards.
                * * * * *
                 (c) For model years 2012-2026, a manufacturer's passenger
                automobile fleet shall comply with the fleet average fuel economy level
                calculated for that model year according to this Figure 2 and the
                appropriate values in this Table III.
                * * * * *
                 Table III--Parameters for the Passenger Automobile Fuel Economy Targets, MYs 2012-2026
                ----------------------------------------------------------------------------------------------------------------
                 Parameters
                 ---------------------------------------------------------------
                 Model year c (gal/mi/
                 a (mpg) b (mpg) ft\2\) d (gal/mi)
                ----------------------------------------------------------------------------------------------------------------
                2012............................................ 35.95 27.95 0.0005308 0.006057
                2013............................................ 36.80 28.46 0.0005308 0.005410
                2014............................................ 37.75 29.03 0.0005308 0.004725
                2015............................................ 39.24 29.90 0.0005308 0.003719
                2016............................................ 41.09 30.96 0.0005308 0.002573
                [[Page 25273]]
                
                2017............................................ 43.61 32.65 0.0005131 0.001896
                2018............................................ 45.21 33.84 0.0004954 0.001811
                2019............................................ 46.87 35.07 0.0004783 0.001729
                2020............................................ 48.74 36.47 0.0004603 0.001643
                2021............................................ 49.48 37.02 0.000453 0.00162
                2022............................................ 50.24 37.59 0.000447 0.00159
                2023............................................ 51.00 38.16 0.000440 0.00157
                2024............................................ 51.78 38.74 0.000433 0.00155
                2025............................................ 52.57 39.33 0.000427 0.00152
                2026............................................ 53.37 39.93 0.000420 0.00150
                ----------------------------------------------------------------------------------------------------------------
                 (d) In addition to the requirements of paragraphs (b) and (c) of
                this section, each manufacturer shall also meet the minimum fleet
                standard for domestically manufactured passenger automobiles expressed
                in Table IV:
                 Table IV--Minimum Fuel Economy Standards for Domestically Manufactured
                 Passenger Automobiles, MYs 2011-2026
                ------------------------------------------------------------------------
                 Minimum
                 Model year standard
                ------------------------------------------------------------------------
                2011.................................................... 27.8
                2012.................................................... 30.7
                2013.................................................... 31.4
                2014.................................................... 32.1
                2015.................................................... 33.3
                2016.................................................... 34.7
                2017.................................................... 36.7
                2018.................................................... 38.0
                2019.................................................... 39.4
                2020.................................................... 40.9
                2021.................................................... 39.9
                2022.................................................... 40.6
                2023.................................................... 41.1
                2024.................................................... 41.8
                2025.................................................... 42.4
                2026.................................................... 43.1
                ------------------------------------------------------------------------
                * * * * *
                0
                14. Amend Sec. 531.6 by revising paragraphs (a) and (b) to read as
                follows:
                Sec. 531.6 Measurement and calculation procedures.
                 (a) The fleet average fuel economy performance of all passenger
                automobiles that are manufactured by a manufacturer in a model year
                shall be determined in accordance with procedures established by the
                Administrator of the Environmental Protection Agency under 49 U.S.C.
                32904 and set forth in 40 CFR part 600. For model years 2017 to 2026, a
                manufacturer is eligible to increase the fuel economy performance of
                passenger cars in accordance with procedures established by the EPA set
                forth in 40 CFR part 600, subpart F, including any adjustments to fuel
                economy the EPA allows, such as for fuel consumption improvements
                related to air conditioning efficiency and off-cycle technologies.
                 (1) A manufacturer that seeks to increase its fleet average fuel
                economy performance through the use of technologies that improve the
                efficiency of air conditioning systems must follow the requirements in
                40 CFR 86.1868-12. Fuel consumption improvement values resulting from
                the use of those air conditioning systems must be determined in
                accordance with 40 CFR 600.510-12(c)(3)(i).
                 (2) A manufacturer that seeks to increase its fleet average fuel
                economy performance through the use of off-cycle technologies must
                follow the requirements in 40 CFR 86.1869-12. A manufacturer is
                eligible to gain fuel consumption improvements for predefined off-cycle
                technologies in accordance with 40 CFR 86.1869-12(b) or for
                technologies tested using the EPA's 5-cycle methodology in accordance
                with 40 CFR 86.1869-12(c). The fuel consumption improvement is
                determined in accordance with 40 CFR 600.510-12(c)(3)(ii).
                 (b) A manufacturer is eligible to increase its fuel economy
                performance through use of an off-cycle technology requiring an
                application request made to the EPA in accordance with 40 CFR 86.1869-
                12(d). The request must be approved by the EPA in consultation with
                NHTSA. To expedite NHTSA's consultation with the EPA, a manufacturer
                shall concurrently submit its application to NHTSA if the manufacturer
                is seeking off-cycle fuel economy improvement values under the CAFE
                program for those technologies. For off-cycle technologies that are
                covered under 40 CFR 86.1869-12(d), NHTSA will consult with the EPA
                regarding NHTSA's evaluation of the specific off-cycle technology to
                ensure its impact on fuel economy and the suitability of using the off-
                cycle technology to adjust the fuel economy performance. NHTSA will
                provide its views on the suitability of the technology for that purpose
                to the EPA. NHTSA's evaluation and review will consider:
                 (1) Whether the technology has a direct impact upon improving fuel
                economy performance;
                 (2) Whether the technology is related to crash-avoidance
                technologies, safety critical systems or systems affecting safety-
                critical functions, or technologies designed for the purpose of
                reducing the frequency of vehicle crashes;
                 (3) Information from any assessments conducted by the EPA related
                to the application, the technology and/or related technologies; and
                 (4) Any other relevant factors.
                PART 533--LIGHT TRUCK FUEL ECONOMY STANDARDS
                0
                15. The authority citation for part 533 is revised to read as follows:
                 Authority: 49 U.S.C. 32902; delegation of authority at 49 CFR
                1.95.
                0
                16. In Sec. 533.5, amend paragraph (a) by revising Table VII and
                removing paragraph (k) to read as follows:
                Sec. 533.5 Requirements.
                 (a) * * *
                [[Page 25274]]
                 Table VII--Parameters for the Light Truck Fuel Economy Targets for MYs 2017-2026
                --------------------------------------------------------------------------------------------------------------------------------------------------------
                 Parameters
                 -------------------------------------------------------------------------------------------------------
                 Model year c (gal/mi/ g (gal/mi/
                 a (mpg) b (mpg) ft\2\) d (gal/mi) e (mpg) f (mpg) ft\2\) h (gal/mi)
                --------------------------------------------------------------------------------------------------------------------------------------------------------
                2017............................................ 36.26 25.09 0.0005484 0.005097 35.10 25.09 0.0004546 0.009851
                2018............................................ 37.36 25.20 0.0005358 0.004797 35.31 25.20 0.0004546 0.009682
                2019............................................ 38.16 25.25 0.0005265 0.004623 35.41 25.25 0.0004546 0.009603
                2020............................................ 39.11 25.25 0.0005140 0.004494 35.41 25.25 0.0004546 0.009603
                2021............................................ 39.71 25.63 0.000506 0.00443 NA NA NA NA
                2022............................................ 40.31 26.02 0.000499 0.00436 NA NA NA NA
                2023............................................ 40.93 26.42 0.000491 0.00429 NA NA NA NA
                2024............................................ 41.55 26.82 0.000484 0.00423 NA NA NA NA
                2025............................................ 42.18 27.23 0.000477 0.00417 NA NA NA NA
                2026............................................ 42.82 27.64 0.000469 0.00410 NA NA NA NA
                --------------------------------------------------------------------------------------------------------------------------------------------------------
                * * * * *
                0
                17. Amend Sec. 533.6 by revising paragraphs (b) and (c) to read as
                follows:
                Sec. 533.6 Measurement and calculation procedures.
                * * * * *
                 (b) The fleet average fuel economy performance of all light trucks
                that are manufactured by a manufacturer in a model year shall be
                determined in accordance with procedures established by the
                Administrator of the Environmental Protection Agency under 49 U.S.C.
                32904 and set forth in 40 CFR part 600. For model years 2017 to 2026, a
                manufacturer is eligible to increase the fuel economy performance of
                light trucks in accordance with procedures established by the EPA set
                forth in 40 CFR part 600, subpart F, including any adjustments to fuel
                economy the EPA allows, such as for fuel consumption improvements
                related to air conditioning efficiency, off-cycle technologies, and
                hybridization and other performance-based technologies for full-size
                pickup trucks that meet the requirements specified in 40 CFR 86.1803.
                 (1) A manufacturer that seeks to increase its fleet average fuel
                economy performance through the use of technologies that improve the
                efficiency of air conditioning systems must follow the requirements in
                40 CFR 86.1868-12. Fuel consumption improvement values resulting from
                the use of those air conditioning systems must be determined in
                accordance with 40 CFR 600.510-12(c)(3)(i).
                 (2) A manufacturer that seeks to increase its fleet average fuel
                economy performance through the use of off-cycle technologies must
                follow the requirements in 40 CFR 86.1869-12. A manufacturer is
                eligible to gain fuel consumption improvements for predefined off-cycle
                technologies in accordance with 40 CFR 86.1869-12(b) or for
                technologies tested using the EPA's 5-cycle methodology in accordance
                with 40 CFR 86.1869-12(c). The fuel consumption improvement is
                determined in accordance with 40 CFR 600.510-12(c)(3)(ii).
                 (3) The eligibility of a manufacturer to increase its fuel economy
                using hybridized and other performance-based technologies for full-size
                pickup trucks must follow 40 CFR 86.1870-12 and the fuel consumption
                improvement of these full-size pickup truck technologies must be
                determined in accordance with 40 CFR 600.510-12(c)(3)(iii).
                 (c) A manufacturer is eligible to increase its fuel economy
                performance through use of an off-cycle technology requiring an
                application request made to the EPA in accordance with 40 CFR 86.1869-
                12(d). The request must be approved by the EPA in consultation with
                NHTSA. To expedite NHTSA's consultation with the EPA, a manufacturer
                shall concurrently submit its application to NHTSA if the manufacturer
                is seeking off-cycle fuel economy improvement values under the CAFE
                program for those technologies. For off-cycle technologies that are
                covered under 40 CFR 86.1869-12(d), NHTSA will consult with the EPA
                regarding NHTSA's evaluation of the specific off-cycle technology to
                ensure its impact on fuel economy and the suitability of using the off-
                cycle technology to adjust the fuel economy performance. NHTSA will
                provide its views on the suitability of the technology for that purpose
                to the EPA. NHTSA's evaluation and review will consider:
                 (1) Whether the technology has a direct impact upon improving fuel
                economy performance;
                 (2) Whether the technology is related to crash-avoidance
                technologies, safety critical systems or systems affecting safety-
                critical functions, or technologies designed for the purpose of
                reducing the frequency of vehicle crashes;
                 (3) Information from any assessments conducted by the EPA related
                to the application, the technology and/or related technologies; and
                 (4) Any other relevant factors.
                PART 535--MEDIUM- AND HEAVY-DUTY VEHICLE FUEL EFFICIENCY PROGRAM
                0
                18. The authority citation for part 535 continues to read as follows:
                 Authority: 49 U.S.C. 32902 and 30101; delegation of authority
                at 49 CFR 1.95.
                0
                19. Amend Sec. 535.6 by revising paragraphs (a)(4)(ii) and (d)(5)(ii)
                to read as follows:
                Sec. 535.6 Measurement and calculation procedures.
                * * * * *
                 (a) * * *
                 (4) * * *
                 (ii) Calculate the equivalent fuel consumption test group results
                as follows for spark-ignition vehicles and alternative fuel spark-
                ignition vehicles. CO2 emissions test group result (grams
                per mile)/((8,887 grams per gallon of gasoline fuel) x
                (10-2)) = Fuel consumption test group result (gallons per
                100 mile).
                * * * * *
                 (d) * * *
                 (5) * * *
                 (ii) Calculate equivalent fuel consumption FCL values for spark-
                ignition engines and alternative fuel spark-ignition engines.
                CO2 FCL value (grams per hp-hr)/((8,887 grams per gallon of
                gasoline fuel) x (10-2)) = Fuel consumption FCL value
                (gallons per 100 hp-hr).
                * * * * *
                0
                20. Amend Sec. 535.7 by revising the equations in paragraphs (b)(1),
                (c)(1), (d)(1), (e)(2), and (f)(2)(iii)(E) to read as follows:
                [[Page 25275]]
                Sec. 535.7 Averaging, banking, and trading (ABT) credit program.
                * * * * *
                 (b) * * *
                 (1) * * *
                Total MY Fleet FCC (gallons) = (Std - Act) x (Volume) x (UL) x
                (10-2)
                Where:
                Std = Fleet average fuel consumption standard (gal/100 mile).
                Act = Fleet average actual fuel consumption value (gal/100 mile).
                Volume = the total U.S.-directed production of vehicles in the
                regulatory subcategory.
                UL = the useful life for the regulatory subcategory. The useful life
                value for heavy-pickup trucks and vans manufactured for model years
                2013 through 2020 is equal to the 120,000 miles. The useful life for
                model years 2021 and later is equal to 150,000 miles.
                * * * * *
                 (c) * * *
                 (1) * * *
                Vehicle Family FCC (gallons) = (Std - FEL) x (Payload) x (Volume) x
                (UL) x (10-\3\)
                Where:
                Std = the standard for the respective vehicle family regulatory
                subcategory (gal/1000 ton-mile).
                FEL = family emissions limit for the vehicle family (gal/1000 ton-
                mile).
                Payload = the prescribed payload in tons for each regulatory
                subcategory as shown in the following table:
                ------------------------------------------------------------------------
                 Payload
                 Regulatory subcategory (tons)
                ------------------------------------------------------------------------
                Vocational LHD Vehicles................................. 2.85
                Vocational MHD Vehicles................................. 5.60
                Vocational HHD Vehicles................................. 7.5
                MDH Tractors............................................ 12.50
                HHD Tractors, other than heavy-haul Tractors............ 19.00
                Heavy-haul Tractors..................................... 43.00
                ------------------------------------------------------------------------
                Volume = the number of U.S.-directed production volume of vehicles
                in the corresponding vehicle family.
                UL = the useful life for the regulatory subcategory (miles) as shown
                in the following table:
                ------------------------------------------------------------------------
                 Regulatory subcategory UL (miles)
                ------------------------------------------------------------------------
                LHD Vehicles.............................. 110,000 (Phase 1).
                 150,000 (Phase 2).
                Vocational MHD Vehicles and tractors at or 185,000.
                 below 33,000 pounds GVWR.
                Vocation HHD Vehicles and tractors at or 435,000.
                 above 33,000 pounds GVWR.
                ------------------------------------------------------------------------
                * * * * *
                 (d) * * *
                 (1) * * *
                Engine Family FCC (gallons) = (Std - FCL) x (CF) x (Volume) x (UL) x
                (10-2)
                Where:
                Std = the standard for the respective engine regulatory subcategory
                (gal/100 hp-hr).
                FCL = family certification level for the engine family (gal/100 hp-
                hr).
                CF= a transient cycle conversion factor in hp-hr/mile which is the
                integrated total cycle horsepower-hour divided by the equivalent
                mileage of the applicable test cycle. For engines subject to spark-
                ignition heavy-duty standards, the equivalent mileage is 6.3 miles.
                For engines subject to compression-ignition heavy-duty standards,
                the equivalent mileage is 6.5 miles.
                Volume = the number of engines in the corresponding engine family.
                UL = the useful life of the given engine family (miles) as shown in
                the following table:
                ------------------------------------------------------------------------
                 Regulatory subcategory UL (miles)
                ------------------------------------------------------------------------
                SI and CI LHD Engines..................... 120,000 (Phase 1).
                 150,000 (Phase 2).
                CI MHD Engines............................ 185,000.
                CI HHD Engines............................ 435,000.
                ------------------------------------------------------------------------
                * * * * *
                 (e) * * *
                 (2) * * *
                Vehicle Family FCC (gallons) = (Std - FEL) x (Payload) x (Volume) x
                (UL) x (10-3)
                Where:
                Std = the standard for the respective vehicle family regulatory
                subcategory (gal/1000 ton-mile).
                FEL = family emissions limit for the vehicle family (gal/1000 ton-
                mile).
                Payload = 10 tons for short box vans and 19 tons for other trailers.
                Volume = the number of U.S.-directed production volume of vehicles
                in the corresponding vehicle family.
                UL = the useful life for the regulatory subcategory. The useful life
                value for heavy-duty trailers is equal to 250,000 miles.
                * * * * *
                 (f) * * *
                 (2) * * *
                 (iii) * * *
                 (E) * * *
                Off-cycle FC credits = (CO2 Credit/CF) x Production x VLM
                Where:
                CO2 Credits = the credit value in grams per mile
                determined in 40 CFR 86.1869-12(c)(3), (d)(1), (d)(2) or (d)(3).
                CF = conversion factor, which for spark-ignition engines is 8,887
                and for compression-ignition engines is 10,180.
                Production = the total production volume for the applicable category
                of vehicles
                VLM = vehicle lifetime miles, which for 2b-3 vehicles shall be
                150,000 for the Phase 2 program.
                The term (CO2 Credit/CF) should be rounded to the nearest
                0.0001
                * * * * *
                PART 536--TRANSFER AND TRADING OF FUEL ECONOMY CREDITS
                0
                21. The authority citation for part 536 is revised to read as follows:
                 Authority: 49 U.S.C. 32903; delegation of authority at 49 CFR
                1.95.
                0
                22. Amend Sec. 536.4 by revising paragraph (c) to read as follows:
                Sec. 536.4 Credits.
                * * * * *
                 (c) Adjustment factor. When traded or transferred and used, fuel
                economy credits are adjusted to ensure fuel oil savings is preserved.
                For traded credits, the user (or buyer) must multiply the calculated
                adjustment factor by the number of shortfall credits it plans to offset
                in order to determine the number of equivalent credits to acquire from
                the earner (or seller). For transferred credits, the user of credits
                must multiply the calculated adjustment factor by the number of
                shortfall credits it plans to offset in order to determine the number
                of equivalent credits to transfer from the compliance category holding
                the available credits. The adjustment factor is calculated according to
                the following formula:
                [GRAPHIC] [TIFF OMITTED] TR30AP20.764
                Where:
                A = Adjustment factor applied to traded and transferred credits. The
                quotient shall be rounded to 4 decimal places;
                * * * * *
                0
                23. Amend Sec. 536.5 by revising paragraphs (c) and (d)(6) to read as
                follows:
                [[Page 25276]]
                Sec. 536.5 Trading infrastructure.
                * * * * *
                 (c) Automatic debits and credits of accounts.
                 (1) To carry credits forward, backward, transfer credits, or trade
                credits into other credit accounts, a manufacturer or credit holder
                must submit a credit instruction to NHTSA. A credit instruction must
                detail and include:
                 (i) The credit holder(s) involved in the transaction.
                 (ii) The originating credits described by the amount of the
                credits, compliance category and the vintage of the credits.
                 (iii) The recipient credit account(s) for banking or applying the
                originating credits described by the compliance category(ies), model
                year(s), and if applicable the adjusted credit amount(s) and adjustment
                factor(s).
                 (iv) For trades, a contract authorizing the trade signed by the
                manufacturers or credit holders or by managers legally authorized to
                obligate the sale and purchase of the traded credits.
                 (2) Upon receipt of a credit instruction from an existing credit
                holder, NHTSA verifies the presence of sufficient credits in the
                account(s) of the credit holder(s) involved as applicable and notifies
                the credit holder(s) that the credits will be debited from and/or
                credited to the accounts involved, as specified in the credit
                instruction. NHTSA determines if the credits can be debited or credited
                based upon the amount of available credits, accurate application of any
                adjustment factors and the credit requirements prescribed by this part
                that are applicable at the time the transaction is requested.
                 (3) After notifying the credit holder(s), all accounts involved are
                either credited or debited, as appropriate, in line with the credit
                instruction. Traded credits identified by a specific compliance
                category are deposited into the recipient's account in that same
                compliance category and model year. If a recipient of credits as
                identified in a credit instruction is not a current account holder,
                NHTSA establishes the credit recipient's account, subject to the
                conditions described in Sec. 536.5(b), and adds the credits to the
                newly-opened account.
                 (4) NHTSA will automatically delete unused credits from holders'
                accounts when those credits reach their expiry date.
                 (5) Starting in model year 2021, manufacturers or credit holders
                issuing credit instructions or providing credit allocation plans as
                specified in Sec. 536.5(d), must use the NHTSA Credit Template
                fillable form (OMB Control No. 2127-0019, NHTSA Form 1475). The NHTSA
                Credit Template is available for download on NHTSA's website. If a
                credit instruction includes a trade, the NHTSA Credit Template must be
                signed by managers legally authorized to obligate the sale and/or
                purchase of the traded credits from both parties to the trade. The
                NHTSA Credit Template signed by both parties to the trade serves as an
                acknowledgement that the parties have agreed to trade credits, and does
                not dictate terms, conditions, or other business obligations of the
                parties. All parties trading credits must also provide NHTSA the price
                paid for the credits including a description of any other monetary or
                non-monetary terms affecting the price of the traded credits, such as
                any technology exchanged or shared for the credits, any other non-
                monetary payment for the credits, or any other agreements related to
                the trade. Manufacturers must submit this information to NHTSA in a PDF
                document along with the Credit Template through the CAFE email,
                [email protected]. NHTSA reserves the right to request additional
                information from the parties regarding the terms of the trade.
                 (6) NHTSA will consider claims that information submitted to the
                agency under this section is entitled to confidential treatment under 5
                U.S.C. 552(b) and under the provisions of part 512 of this chapter if
                the information is submitted in accordance with the procedures of that
                part.
                * * * * *
                 (d) * * *
                 (6) Credit allocation plans received from a manufacturer will be
                reviewed and approved by NHTSA. Starting in model year 2021, use the
                NHTSA Credit Template (OMB Control No. 2127-0019, NHTSA Form 1475) to
                record the credit transactions requested in the credit allocation plan.
                The template is a fillable form that has an option for recording and
                calculating credit transactions for credit allocation plans. The
                template calculates the required adjustments to the credits. The credit
                allocation plan and the completed transaction template must be
                submitted to NHTSA. NHTSA will approve the credit allocation plan
                unless it finds that the proposed credits are unavailable or that it is
                unlikely that the plan will result in the manufacturer earning
                sufficient credits to offset the subject credit shortfall. If the plan
                is approved, NHTSA will revise the respective manufacturer's credit
                account accordingly. If the plan is rejected, NHTSA will notify the
                respective manufacturer and request a revised plan or payment of the
                appropriate fine.
                PART 537--AUTOMOTIVE FUEL ECONOMY REPORTS
                0
                24. The authority citation for part 537 is revised to read as follows:
                 Authority: 49 U.S.C. 32907; delegation of authority at 49 CFR
                1.95.
                0
                25. Amend Sec. 537.5 by redesignating paragraph (d) as paragraph (e)
                and adding a new paragraph (d) to read as follows:
                Sec. 537.5 General requirements for reports.
                * * * * *
                 (d) Beginning with model year 2023, each manufacturer shall
                generate reports required by this part using the NHTSA CAFE Projections
                Reporting Template (OMB Control No. 2127-0019, NHTSA Form 1474). The
                template is a fillable form.
                 (1) Select the option to identify the report as a pre-model year
                report, mid-model year report, or supplementary report as appropriate;
                 (2) Complete all required information for the manufacturer and for
                all vehicles produced for the current model year required to comply
                with CAFE standards. Identify the manufacturer submitting the report,
                including the full name, title, and address of the official responsible
                for preparing the report and a point of contact to answer questions
                concerning the report.
                 (3) Use the template to generate confidential and non-confidential
                reports for all the domestic and import passenger cars and light truck
                fleet produced by the manufacturer for the current model year.
                Manufacturers must submit a request for confidentiality in accordance
                with part 512 of this chapter to withhold projected production sales
                volume estimates from public disclosure. If the request is granted,
                NHTSA will withhold the projected production sales volume estimates
                from public disclose until all the vehicles produced by the
                manufacturer have been made available for sale (usually one year after
                the current model year).
                 (4) Submit confidential reports and requests for confidentiality to
                NHTSA on CD-ROM in accordance with Part 537.12. Email copies of non-
                confidential (i.e., redacted) reports to NHTSA's secure email address:
                [email protected]. Requests for confidentiality must be submitted in a PDF
                or MS Word format. Submit 2 copies of the CD-ROM to: Administrator,
                National Highway Traffic Administration, 1200 New Jersey Avenue SE,
                Washington, DC 20590, and submit emailed reports electronically to
                [[Page 25277]]
                the following secure email address: [email protected];
                 (5) Confidentiality Requests. Manufacturers can withhold
                information on projected production sales volumes under 5 U.S.C.
                552(b)(4) and 15 U.S.C. 2005(d)(1). In accordance, the manufacturer
                must:
                 (i) Show that the item is within the scope of sections 552(b)(4)
                and 2005(d)(1);
                 (ii) Show that disclosure of the item would result in significant
                competitive damage;
                 (iii) Specify the period during which the item must be withheld to
                avoid that damage; and
                 (iv) Show that earlier disclosure would result in that damage.
                * * * * *
                0
                26. Amend Sec. 537.6 by revising paragraphs (b) and (c) to read as
                follows:
                Sec. 537.6 General content of reports.
                * * * * *
                 (b) Supplementary report. Except as provided in paragraph (c) of
                this section, each supplementary report for each model year must
                contain the information required by Sec. 537.7(a)(1) and (a)(2), as
                appropriate for the vehicle fleets produced by the manufacturer, in
                accordance with Sec. 537.8(b)(1), (2), (3), and (4) as appropriate.
                 (c) Exceptions. The pre-model year report, mid-model year report,
                and supplementary report(s) submitted by an incomplete automobile
                manufacturer for any model year are not required to contain the
                information specified in Sec. 537.7 (c)(4) (xv) through (xviii) and
                (c)(5). The information provided by the incomplete automobile
                manufacturer under Sec. 537.7(c) shall be according to base level
                instead of model type or carline.
                0
                27. Amend Sec. 537.7 by revising paragraph (a) to read as follows:
                Sec. 537.7 Pre-model year and mid-model year reports.
                 (a)(1) Provide a report with the information required by paragraphs
                (b) and (c) of this section for each domestic and import passenger
                automobile fleet, as specified in part 531 of this chapter, for the
                current model year.
                 (2) Provide a report with the information required by paragraphs
                (b) and (c) of this section for each light truck fleet, as specified in
                part 533 of this chapter, for the current model year.
                 (3) For model year 2023 and later, provide the information required
                by paragraphs (a)(1) and (2) of this section for pre-model and mid-
                model year reports in accordance with the NHTSA CAFE Projections
                Reporting Template (OMB Control No. 2127-0019, NHTSA Form 1474). The
                required reporting template can be downloaded from NHTSA's website.
                * * * * *
                0
                28. Amend Sec. 537.7 by revising paragraphs (b)(3), (c)(1), (c)(3),
                (c)(7)(i), (c)(7)(ii), and (c)(7)(iii) to read as follows:
                Sec. 537.7 Pre-model year and mid-model year reports.
                * * * * *
                 (b) * * *
                 (3) State the projected required fuel economy for the
                manufacturer's passenger automobiles and light trucks determined in
                accordance with Sec. Sec. 531.5(c) and 533.5 of this chapter and based
                upon the projected sales figures provided under paragraph (c)(2) of
                this section. For each unique model type and footprint combination of
                the manufacturer's automobiles, provide the information specified in
                paragraph (b)(3)(i) and (ii) of this section in tabular form. List the
                model types in order of increasing average inertia weight from top to
                bottom down the left side of the table and list the information
                categories in the order specified in paragraphs (b)(3)(i) and (ii) of
                this section from left to right across the top of the table. Other
                formats, such as those accepted by the EPA, which contain all the
                information in a readily identifiable format are also acceptable. For
                model year 2023 and later, for each unique model type and footprint
                combination of the manufacturer's automobiles, provide the information
                specified in paragraph (b)(3)(i) and (ii) of this section in accordance
                with the CAFE Projections Reporting Template (OMB Control No. 2127-
                0019, NHTSA Form 1474).
                 (i) In the case of passenger automobiles:
                 (A) Beginning model year 2013, base tire as defined in Sec. 523.2
                of this chapter,
                 (B) Beginning model year 2013, front axle, rear axle, and average
                track width as defined in Sec. CFR 523.2 of this chapter,
                 (C) Beginning model year 2013, wheelbase as defined in Sec. 523.2
                of this chapter, and
                 (D) Beginning model year 2013, footprint as defined in Sec. 523.2
                of this chapter.
                 (E) The fuel economy target value for each unique model type and
                footprint entry listed in accordance with the equation provided in part
                531 of this chapter.
                 (ii) In the case of light trucks:
                 (A) Beginning model year 2013, base tire as defined in Sec. 523.2
                of this chapter,
                 (B) Beginning model year 2013, front axle, rear axle, and average
                track width as defined in Sec. 523.2 of this chapter,
                 (C) Beginning model year 2013, wheelbase as defined in Sec. 523.2
                of this chapter, and
                 (D) Beginning model year 2013, footprint as defined in Sec. 523.2
                of this chapter.
                 (E) The fuel economy target value for each unique model type and
                footprint entry listed in accordance with the equation provided in part
                533 of this chapter.
                * * * * *
                 (c) * * *
                 (1) For each model type of the manufacturer's automobiles, provide
                the information specified in paragraph (c)(2) of this section in
                tabular form. List the model types in order of increasing average
                inertia weight from top to bottom down the left side of the table and
                list the information categories in the order specified in paragraph
                (c)(2) of this section from left to right across the top of the table.
                For model year 2023 and later, CAFE reports required by part 537 of
                this chapter, shall for each model type of the manufacturer's
                automobiles, provide the information in specified in paragraph (c)(2)
                of this section in accordance with the NHTSA CAFE Projections Reporting
                Template (OMB Control No. 2127-0019, NHTSA Form 1474) and list the
                model types in order of increasing average inertia weight from top to
                bottom.
                * * * * *
                 (3) (Pre-model year reports only through model year 2022.) For each
                vehicle configuration whose fuel economy was used to calculate the fuel
                economy values for a model type under paragraph (c)(2) of this section,
                provide the information specified in paragraph (c)(4) of this section
                in accordance with the NHTSA CAFE Projections Reporting Template (OMB
                Control No. 2127-0019, NHTSA Form 1474).
                * * * * *
                 (7) * * *
                 (i) Provide a list of each air conditioning efficiency improvement
                technology utilized in your fleet(s) of vehicles for each model year.
                For each technology identify vehicles by make and model types that have
                the technology, which compliance category those vehicles belong to and
                the number of vehicles for each model equipped with the technology. For
                each compliance category (domestic passenger car, import passenger car,
                and light truck), report the air conditioning fuel consumption
                improvement value in gallons/mile in accordance with the equation
                specified in 40 CFR 600.510-12(c)(3)(i).
                 (ii) Provide a list of off-cycle efficiency improvement
                technologies
                [[Page 25278]]
                utilized in your fleet(s) of vehicles for each model year that is
                pending or approved by the EPA. For each technology identify vehicles
                by make and model types that have the technology, which compliance
                category those vehicles belong to, the number of vehicles for each
                model equipped with the technology, and the associated off-cycle
                credits (grams/mile) available for each technology. For each compliance
                category (domestic passenger car, import passenger car, and light
                truck), calculate the fleet off-cycle fuel consumption improvement
                value in gallons/mile in accordance with the equation specified in 40
                CFR 600.510-12(c)(3)(ii).
                 (iii) Provide a list of full-size pickup trucks in your fleet that
                meet the mild and strong hybrid vehicle definitions. For each mild and
                strong hybrid type, identify vehicles by make and model types that have
                the technology, the number of vehicles produced for each model equipped
                with the technology, the total number of full-size pickup trucks
                produced with and without the technology, the calculated percentage of
                hybrid vehicles relative to the total number of vehicles produced, and
                the associated full-size pickup truck credits (grams/mile) available
                for each technology. For the light truck compliance category, calculate
                the fleet pickup truck fuel consumption improvement value in gallons/
                mile in accordance with the equation specified in 40 CFR 600.510-
                12(c)(3)(iii).
                * * * * *
                0
                29. Amend Sec. 537.8 by revising paragraph (a)(3), adding paragraphs
                (a)(4) and (b)(4), and revising paragraph (c)(1) to read as follows:
                Sec. 537.8 Supplementary reports.
                 (a) * * *
                 (3) For model years through 2022, each manufacturer whose pre-model
                or mid-model year report omits any of the information specified in
                Sec. 537.7(b) or (c) shall file a supplementary report containing the
                information specified in paragraph (b)(3) of this section. Starting
                model year 2023, each manufacturer whose pre-model or mid-model year
                report omits any of the information shall resubmit the information with
                other information required in accordance with the NHTSA CAFE
                Projections Reporting Template (OMB Control No. 2127-0019, NHTSA Form
                1474).
                 (b) * * *
                 (4) The supplementary report required by paragraph (a)(4) of this
                section must contain:
                 (i) All information omitted from the pre-model or mid-model year
                reports under Sec. 537.6(c)(2); and
                 (ii) Such revisions of and additions to the information submitted
                by the manufacturer in its pre-model or mid-model year reports
                regarding the automobiles produced during the current model year as are
                necessary to reflect the information provided under paragraph (b)(4)(i)
                of this section.
                 (c)(1) Each report required by paragraphs (a)(1), (2), (3), or (4)
                of this section must be submitted in accordance with Sec. 537.5(c) not
                more than 45 days after the date on which the manufacturer determined,
                or could have determined with reasonable diligence, that the report was
                required.
                * * * * *
                 Dated: March 30, 2020.
                Andrew Wheeler,
                Administrator, Environmental Protection Agency.
                 Issued on March 30, 2020 in Washington, DC, under authority
                delegated in 49 CFR 1.95 and 501.5
                James Clayton Owens,
                Acting Administrator, National Highway Traffic Safety Administration.
                [FR Doc. 2020-06967 Filed 4-20-20; 4:15 pm]
                BILLING CODE 4910-59-P
                

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT